Room Acoustics Forum by Rives Audio

Welcome! Need support, you got it. Or share you ideas and experiences.

Return to Room Acoustics Forum by Rives Audio

Message Sort: Post Order or Asylum Reverse Threaded

stuck with a cube shaped room

70.218.44.67
Posted on January 16, 2008 at 06:08:00
jimgrocecfp@hotmail.com
 The worst possible room I guess, 12X12X12. Interested in slowly adding room treatment. So far I have Corner Busters in all 4 corners where ceiling meets wall, and I have a book case app 5' tall in each corner on the front wall.
I have had my listening seat against the real wall until recently, when I move it out app. 18" into the room, and also moved the speakers back by the same amount. So currently the speakers are about 3' out from front wall, 2.5' from side walls, app. 6.5' apart, and I sit about 6' away. Not entirely sure I like this better than sitting against the back wall.
Besides the room treatments I have plans to upgrade the electronics late this year. Photos can be seen at
http://forum.audiogon.com/cgi-bin/fr.pl?vdone&1183773812&read&3&4&
So, do I treat the front wall behind the speakers? Side walls? Move the book cases and get bass traps? Would love to hear what you think.
Jim

 

Hide full thread outline!
    ...
RE: stuck with a cube shaped room, posted on January 16, 2008 at 13:01:29
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 A cube shaped room will have one single set of extremely strong axial resonances since each axis of the room has the same dimension.

I would strongly recommend bass traps in the corners, most definitely in the corners behind the speakers if you can only manage 2 corners but preferably in all 4 where the room allows it, and floor to ceiling if you can manage them. You've got a door in one corner by the look of the photos and that makes bass traps in that corner difficult. You could use stand mounted panel traps placed diagonally across the corner with the door closed but you would have to be prepared to move them each time some one comes into or leaves the room which is a pain. If you leave that corner without traps you may wish to leave the other corner behind you without them as well for the sake of balance.

If you can't fit traps into the two wall corners behind you, try fitting them into the wall ceiling corners there. Note that I'm not talking simple corner treatments, I'm talking about sizeable bass traps running lengthwise along the walls at the ceiling interface there.

Bass traps would be my primary treatment suggestion. I would also recommend treatment of the first reflection points but there's certainly an element of preference as to how you like recordings to sound when it comes to that. I would definitely recommend that you either listen to a system in a room where the first reflection points have been treated or try it yourself with some makeshift absorption at the first reflection points of your room to see whether you like it. I'd recommend free standing panels for first reflection points since their location will change with changes to speaker and listening position placement and free standing panels make dealing with such changes easy. You can use a thick rug on the floor for the floor to treat the floor reflection points and you may get some benefit from using a rug as well as the existing carpet. Ceilings are trickier since you probably need to use relatively permanent attachment to any reflection treatments there so make sure you've finalised your speaker and listening position placements before treating the ceiling points if you decide to do that.

If bass traps don't smooth the room response sufficiently, with or without treatment at the first reflection points depending on your preferences there, then I'd add some large areas of absorption capable of broad band absorption, especially in the bass region. You will get more benefit from the same area of absorption if areas on opposite walls do not line up exactly opposite to each other so asymmetric placement on opposite walls is a good idea. Consider using areas of absorption at least 4' x 2' in size at each point, and preferably at least 3' wide or so since area certainly helps improve bass absorption but the thickness and density of the materials used are the primary factors in getting decent bass absorption.

Some diffusion would probably help also given the relatively small (with the exception of ceiling height) room dimensions.

You've obviously gone to some effort to create a good looking room and acoustic treatments are highly visible. You can try matching the colour of the fabric covers to the walls in order to reduce the visual impact. Unfortunately as a rule of thumb the effectiveness of acoustic treatments depends strongly on the surface area and small, visually unobtrusive treatments are going to be less effective than larger ones which are harder to make visually unobtrusive. It's a pity in many ways but size does count here.



David Aiken

 

RE: stuck with a cube shaped room, posted on January 17, 2008 at 07:52:02
jimgrocecfp@hotmail.com
 Thanks very much for the detailed reply. Will probably start with bass traps in the front corners. I have two more questions if you don't mind;

- what is it about treating 1st reflection points that I may not like? I really like detail, imaging, and really enjoy hearing the soundstage extend outside the speakers. I guess I like everything! (incidentally I hear the soundstage often extending on the right side where the boundary is bare drywall, almost never on the left where the boundary is a large blinds-covered window)

- a more loaded question....is there any particular brand I should consider? echo buster vs realtraps vs eigth nerve etc.

 

RE: stuck with a cube shaped room, posted on January 17, 2008 at 10:17:24
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 Jim,

> echo buster vs realtraps vs eigth nerve etc. <

As an owner of RealTraps I don't usually chime in on questions like this, but this one is a no-brainer. The best bass trap is the one with the best absorption specs at bass frequencies.

--Ethan

 

RE: stuck with a cube shaped room, posted on January 17, 2008 at 12:45:59
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 "what is it about treating 1st reflection points that I may not like? I really like detail, imaging, and really enjoy hearing the soundstage extend outside the speakers. I guess I like everything! (incidentally I hear the soundstage often extending on the right side where the boundary is bare drywall, almost never on the left where the boundary is a large blinds-covered window)"

Well, not everyone likes the things you like. Some people are distracted by the soundstage and imaging and prefer them minimised rather than maximised. Pick any parameter and give people a choice and some will go for each option. In this case treating the first reflection points will most certainly enhance the soundstage and imaging you get, and let you hear a lot more of what the people who mixed the recording intended you to hear but their preferences don't have to agree with your preferences and you're the person your system has to satisfy. If you appreciate what a good soundstage and imaging adds to things, then definitely treat the reflection points. If you don't like those things, then don't use absorption at the reflection points.

As for the reason for your L-R difference, I suspect what may be going on is that you've got a stronger reflection on the right side which will tend to pull the overall balance a little to the right. It may even be strong enough to cause image width to broaden towards the wall. If you use an absorptive panel on each side so that reflections on both sides are treated equally, you should get a similar effect on each side. As far as images to the outside of the speaker go, a fair bit depends on the recording and some recordings deliver a lot more there than do others. Don't expect to get the same results with every disc but you should get imaging to the outside of the speakers with a lot of discs.

As to what brands, I'm in Australia and while the brands you mention are all available here, they're only available from dealers in other states and I've never had a chance to do any sort of comparison. All of my treatments at present are DIY but I've recently ordered RealTraps to replace them. Apart from comparative data on Ethan's RealTraps site I really have no idea how any of these products compare.

I've also never seen anyone do an A/B type comparison of different room treatments. All of the reviews I've read have contrasted a specific treatment with the sound of the previously bare room. At least in several weeks I should be able to do my own A/B comparison of my DIY vs RealTraps but that will really be of academic interest since there's no real uniformity when it comes to DIY so the comparison may not be too useful to others.

I confess to being cautious about the 8th Nerve products. As I said in my original post, size counts at low frequencies. The 8th Nerve products don't have the size I expect in products that are effective at low frequencies and they claim that their products work on a different basis to absorption. Absorption is the bass treatment method I understand so I gravitate towards traditional treatment products rather than those based on things that I don't understand and which offer explanations that are often not too clear.

In a recent response Ethan said to a poster that there were 3 factors with room treatment products: size, effectiveness and attractiveness, and that you could get any combination of 2 of them in a product but not all 3. I think he's right with the possible single exception that I can't think of an attractive and large product that's ineffective because I can't think of an attractive large product. When it comes to large products I think the best you can hope for is that some are less visually intrusive than others in a particular room. Which product is less intrusive will depend on the appearance of the room as well as that of the treatment product and also on the purchaser's own preferences regarding visual styling.




David Aiken

 

You don't need to treat reflections!!!, posted on January 18, 2008 at 21:52:36
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 Floyd Toole, arguably the world's leading expert when it comes to the loudspeaker/room interface, says that reflections are not problems. You either believe recommendations based on incorrect and inexisting evidence or you believe someone who spent his entire life investigating these matters.

Toole, “Loudspeakers and rooms for sound reproduction – a scientific review”, J. of the Audio Engineering Society 2008, p.451

Klaus

 

In assessing Klaus' views on reflections…, posted on January 19, 2008 at 15:18:33
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 You need to be aware of several things:

1- the bit on Toole's views that is contained in the document he links to says in full:

"Toole has concluded that normal reflections in a typical small living room seem not to interfere with perception of the recorded space. He has also determined that early lateral reflections (<50ms) have a beneficial effect on intelligibility similar to raising the dialog level, and that the reflection pattern is more important than reverberation.

This has led to Toole’s recommendation that too many or too few reflections can be a problem. In particular, acoustic absorption, diffusion, and reflection must be broadband, ideally starting below 200Hz. He pointed out that the typical 1˝ or 2˝ sound panel most often affixed to walls works only at relatively high frequencies, and acts to effectively turn down the tweeter with no effect on the midrange or upper bass, thus unbalancing the sound.

Toole reminded us that many of the practices to deaden a room’s acoustics came from standards for broadcast and recording control rooms, where sound details must be heard more clearly, not for general listening rooms."

Klaus interprets the first paragraph of that section as saying that reflections aren't a problem, period. It doesn't go quite that far and the second paragraph points out that too many or too few reflections can be a problem. What is the situation in your room as opposed to Klaus ignoring that part of the discussion and saying they're never a problem? Do you have too few or too many reflections? Bear in mind that there isn't a fixed amount of reflection you should have but rather there's a range and different people can and do prefer different points on that range. Your room may fall within the range but at the other end of the range to your preference, or your room may fall outside the range. Klaus doesn't know or care about that. He just says there's no need to treat reflections and totally ignores the qualifications inherent in that quoted passage.

Note also that the quote is Linkwitz's take on what Toole said and not what Toole himself said. It's a summary and summaries always gloss over the fine detail. Klaus's statement that Toole says that reflections are not a problem is also a summary which drops out all of the fine detail and every qualification Toole makes. That kind of summary is absolutely the worst summary you can get. You can't assume that such a simplistic summary will apply to your situation. Linkwitz's summary is better and raises questions which should tell you that you shouldn't rely on Klaus's more extreme summary.

2- Next, treating first reflections does make a difference and it certainly improves imaging, soundstage, and the sense of the acoustic of the recorded venue in my room and in the one other room where I've seen first reflections treated. Other people who have done it report the same and at least one author, Everest, recommends treating of first reflections for that purpose. Note that Everest is a professional acoustician and author whilst Klaus is not. He's widely read but he hasn't conducted research and does not appear to work in the field of acoustics.

3- Klaus always conducts this debate in terms of "need". Do you NEED to treat first reflections? The answer is no because you can get quite enjoyable sound without such treatment and, in fact, without any treatment of your room. Can you get sound that you will enjoy more if you treat your room? Yes, most certainly and all you need to do to prove that for yourself is to try listening in a few treated rooms. You will probably like the results more in some rooms than in others depending on how well the treatment was done and what goal the person treating the room had for their results. If you're not interested in refining imaging and soundstage, don't treat first reflections with absorption but rather consider diffusion in those locations and at the speaker end of the room. There's more than one way to treat a room and you can certainly make your room deliver more of what you like in terms of sound quality than you get in an untreated room.

If you're thinking about room treatments then there must be some aspect of the sound you're currently getting that doesn't satisfy you and that you want to improve so the question is can room treatment do that for you. It can't fix every problem and it may take more treatment than you can afford or want to have in your room to fix some problems, but it can make a positive difference with many problems and there's more than enough people who've experienced the benefits to prove that claim. Klaus views their 'evidence' as anecdotal and chooses to ignore it but there's a lot of that evidence, it's consistent in its nature, and it agrees with the treatment recommendations of a number of professionals. You certainly don't NEED acoustic treatment in most cases but it certainly makes an improvement in most cases and, when it doesn't make an improvement, the reason is often because people ignored the recommendation of Toole and other knowledgeable people that absorption should always be broad band.

Its your room and your decision so you're free to choose to do what you want. Certainly consider Klaus' view but also consider those of others and decide on balance which side you agree with. If you're going to read references, certainly read those Klaus refers to but chase a few more as well and do include Everest's book, the Master Handbook of Acoustics, which just happened to be written to provide sensible advice on room treatment for people wanting to treat listening rooms and build their own recording studios.




David Aiken

 

RE: You don't need to treat reflections!!!, posted on January 20, 2008 at 11:39:23
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 Klaus,

You'd be a lot more convincing if you offered your own research rather than always quoting someone else. By the way, that logical fallacy is called Argument from authority. This is really just a special case of a Non-Sequitur since it does not necessarily follow that those who speak from authority are always correct.

I'll add that Floyd Toole is not the only expert in audio, acoustics, and perception. A lot of true experts, including me, disagree strongly that early reflections in a small room are not a problem. In fact, Floyd is in a very small minority of experts with that view. Not that being in the minority makes him wrong. But since you said he is "arguably the world's leading expert" I'll be glad to read you argue that point for us. Over to you...

--Ethan

 

From a report on the recent AES Convention, posted on January 20, 2008 at 14:17:14
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 From Scott Dorsey's report on the AES convention—see link for his full report:

"Sean Olive from Harman International and William Martens from McGill University did some experiments in Interaction between Loudspeakers and Room Acoustics Influences Loudspeaker Preferences in Multichannel Audio Reproduction. They used a multi-driver speaker which could be easily configured into several different kinds of speaker on the fly by enabling and disabling various redundant drivers. Some of the things they found are things you'd expect: for example, the reflectance of the room affects listeners' preference in speaker, and so does the program material. But they also found that adoption to a different room acoustic was very rapid, and that the experience of the listener dramatically affects configuration preference. Interesting if you are at all curious about speaker testing, even if much of it may seem like common sense on the face of it; common sense often turns out to be wrong, so it's good to see measurements like these made. Preprint 7196."

So if speaker preference is influenced by room reflectance and you can't change your speaker to one that works better with your room, perhaps you'd like to change the reflectance of the room to suit the speaker you have.

I accept the point about adaption to a different room acoustic being rapid but that won't affect preference for a particular sort of acoustic. While you will adapt rapidly to the room you're in, I'm prepared to bet that if you have 2 rooms of the same dimensions with different acoustics, whether that be because of surface materials or acoustic treatment, and identical speakers you will have a preference for one room over the other even though you may adapt rapidly to the room you're in.

And since Olive is a long time associate of Toole, this paper should be of interest to you, especially since it seems to be multi-channel.

It seems reflections do make a difference.




David Aiken

 

I don't have the anechoic chamber needed for such research!, posted on January 21, 2008 at 04:58:33
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 If the true experts disagree, then why is there no comment in JAES on Toole's paper? Why is there no peer-reviewed paper by one of these experts in one of the relevant journals showing that and why early reflections ARE a problem? Why don't you, as an expert, go public in these journals and prove your case?

Re: argument from authority
Authority in this case is the combined mass of data from the various technical/scientific papers, not Toole's opinion alone. I've provided the list of papers, read them and decide which conclusions can be drawn.

Re: leading expert
Toole has authored or co-authored 27 AES papers (both peer-reviewed and convention), 17 JASA papers (both peer-reviewed and meetings), most of them about the loudspeaker/room interface. Other names that come to mind are Bech and Damaske. What other names would you propose for this particular discipline?

Further, a true expert is aware of what's going on in his discipline. You are calling yourself a true expert, I hence suppose that you are subscribing to JAES, JASA, J. of Applied Acoustics, J. of Building Acoustics, Acustica/acta acustica, just to name a few.


Klaus

 

123rd AES convention, posted on January 21, 2008 at 05:04:18
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 I would not read to much into Dorsey's summary. I know of this paper as well of Linkwitz' paper about reflections. These papers are not yet available to me, so I have to wait until they are which is around May this year.


Klaus

 

It is Toole's views, not mine!!!, posted on January 21, 2008 at 06:40:57
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 It’s Toole’s views I’m talking about here all the time, not mine!!!!!! It’s Toole who says that early reflections are not problems, not me? I’m NOT interpreting anything at all, I’m repeating what an expert states. And I'm not talking about Linkwitz summary, I'm talking about Toole's AES paper. Toole's paper actually leaves no room for interpretation.

The text in the link is a summary provided by Linkwitz, the interested reader should read Toole’s AES paper, of which I provided the bibliographic data. In this paper Toole says : “The inevitable conclusion is that, in natural listening, room reflections are not problems.” This doesn't leave any room for interpretation, or does it?


Toole further concludes that “Individual reflections, in general, appear to be flattering to both music and speech sounds, and those occurring naturally in small rooms are, if anything, too low in level to have an optimal effect.”

I must, again, note that you adamantly refuse to read that paper!!! I wonder what the reason might be.



>Note also that the quote is Linkwitz's take on what Toole said and not what Toole himself said. It's a summary and summaries always gloss over the fine detail.<

Correct, and that’s why I say: “Read the paper”. That’s why I give bibliographic data. In Linkwitz’ summary, in the phrase about too many or too few reflections, too many stands for reverberation. This is a point Toole makes in his paper: excessive reverberation is a problem and measures should be taken.


>Klaus's statement that Toole says that reflections are not a problem is also a summary which drops out all of the fine detail and every qualification Toole makes. That kind of summary is absolutely the worst summary you can get.<

It’s not my summary, it is Toole’s. It’s in his paper, if you cared to read, you would find that phrase. Since it’s Toole’s summary you should argue with Toole, not me. Ot maybe with the scientists Toole is basing his conclusion on.


>Next, treating first reflections does make a difference and it certainly improves imaging, soundstage, and the sense of the acoustic of the recorded venue in my room and in the one other room where I've seen first reflections treated.<

Funnily enough, you say that Olive’s and Toole’s results of the single loudspeaker/single reflection situation apply directly to the stereo situation, but when they say, in the same paper, that “normal reflections in typical living rooms do not interfere with perception of the recorded space”, then that does not apply. When you treat reflections you will have possible effects in both timbre and spatial impression. Where is the careful investigation that shows that “it certainly IMPROVES imaging etc.?



>Other people who have done it report the same and at least one author, Everest, recommends treating of first reflections for that purpose. Note that Everest is a professional acoustician and author whilst Klaus is not. He's widely read but he hasn't conducted research and does not appear to work in the field of acoustics.<

I haven’t conducted research, nor has Everest, but Toole has. You seem to be consistently ignoring Toole's expertise and conclusion. Toole certainly struck a nerve here and many people, like you, will prefer to simply ignore what he says. And Everest made an serious mistake w.r.t Olive’s results as I have shown at least on two occasions.


>If you're thinking about room treatments then there must be some aspect of the sound you're currently getting that doesn't satisfy you and that you want to improve so the question is can room treatment do that for you.<


Be honest, most people think about room treatment simply because this is the message you usually get, not because there is a problem. I think that in view of the evidence discussed by a leading expert in the field this message is fundamentally wrong.


Klaus

 

RE: I don't have the anechoic chamber needed for such research!, posted on January 21, 2008 at 10:34:01
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 Klaus,

> I don't have the anechoic chamber needed for such research! <

You don't need that! All you need is a rectangle room and some absorption panels to alternately apply and remove. You really owe it to yourself to at least hear how the sound changes with added absorption!

> Toole has authored or co-authored 27 AES papers <

Do you have any idea how many papers Deepak Chopra and Andrew Weill have published? :->)

> Why is there no peer-reviewed paper by one of these experts in one of the relevant journals showing that and why early reflections ARE a problem? <

A Google search on 'audio first reflections' gives 229,000 hits. Do you really need more than that?

Tell you what Klaus. I've explained as best I can and you've refused to accept it even though you have never actually tried this yourself. So you go right ahead and continue to state early reflections are not a problem. I might pipe up now and again to disagree with you for the benefit of a newbie, but hopefully that won't stop you!

--Ethan

 

Here's my challenge to you., posted on January 21, 2008 at 14:16:08
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 I'm going to pull the discussion about reflections and whether or not they're a problem back to the specific case of the original query in this thread.

You comment that the "too many reflections" mentioned in Linkwitz's summary refers to reverberation but you jumped into this thread concerning a cube shaped room, of all shapes, and said nothing more than "Floyd Toole, arguably the world's leading expert when it comes to the loudspeaker/room interface, says that reflections are not problems." Standing wave behaviour is a reflection phenomena and standing waves in a cube are particularly problematic because there is only one axial frequency and it's strongly reinforced because all 3 axes have the same frequency. Room response also smooths out as frequency increases because there are more standing waves per octave. With different axial lengths supplying different standing wave frequencies, that response will smooth out more quickly in a room with 3 different axial dimensions than it will in a cube so not only will room response in a cube be more problematic because of the extreme strength of the standing waves it generates, that behaviour will also be problematic over a wider range than it would in most normal, rectangular rooms.

Tell me, do you think Toole would say that there are no reflection problems in a cube given standing wave behaviour in a cube? I somehow doubt it.

Of course you may have meant to say that first reflections aren't a problem rather than standing waves but you didn't say that. You didn't qualify your statement in any way. You simply said to a person with a cube for a room that Toole says reflections are not problems, thereby implying that there are no problems at all.

So rather than fight any more about whether reflections are ever a problem when we both agree that they sometimes are (you are committed to that in the case of reverberation at the very least) let's change the nature of the discussion here very simply and give you the chance to say something of practical value.

What advice would you give to a person with a cube shaped room who wants to use that space as a listening room? You can cover speaker placement, listening location, and room treatment. Where would you put the speakers, where would you put the listening chair, and would you do anything about acoustic room treatment in a cube with 12' sides and height?

In this particular case the original poster also said "I really like detail, imaging, and really enjoy hearing the soundstage extend outside the speakers. I guess I like everything! (incidentally I hear the soundstage often extending on the right side where the boundary is bare drywall, almost never on the left where the boundary is a large blinds-covered window)" so please take those tastes and the left/right asymmetry into consideration in your response.

I'm not going to bother responding to any more theoretical points here, but I'm prepared to discuss practical suggestions for the original poster if you have any. You've done the reading so now here's your chance to apply it and offer practical advice. Ethan and I have had our say. Since you disagree with our views what would you do in this specific room? Don't say what is or isn't a problem unless you're prepared to say why it isn't a problem in this particular room and give reasons in terms of acoustic behaviour, not "Toole says it isn't a problem" but you can, if you like, say WHY Toole or anyone else thinks that a particular aspect is or is not a problem in a cube.

This forum was intended to be a practical one and theory is really only useful if it can be applied practically so here's your chance to apply theory as you understand it and to provide helpful advice. If you can't do that, then all of your reading and theory is worthless in this forum. Criticising others is of value here only if you can offer alternatives with practical value and so far I haven't ever seen you offer any practical advice to anyone. You only ever seem to say why you think the advice others offer is wrong and you never seem to offer any practical advice of your own. Here's your opportunity to do so. There tends to be general agreement that a cube is the worst shape for a listening room and we have a cube here. What would you do with it?



David Aiken

 

RE: I don't have the anechoic chamber needed for such research!, posted on January 21, 2008 at 14:49:16
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 As Ethan says, you don't need an anechoic chamber to conduct research on whether or not reflections are a problem in a normal listening room. You actually want a room with reasonably normal reflection patterns and the ability to treat the room in order to reduce or eliminate reflections. Your own room would probably be quite suitable.

If you say that single speaker tests can't be generalised to multi-speaker situations, then surely anechoic chamber results with pseudo reflections generated by a second speaker with adjustable delay and level can't be generalised to a normal room with multiple reflections.



David Aiken

 

RE: Here's my challenge to you., posted on January 21, 2008 at 15:20:21
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 "Criticising others is of value here only if you can offer alternatives with practical value and so far I haven't ever seen you offer any practical advice to anyone. You only ever seem to say why you think the advice others offer is wrong and you never seem to offer any practical advice of your own."

Hear hear. :->)

 

RE: Here's my challenge to you., posted on January 23, 2008 at 04:21:27
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 I jumped into the thread after you had recommended, amongst other things, to treat first reflection points and the questioner had asked a specific question regarding this issue.

Re: Standing waves
You have standing waves at ALL frequencies by superposition of a wave with its reflection from a room boundary which creates an interference pattern:

Waterhouse, “Interference patterns in reverberant sound fields”, J. of the Acoustical Society of America 1955, vol.27, no.2, p.247

When you speak of reflection treatment in general terms you don't think of standing waves, do you?

Be that as it may, standing waves come in two forms, the common standing wave and, when particular requirements are fulfilled, the resonant standing wave or mode. When the sound source is stopped, only the modes will continue to ring, their decay time is longer than for the non-modes.

Standing wave is not equivalent to mode so you are using the term incorrectly.

When I play pure sine tones I detect the interference pattern very easily when walking around the room. When I play synthesizer music where sustained notes are common such pattern is somewhere between very weak and non-existing. When I play pop, rock, jazz or classical, I don’t detect any pattern at all. So I wonder how big a problem interference patterns are in real listening situations.

> that [modal] behaviour will also be problematic over a wider range than it would in most normal, rectangular rooms.<

The frequency range where modes are widely spaced as to possibly cause problems is determined by the room’s size, not by its shape. In the room in question (12x12x12) the transition to closely spaced modes will, according to Schroeder’s formula fc = 2000 sqrt RT60/V occur at about 240 Hz (assuming 0.7 s for RT60).

> Tell me, do you think Toole would say that there are no reflection problems in a cube given standing wave behaviour in a cube? I somehow doubt it.<

Toole says : “The inevitable conclusion is that, in natural listening, room reflections are not problems.” Of course, he did not mean those reflections which generate room modes. Of course, he did not mean late reflections in the case when there are too many.

> What advice would you give to a person with a cube shaped room who wants to use that space as a listening room? There tends to be general agreement that a cube is the worst shape for a listening room and we have a cube here. What would you do with it?<

There is also general agreement about optimum room dimensions. That these people are fundamentally wrong has been shown.

Advice: First, I won’t gove the advice to treat first reflection points becausev there is no evidence that first reflections are a problem.

Second, I would give the advice to first try to find out whether or not the cube shaped room creates mode problems at all. Everyone seems to be sure that a cube shaped room is a problem, whatever the circumstances are. In order to have a problem, in any room, the modes have first to be excited and then they have to be perceived at the listening position. So my advice would be to place speakers and listening chair where it suits best and listen to normal music. If you play music instead of test signals only those modes that correspond to frequencies in the music will be driven, and only those modes will ring on when the signal stops. However, music is a continuous and the passage following the one which is able to excite a room mode may mask the ringing. So maybe the speakers are in a position where they don’t excite the modes and/or the listening chair is in a position where the excited modes are not perceived in full glory.

Further, when you measure SPL at listening position you always must keep in mind that a mike operates linearly, unlike human hearing which operates non-linearly. If you look at curves of equal loudness you see that for the same perceived loudness the absolute SPL may differ by up to 20 dB in the sub 100 Hz range. Which means that you have to apply A-weighting or similar to the absolute SPL figures.

Should there be a mode problem at listening position when playing music, the obvious solution is to change positions of loudspeakers and listening chair. In my room I have detected so far one mode, when I move my head 20 cm forward, I’m out of the high pressure zone.

Should nothing of the above work, one could apply treatments as proposed already. Another solution could be to use narrow-band equalisation or four subwoofers placed on the walls midway between the corners:

Welti, “How many subwoofers are enough”, Audio Eng. Society preprint 5602 (2002)
Welti, “In-room low frequency optimization”, Audio Eng. Society preprint 5942 (2003)
Welti, “Low-frequency optimization using multiple subwoofers”, J. of Audio Eng. Society. 2006, p.347


>In this particular case the original poster also said "I really like detail, imaging, and really enjoy hearing the soundstage extend outside the speakers.<

Since I could not find any research at all on this issue I would not know what actions have which effect.

> (incidentally I hear the soundstage often extending on the right side where the boundary is bare drywall, almost never on the left where the boundary is a large blinds-covered window)<

This looks like a matter of higher SPL from the right side which generates an interaural level difference such that the apparent source is shifted to that side. That’s perfectly in line with psychoacoustic knowledge. The measure to take would be as a simple as using the balance button on the amp! If there’s no such button, a change in toe-in might work or an increase in distance of right speaker to listening chair. Placing the chair off-center might work as well. You obviously could envisage to create a situation where right and left reflection points (read: wall surfaces) are of similar (or identical) acoustic nature.



Klaus

 

RE: I don't have the anechoic chamber needed for such research!, posted on January 23, 2008 at 04:46:53
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 Soren Bech used a system with at least 18 loudspeakers, in the 60ies/ 70ies the Göttingen acousticians used a system with 65 loudspeakers for simulating a complete sound field:




Furhermore, when you read books like Bech's "Perceptual audio evaluation - theory, method and application" you will see that performing meaningful listening tests is anything but simple.


Klaus

 

RE: I don't have the anechoic chamber needed for such research!, posted on January 23, 2008 at 05:22:51
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 Though I'm not a professional researcher, I've had to do research in a non-audio health related area for a post-grad qualification and I did manage to do it well enough to present it at a professional conference and have it published in a peer-reviewed journal.

You're right, performing meaningful tests—whether they be listening tests or other tests—isn't easy but that doesn't mean that you need complicated setups for all tests. You don't need an anechoic chamber to study the effects of the treatment of reflections in a listening room. In fact using an anechoic chamber would defeat the purpose of your tests if you're concerned with the effects of treating normal home listening rooms. What you need is:

- a hypothesis that states how listeners are expected to perceive the change in sound quality resulting from the acoustic treatment of a room. This is the first of the 2 really critical factors in the study;

- a reasonably representative average home listening room;

- a sound system, either stereo or surround but it would certainly be preferable to do different tests for each;

- a set speaker and listening position setup that doesn't vary through the tests;

- some way of treating reflections; professional treatments with known acoustic properties from independent tests would be best, plus a standard set of treatment placements for individual evaluation;

- subjects;

- and finally a good protocol for getting subject responses, one that doesn't have hidden biases in the questions that would tend to colour the responses This is the second of the 2 critical factors.

The first and last points are the hard ones because they're the ones that will really make or break the validity of your findings. The physical aspect of the test set-up is simple and remains fixed throughout the tests. It's also described in the report of the study so that others can duplicate the setup in order to replicate the study.

Your hypothesis and protocol for gathering subject responses really are the critical factors, not the physical set-up. Get the hypothesis wrong and you end up testing for the wrong things. Ask the wrong questions or ask the questions in the wrong way and you run the risk of influencing the subject's responses in ways that reduce the reliability of your data and/or make it unclear just what interpretation should be placed on the data. These are the parts that everyone doing research that relies on test subject responses has to really sweat on. The rest of it, the phsyical set-up in the sort of case we're talking about, can be a cake-walk in comparison.

And that leaves only the final bit: analysis and interpretation of the data. Provided the hypothesis is quite clear and the data protocol is appropriate, the interpretation of the data should be relatively straightforward.

As I said, you could do it in your own listening room PROVIDED your test design has a clear and testable hypothesis and a good protocol for gathering data on subject responses. You could do it in your room with your system and your listening set-up and do it well enough to have it published in the peer-reviewed research journals provided you get the protocol and data gathering right. You don't need a fancy lab set-up to test the responses of subjects to treatment of reflections in the average real world listening room.



David Aiken

 

RE: Here's my challenge to you., posted on January 23, 2008 at 06:45:04
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 Klaus,

I'm not going to respond to anything prior to your comments before you quoted my question about what advice you would give the original poster. I'm not going to argue theory with you any further and I said that.

I have agreed with you in the past about the non-optimum nature of the so-called 'optimum room dimensions', but the fact that the generally recommended ratios aren't all that significant does not mean that some room shapes are not worse than others. You've offered no proof that cube shaped rooms aren't more problematic than others, you've simply begged that question.

You then say that there's no evidence that first reflections are a problem and, several paragraphs later talk about the L/R channel balance issue as being "a matter of higher SPL from the right side which generates an interaural level difference such that the apparent source is shifted to that side.". Of course it is just that, but the main contributor to that difference will be the first reflection which is the strongest reflection from the area and which arrives first so there's your evidence that, in this case, we have a first reflection problem.

Now to your recommendations for dealing with it. Yes, a balance control might do the trick but a lot of equipment no longer has one. While that will fix the location of the centre image it won't control the sideways shift of the apparent source from the speaker to the reflection point on the wall. Toe-in is unlikely to fix any shifts of the centre image or the sideways shift of the apparent source nor will moving the speaker though all of these suggestions will certainly change the size of the apparent shift.

You carefully ignore the room dimensions when you talk about speaker movement. In a 12' x 12' floor area there isn't much in the way of scope available to move the speakers without putting them too close together to generate a wide soundstage and one of the original poster's requests was for a solution with a wide soundstage with good spread to the outside of the speakers. Moving a speaker away from a wall moves it closer to the other speaker, narrowing the soundstage. That's counterproductive in this case.

I would suggest that treatment of the first reflection area on both side walls would accomplish the poster's desired results better than your suggestions in a room of this size. In a room with a much larger floor area where the speakers could be placed a reasonable distance apart and the distance to the offending side wall could be increased sufficiently to weaken that first reflection significantly in relation to the strength of the direct sound, you may well be able to get by with moving the speaker as you suggest. Your other suggestions, apart from the balance control which depends on the equipment rather than the room, would also be likely to have more effect in a larger room but in a small room such as this one I'd suggest that they are unlikely to have any significant effect.

I'd also suggest that treating the first reflection points will assist with detail, imaging, and breadth of soundstage, the areas on which your response which was: "Since I could not find any research at all on this issue I would not know what actions have which effect." I would strongly suggest to you that all of those anecdotal reports you choose to reject, plus some of the material in Everest who you criticise for not having done any research himself, strongly support the view that treating the first reflections would assist in meeting the original poster's needs.

You're right also about room modes needing to be excited and that you have to be in a location where the pressure is highest for them to be heard at their worst. As already stated I'd suggest that your options are limited in a room of this size. The fundamental mode associated with a 12' dimension is roughly 47.2 Hz which falls roughly mid-way between F#1 and G1. we're lucky that it's "in the cracks" as it were so provided the music listened to is tuned to A=440Hz there may not be a problem. A lot of baroque period music is played around a quartertone to a semitone flat and may excite that mode a lot more.

You demand measurements in proof of a problem but there's more than sufficient practical experience around in my view to support the claim that room modes do cause problems in most small rooms and that dealing with them in some way is beneficial.

There isn't much room to vary speaker or listening position placement to minimise problems in this room yet that's your only suggestion. Bass traps are an effective solution and in this case I actually said they were my primary suggestion for treatment. Bass traps will certainly reduce the reverberation time associated with any modes and reverberation time is likely to be longer in a room where each dimension has the same modal frequency because the three modes will reinforce each other. Even if the worst aspects of the mode can be avoided by your suggestions, the increased reverberation at modal frequencies may well be problematic.

You suggest that the room response should have smoothed out by 240 Hz. Since the fundamental mode for the room is roughly 48 Hz and it's usually taken that response will be smoothing out by the time there are 5 modes per octave and there will be 5 modes in the octave from 240 Hz to 480 Hz, you're probably right on that. What I said was that it would smooth out more quickly in a room with 3 different axial modes rather than a single mode replicated in all 3 axes. I stand by that.

Finally we come to the section on detail, imaging, and breadth of soundstage and your response which was: "Since I could not find any research at all on this issue I would not know what actions have which effect." I would strongly suggest to you that all of those anecdotal reports you choose to reject, plus some of the material in Everest who you criticise for not having done any research himself, would strongly suggest that treating the first reflections would assist in meeting the original poster's needs.

All in all, not bad for a first effort at practical application of your knowledge but you really do need to consider the room size and just what that means for what a person can and can't do in relation to set-up issues and the effect that changes in set-up can reasonably be expected to achieve. All you can do is small rooms is make small changes and small changes aren't going to have big effects on a pronounced L/R imbalance.

Now how about you continue the good work by trying to help the people asking questions instead of simply sniping at the people making suggestions, especially when the suggestions you're getting upset about, ie treatment of lateral reflections, are often intended to address those areas where you said you were unaware of any research. You may not have seen any research but there is a large body of practical experience on dealing with those issues and since you have no research on the topic, you've also got nothing to prove that practical experience to be wrong.



David Aiken

 

The advantage of a synthetic sound field..., posted on January 23, 2008 at 07:15:25
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 is that you can change the field without delay. Auditory memory is notoriously bad and listening to music without treatment, then installing the treatment, then listening again certainly requires more time than you are able to keep your impressions reliably in your memory. In addition, a synthetic field would allow to perform the test blind.

Further I don't know if you need a hypothesis to be tested: just test for changes in timbre and spatial aspects, that's how Bech ran his tests in the 90ies, using a synthetic sound field.

Further,when you treat reflections you are lowering the overall loudness. This could lead to confusion in what is actually perceived in terms of timbre or spatial aspects.



Klaus

 

RE: The advantage of a synthetic sound field..., posted on January 23, 2008 at 13:08:27
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 A synthetic sound field is always open to doubts about how accurately it mimics a real life situation and it isn't necessary if you can use a real life situation.

Your point about audio memory is valid, but it assumes that you're going to gather data based on subjects doing comparisons. That's not the only way to do the test. You can ask after exposure in one situation how the subject felt about certain qualities of the sound they heard, and get ratings on a 5 or 10 point scale. If there is a significant difference in the ratings for the treated room vs the untreated room, that points to a change in how the sound is perceived.

You always need a hypothesis when designing a study. It's a basic part of study design. Obviously you haven't had to do any research at all if you suggest that you don't need one in this sort of study.

Yes, treating reflections will change overall sound levels but relatively insignificantly. If the reflected sound field is equal in strength to the direct sound field at the listening level, the total level will be 3 dB higher than both. Treating lateral first reflection points, for example, can be done with two 2' x 4' panels, a relatively small area in relation to the room's total reflective surface. Even if you managed to get total absorption at those panels, which you won't, you still aren't going to lower the total level of the overall sound field by a great amount though with some affect on bass and smoothing of low frequency room response if the panels are genuinely broad band, there may be significant level changes at some frequencies in the bass region. Overall loudness isn't going to change significantly but even if it does that won't be an issue if the study does not use direct comparison. In fact, if you're not doing direct comparisons you could do listening tests in the untreated room on one day and in the treated room on another. Small variations in overall loudness would then not be a problem. You also have the possibility then of even allowing subjects to adjust overall volume to their own preference since they're only rating sound quality in one room situation at a time.

There are objections that can be raised to using a normal room and to using a synthetic sound field. There's no perfect choice but there are usually ways of addressing the objections in either case. If the study is intended to research how listeners react to real world treated or untreated rooms, use of a real world room is preferable if it can be managed and it can be managed.



David Aiken

 

Additional comments…, posted on January 23, 2008 at 15:07:04
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 In relation to bass problems you said "Another solution could be to use narrow-band equalisation or four subwoofers placed on the walls midway between the corners" and I failed to comment on those suggestions.

Narrow band equalisation will work for peaks but not for nulls. You can't correct for a null by equalisation without running the risk of overloading both amplifier and speaker, and all high quality digital room EQ products work on the basis of correcting the peaks but leaving the nulls which often tend not to be noticed. Physical treatments such as bass traps actually smooth the response by absorbing energy that contributes to both the peaks and nulls. You need test equipment to get the best out of narrow band EQ but you don't for physical treatments though you do need test equipment for both if you want to verify the scale of the change they make rather than simply relying on your ears and subjective evaluation. Subjective evaluation isn't necessarily bad in this case because the changes from both approaches will be clearly audible if both are implemented well.

The multiple subwoofer option will certainly work and I suspect it will deliver better results than EQ because it is likely to smooth out some of the nulls as well as the peaks, but I wonder how preferable it is on a practical level to physical treatments. Subwoofers cost at the purchase stage and also to run whereas physical acoustic treatments have a purchase cost and no running cost. A decent bass trap costs about half the price of a decent sub here in Australia and even if you're using 2 traps per corner and treating 4 corners, the bass trap option is going to be comparable in price to 4 subs in many cases. It will be dearer if you're buying cheap subs and cheaper if you're buying top flight subs but price can reasonably be considered comparable at the purchase stage. From there on in bass traps are definitely cheaper to run. There is the additional problem with subs of actually getting decent integration with the speakers. It's not necessarily easy. I thought I had done a reasonable job integrating my REL Storm III into my audio system which uses an integrated amp and the REL input was taken from the speaker outputs of the amp. After considerable time and effort over a period of months, I felt reasonably happy that I had achieved a good balance both tonally and with respect to level. Out of curiosity I moved it to a separate 2 channel HT system that used a surround sound AV receiver with sub outputs and automated set-up and equalisation. I ran the auto set-up and sat back staggered at the results which surpassed anything I had achieved in months in the audio system. The big difference was the fact that the receiver matched signal arrival times for all speakers and the effect of that on transient response and overall clarity was simply surprising. I could not achieve that result in the audio system though I could now go closer towards achieving it with some of the information I learned by moving the sub to the HT system where it still remains. The audio system also sounds better in respect of transient response and clarity in the bass region without the sub though I've definitely lost low bass extension. On balance I've decided I can do without the extension in my system and that the gain in transient response and clarity is more beneficial since my mains produce useful response down to close to 30 Hz anyway.

Set-up problems will be magnified with multiple subs. Receivers are now starting to appear that will deal with 2 subs and handle their integration but I've yet to hear of a receiver that will do that for 4 subs. Without test instruments and a lot of work I doubt that anyone will achieve really good results integrating 4 subs into most systems because of the differences in listening distance to them amongst other factors. In general my recommendation on multiple subs would be that they may be a worthwhile option in a surround sound system where you're going to have at least 1 sub anyway but they aren't anywhere near as likely to be a worthwhile option in an audio system, especially if the speakers there already have sufficiently good extension for the listener's preferred musical choices. Even in a surround sound system I think that multiple subs only start to become a really worthwhile option for most people when the receiver or AV controller is capable of controlling them individually and that, at the moment, would seem to limit choices to a 2 sub set-up though that will almost certainly change over time.

So, on balance and for entirely practical reasons I tend to prefer physical acoustic treatments over equalisation or multiple subs for resolving low frequency room problems. I have no doubt that quite effective results can be achieved with all 3 of those approaches and one may certainly be better than the others in theory but I don't know which. In practice I think physical treatments are the simplest to implement, the cheapest to run, and comparable in purchase price to the other options at similar quality levels.

You may certainly have a different view to me on those comparisons and my preference for physical treatments but I do stress that the main basis of my preference is practical with the proviso that the physical room treatments are appropriate to the scale of the problems. My main reason for stressing appropriateness here is simply the fact that there seem to be an increasing number of room treatment products coming out that seem to be designed for aesthetics first and with questionable efficacy. I'm talking about standard absorption based commercial products with proven effectiveness and those DIY treatments with a sound theoretical basis that are known to be effective at low frequencies.



David Aiken

 

RE: Here's my challenge to you., posted on January 24, 2008 at 04:50:25
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 >You've offered no proof that cube shaped rooms aren't more problematic than others…<

Well, since a cube is, at least in theory, the worst possible case the approach is very much the same as for room with optimum dimensions: the only manner to have the worst possible conditions is to place the loudspeakers and yourself in a corner. Here you excite and hear all modes.

There is one question I’m asking myself: does it matter where on the room mode pressure curve you place the loudspeaker? If you place it in a maximum, you fully excite the mode, if you place it in a minimum, you don’t excite it at all (at least in theory). What happens if you place the loudspeaker somewhere between maximum and minimum? Do get full excitement, partial excitement, or what? I did not yet start investigating the relevant literature, so I don’t know the answer. Everest says that the modes are driven proportionally, but what exactly does that mean?

>Of course it is just that, but the main contributor to that difference will be the first reflection which is the strongest reflection from the area and which arrives first so there's your evidence that, in this case, we have a first reflection problem.<

Not necessarily, the quality of the loudspeakers may be such that pair matching is bad, in which case the louder speaker would dominate apparent source location and image width.

Then there is the fact that not everyone hears the same: in-ear measurements have shown that there are substantial differences between individuals because of different shapes of head, torso, outer ear, different hair, etc. What the questioner is perceiving may be different from what you or I would perceive. I’m just reading Newell, “Recording studio design” and the author mentions personal anecdotal evidence. See also

Shaw, “Ear canal pressure generated by a free-field sound”, J. of Acoust. Soc. America 1966, p.465

>Yes, a balance control might do the trick but a lot of equipment no longer has one. While that will fix the location of the centre image it won't control the sideways shift of the apparent source from the speaker to the reflection point on the wall. Toe-in is unlikely to fix any shifts of the centre image or the sideways shift of the apparent source nor will moving the speaker though all of these suggestions will certainly change the size of the apparent shift.<

Balance control will lower the SPL of both the speaker and the reflection. The apparent source, the phantom source, is located between the speakers, ideally it is centred, but that depends on program material. With speech mono recordings the phantom source is dead centre, if not, you have a problem.

Toe-in will change the reflection’s SPL, depending on the speaker’s off-axis behaviour. Changing the distance speaker-listener will change SPL directly, both of direct sound and reflection. All these measures might be sufficient and they are for free. One measure I forgot is to shift both speakers to the left side, this increases the SPL from the left while decreasing SPL from the right.

When you consider the setup with the dimensions and distances as indicated by the questioner, especially the 2.5' from side walls, this means that the reflection is delayed by about 2.8 ms. The lower limit of the precedence effect for artificial signals is 0,8 ms, for music I could not find any data. It may well be that the figure of 2-3 ms is critical in this respect in which case a location and soundstage issue would be obvious. Anyway, Toole actually says that the reflections falling within this range are least beneficial, evidence provided by Ando.

>You carefully ignore the room dimensions when you talk about speaker movement. In a 12' x 12' floor area there isn't much in the way of scope available to move the speakers without putting them too close together to generate a wide soundstage and one of the original poster's requests was for a solution with a wide soundstage with good spread to the outside of the speakers.<

Agreed, a room of this size limits the possibilities. However, without any treatment the questioner states that he “hears the soundstage often extending on the right side”. So apparently the bare wall on the right does something he likes. Yet what you are proposing would destroy this. What is more, here we have anecdotal evidence for the beneficial effect of early reflections. And checkmate!

Instead of treating the reflections in the usual manner, it would seem more appropriate to take the logical step and provide the same conditions on the left side, the soundstage logically will extend also beyond the left speaker.

>You demand measurements in proof of a problem but there's more than sufficient practical experience around in my view to support the claim that room modes do cause problems in most small rooms and that dealing with them in some way is beneficial<

I’m not demanding measurements, I’m saying that measurements will not prove anything unless you consider the mechanisms our hearing is using such as shown in the curves of equal loudness. Most people, I guess, get themselves a SPL meter, measure and find that there are huge differences, whilst blissfully ignoring or forgetting that our hearing will smooth out a great deal.

>There isn't much room to vary speaker or listening position placement to minimise problems in this room yet that's your only suggestion.<

The questioner hasn’ even reported a mode problem yet you suggest treatment right away. In my room I haven’t any mode treatment whatsoever except for the floor-to-ceiling modes, I did not pay any attention whatsoever w.r.t placement of speakers and sofa, yet so far I discovered only one mode that is actualy excited, the hazard has it that the responsible track is called “Pressure points”

Who says that under all circumstances there are mode problems, even in a cubus? General agreement again?

>Bass traps will certainly reduce the reverberation time associated with any modes and reverberation time is likely to be longer in a room where each dimension has the same modal frequency because the three modes will reinforce each other.<

You seem to assume that when one of the axial modes is driven by the speaker because of unlucky placement, all three modes are automatically driven. Let’s consider the case where the speakers are against the front wall, 2.5’ from the side walls. The first front-to-rear wall mode will be driven because the speaker is in the pressure maximum of that mode. With floorstanding speakers the woofer is low but not on the floor. Will it drive the first floor-to-ceiling mode? The speakers are 2.5’ from the side wall, hence not in a pressure maximum of the first right-to-left mode? Everest says that modes are driven proportionally. Which clarly means that there are loudspeaker positions where the three axial modes are not driven fully. Determine these points and then place the chair close to a null.

>it's usually taken that response will be smoothing out by the time there are 5 modes per octave<

The Schroeder frequency is a concept that does not work in small rooms due to the very different nature of the sound field. It has further been shown that the transition from the mode region to the non-mode region is at frequencies much higher than the calculated Schroder frequency:

Baskind et al., “Sound power radiated by sources in diffuse field”, Audio Eng. Soc. preprint 5146 (2000)

> What I said was that it would smooth out more quickly in a room with 3 different axial modes rather than a single mode replicated in all 3 axes. I stand by that.<

What you said was “that behaviour will also be problematic over a wider range”. Wider range is commonly understood as frequency range, not time span.

>Finally we come to the section on detail, imaging, and breadth of soundstage and your response which was: "Since I could not find any research at all on this issue I would not know what actions have which effect." I would strongly suggest to you that all of those anecdotal reports you choose to reject, plus some of the material in Everest who you criticise for not having done any research himself, would strongly suggest that treating the first reflections would assist in meeting the original poster's needs.<

Everest is misinterpreting Olive’s results and statements. There is no reliable research on how reflections treatment changes detail, imaging, image width etc. Your anecdotal evidence certainly classifies the changes as improvement. For good reasons Everest addresses this issue in vague, general hence imprecise terms. What Everest does say is that interaural cross-correlation is a factor which affects sense of spaciousness, the weaker it is the better the effect. This is fully confirmed by psychoacoustic research yet nobody takes this into account when making treatment suggestion. The (wrong) general agreement is that the setup has to be acoustically symmetric.


> the suggestions you're getting upset about, ie treatment of lateral reflections, are often intended to address those areas where you said you were unaware of any research.<

There is research re: early reflections and the conclusions are that these are not a problem and hence don’t need treatment. As long as I’m convinced that this conclusion is true I will give advice accordingly.

Summarizing it up: the questioner has not reported any mode problem, yet you advise treatment. He says that he likes the soundstage extending beyond the speakers, this is what happens on the right side because of the bare wall, yet you advise treatment.

My advice would be: since there is no apparent mode problem, leave things as they are. For the soundstage extending beyond the speakers, first check if that is not a system problem (bad pair matching: swap speakers; amplifier output: swap speaker cables), then try and move the speakers in order to decrease SPL on the right side/increase SPL on the left side, if that doesn’t do it, what works on the right side should logically also work on the left side so try to have a similar surface at the left reflection point.

Klaus

 

Subs vs acoustic treatment, posted on January 24, 2008 at 05:02:29
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 Of course, it's a matter of budget etc. but it's a way to address problems. Harman has a patent on bass management system so maybe they also have gear which uses the technology.

But first of all I would try to find out whether or not room modes are a problem at all. It's not because every textbook has a chapter on modes that it's a problem in real listening situations when listening to music. As I said in my other reply, apart from the acoustic ceiling I have no treatment whatsoever and in 6 years time I have discovered only one mode that actually gets excited and is perceived at listening position, move the head 20 cm and it's gone. In this particular case the questioner has not reported problems, so why treat?


Klaus

 

RE: Subs vs acoustic treatment, posted on January 24, 2008 at 09:51:48
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 Klaus,

> I would try to find out whether or not room modes are a problem at all.

If you have a domestic size room, modes are always a problem. Unless you consider peak/null spans of 20 to 30 dB or even more, and ringing of half a second or longer, not to be a "problem."

> apart from the acoustic ceiling I have no treatment whatsoever and in 6 years time I have discovered only one mode that actually gets excited and is perceived at listening position <

You've never actually measured your room using appropriate software like ETF, right? :-)

--Ethan

 

RE: Here's my challenge to you., posted on January 24, 2008 at 09:53:06
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 > As long as I’m convinced that this conclusion is true I will give advice accordingly. <

You go girl!

 

RE: Here's my challenge to you., posted on January 24, 2008 at 21:15:03
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 What I will say, however, is that when it comes to dealing with situations in an actual room, the research often has to be taken with a grain of salt. Much research is undertaken in rooms that aren't like real rooms, they're symmetrical with uniform reflection behviour around the walls and across the ceiling. Real rooms often aren't symmetrical and/or don't have symmetrical reflection patterns due to placement of windows on one side and not the other, asymmetric door locations, doorways that are simply permanently open entrances, and even placement of furniture if the room is a living room or is also used for some other purpose. A finding that reflections aren't normally a problem in a symmetrical room with the same reflection behaviour on the left and right sides really may not count for much in an actual listening room with windows or an open doorway on one side wall, and/or an asymmetric shape. Bass behaviour will not be as predicted in sealed room studies if there are open entrances connecting the room to other areas or if the room is asymmetric. Research findings are certainly important but they do have to be interpreted in the light of the room under consideration and that room is probably going to vary in quite a few ways from the room the research was conducted in. You simply can't expect to take research findings and be able to apply them without variation or modification to most real world situations because of the range of variation in rooms and furnishings. That, to my mind, is your failing in a lot of our discussions. You expect the rooms under discussion to always behave the way the rooms in the research studies did and I think that occurs less than half the time.


David Aiken

 

Re:If you have a domestic size room, modes are always a problem, posted on January 25, 2008 at 22:05:14
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 I've got a test CD with sine waves 0-100 Hz in one Hz steps. The first width modes are at 37 and 72 Hz, per equation. When playing the tracks I get strongest response at 36 and 73 Hz, the 73 Hz mode being a bit weaker than the other.

My speakers have two separate mute funcions, actuable via remote control, both can be set to any value between 0 and 128 dB attenuation. By default they are set to 10 and 20 dB, respectively. When I compare the strongest response at 36 Hz with the weakest response I find that the difference is more like 10 dB, not 20. When I switch the signal off, the mode is gone within one blink of the eye but then, my speakers don't ring

If you have problems with modal ringing, you should perhaps buy better speakers!

When playing music I don't perceive standing waves let alone modes, apart from the one I mentioned. There must be a difference between theory and practice, quite obviously.

Why should I make measurements: these use a single mike, I have two ears with a head inbetween, and I have a torso, hair, beard, eye-glasses. Microphones are not sensitive to the direction (azimuth, elevation) from which the signal comes, human hearing, on the contrary, is. The sound waves that hit the membrane of a mike are very different from those that hit the ear drums, because of torso, hair and beard, pinnae, ear canals. And the in-ear difference between indivduals are significant which explains the fact that you may hear things which I don't and vice versa.

Further, measurements use artificial signals, I rarely listen to pure sine waves, tone bursts, pink noise. Measurements may valuable diagnostic tools once you have detected an audible problem, but that's about it.

Standing wave and modes are steady-state phenomena, how long an impulse does it take to excite a mode? My gear allows to prepare impulses of 10 ms and multiples thereof. I'll see.


Klaus

 

RE: Here's my challenge to you., posted on January 25, 2008 at 22:39:59
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 >Real rooms often aren't symmetrical and/or don't have symmetrical reflection patterns <

Asymmetry in itself is not necessarily a bad thing, on the contrary

Schroeder (1979), “Binaural dissimilarity and optimum ceilings for concert halls: More lateral sound diffusion”, J. of the Acoustical Society of America, Vol. 65, No. 4, S.958
Kurozumi et al. (1983), „The relationship between the cross-correlation coefficient of two-channel acoustic signals and sound image quality”, J. of the Acoustical Society of America, Vol. 74, No. 6, S.1726


There may be cases of very strong asymmetry, I agree, which need appropriate action. This action, like in the particular case, could however consist in creating a reflective zone, not in the ubiquitous absorption. So your standard advice would appear to miss the target in this case.

> Bass behaviour will not be as predicted in sealed room studies if there are open entrances connecting the room to other areas or if the room is asymmetric.<

Bass behaviour will not be as predicted because the formula assume ideal conditions, i.e. an empty room with no openings and perfectly rigid walls. Real rooms are far from that ideal, so the frequencies are not as calculated and, what is maybe more important, the amplitude and decay times of the modes are not what one might expect from theory. There is research like

De Melo et al., “Sound absorption at low frequencies: room contents as obstacles”, J. of Building Acoustics 2007, vol. 14, no. 2, p.143
De Melo et al., "Finite element model of absorbent furniture in small rooms at low frequencies", 9th International Congress on Sound and Vibration, Orlando, USA, 8-11 July 2002

which any true expert should have read.

> You simply can't expect to take research findings and be able to apply them without variation or modification to most real world situations because of the range of variation in rooms and furnishings.<

And you can’t take the theoretical approach, which is found in every handbook on acoustics, and say that there MUST be problem in real rooms when playing music. To use you own words, that, to my mind, is your failing in these discussions. You expect the rooms under discussion to always behave like theory or the general agreement predicts.

As you could see in the case of opimum room dimensions, textbooks and the general agreement are wrong or misleading. I have the strong suspicion that they are also wrong or misleading in the case of cube shaped rooms. I have the strong suspicion that they are wrong or misleading in many other cases. Textbooks base their case on theory, idealized conditions and measurements. Real rooms are not ideal, so the theory has to take into account these different conditions. Measurements not necessarily relate to how humans hear, a nice example being the curves of equal loudness to which you have to compare SPL data of a linearly operating microphone.


Klaus

 

RE: Here's my challenge to you., posted on January 26, 2008 at 00:35:24
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 >Real rooms are not ideal, so the theory has to take into account these different conditions.<

Agreed, but are you sure that the research you quote so often took into account those different conditions, and do you take them into account every time you reject treatment of reflections with the simple unqualified statement that "There is no evidence that reflections cause problems"?



David Aiken

 

RE: Re:If you have a domestic size room, modes are always a problem, posted on January 27, 2008 at 18:14:45
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 Taking the quotes out of order:

"Why should I make measurements:…"

"The first width modes are at 37 and 72 Hz, per equation. When playing the tracks I get strongest response at 36 and 73 Hz, the 73 Hz mode being a bit weaker than the other.
My speakers have two separate mute funcions, actuable via remote control, both can be set to any value between 0 and 128 dB attenuation. By default they are set to 10 and 20 dB, respectively. When I compare the strongest response at 36 Hz with the weakest response I find that the difference is more like 10 dB, not 20."

I assume you're saying that the difference between the 2 modes is more like 10 dB. My response is that if you're not measuring it with a meter or software, you really don't know. Estimating differences in SPL by ear is problematic at the best of time and doing so at different bass frequencies is especially problematic because of the ear's non-linear response (see Fletcher Munson curves including the change in the response curves at different SPLs). The difference in the ear's sensitivity at 36 Hz and 72 Hz can easily be 10 dB or more depending on SPL. You simply can't reliably judge the difference in level of your 2 modes by ear, you need measurements in order to get a reliable result.

Of course you could be right but if you did it by ear that would be sheer luck rather than anything else.



Now to some real curve balls:

" The sound waves that hit the membrane of a mike are very different from those that hit the ear drums, because of torso, hair and beard, pinnae, ear canals."

I think those differences are basically irrelevant. Microphones/measuring equipment respond differently to the way the ear/brain system responds and the differences in how they respond are really so great as to make any differences between the actual sound waves impinging on the mic diaphragm and the ear drum irrelevant.


"When I switch the signal off, the mode is gone within one blink of the eye but then, my speakers don't ring"

Well, I think every driver exhibits some ringing but the reason the mode is gone is not because the speaker "doesn't ring" but because the decay time of the mode in the room is short enough for you not to notice the time it takes for the sound to decay to inaudibility after the speaker stops emitting a tone. If we're talking about modal behaviour in a room, we're not interested in how quickly the speaker stops making sound but how long the room's modal behaviour causes the tone to persist after the speaker stops making a sound.

Of course, the "blink of an eye" does take a finite period in which to occur and average living room reverberation times range from .69 sec at 125 Hz to .4 sec at 8 kHz. Reverberation times are based on time taken for the SPL to decay by 60 dB and, given the background noise levels, it's unlikely that sound in most of our rooms would have to decay by 60 dB in order for us to regard it as inaudible so I'm really not certain that the fact that modal excitation disappears in "the wink of an eye" really tells us anything we wouldn't expect to find in most cases. I just tried and I find it hard to blink twice in a second. As an observation, measuring decay times in terms of eye blinks seems particularly unhelpful since the degree of resolution of the measurement unit is simply too gross for the measurement to be useful or informative.


"If you have problems with modal ringing, you should perhaps buy better speakers!"

And just how, pray tell, is a change of speakers going to change how fast a mode decays in a room?

Confusing speaker behaviour and room behaviour in this discussion is simply bizarre.


"Standing wave and modes are steady-state phenomena, how long an impulse does it take to excite a mode? "

I think that's the wrong question. In my view the important question is how long does it take a mode to decay once it has been activated.





David Aiken

 

Research and real life, posted on January 28, 2008 at 06:38:00
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 You ask me whether or not research applies to real life situations. You don’t ask this question when “true experts” like Ethan use measurements to “prove” that early reflections cause audible comb filter coloration yet these measurements are totally irrelevant simply because they don’t take into account binaural hearing mechanisms, in particular binaural decoloration. You blindly accept this line of reasoning perhaps because it looks good on paper and it's in line with the general opinion.

That simple statement, which actually is Toole’s statement, is not unqualified at all. Read Toole’s paper, read my own literature overview, read the literature if you really care, then you are in a position to draw conclusions.

As for the anecdotal evidence you are so obviously favoring, in his 1998 JASA paper “Spatial aspects of reproduced sound in small rooms” Bech says the following:

“When the level of the [single] reflection was set to its initial level there was a definite perceived difference between the standard and the comparison stimulus. The difference consisted of a loudness difference, the appearance of a second sound source and a timbral difference. A decrease in the reflection level caused the loudness difference to disappear, and a further decrease caused the differences related to the second sound source to disappear. The difference in timbre was the last to disappear, which means that the threshold of interest in the present experiments was reached while the timbral differences still existed. This means that the chosen experiment should allow the subject to discriminate between two coexisting cues.”

Bech takes this situation into account by using very specific investigation methods. Those folks who report on "improvements" when applying reflection treatment are not using any experimental method at all, yet you are sure that they are capable of producing meaningful results.


Klaus

 

RE: Research and real life, posted on January 28, 2008 at 13:52:41
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 At least some of Toole's studies on reflections did not employ reflections at all. They employed 1 speaker as a 'source' and a second speaker, displaced to the side with control of time delay and signal level, to simulate a reflection so that they could easily vary the 2 parameters they considered important, delay in reflection arrival time and level of the reflection relative to that of the direct sound. From the description of the Bech study you quote, something similar appears to have been done in it.

I'm not saying that such studies don't produce useful information, and you know I've quoted the Toole study as a source for information supporting treatment of first reflections even if we disagree about whether Toole's study does provide support for my views, but they certainly don't tell the whole story of what goes on in a real life situation where there are actually multiple reflections being heard, not just the single simulated one. Bech acknowledges that in your quote.

You say "Those folks who report on "improvements" when applying reflection treatment are not using any experimental method at all, yet you are sure that they are capable of producing meaningful results."

I don't think you have any real idea of what experimental method really is so I'll ask: have you ever done any research at all in any field? I have. only once as part of the requirement for a post graduate qualification in occupational health, but it was good enough to get published in one of the peer reviewed journals and to be presented at a scientific conference so at least I've got some idea. It may not have been acoustics but the basic approach to experimentation and research studies is similar in all fields.

If you can apply and remove reflection treatment, and if you can observe whether or not its presence makes a difference, you've got the basics of the scientific experimental method operating and if Toole or Bech or someone else were to undertake a study on the effects of treating first reflections in a normal room, that's certainly what they would be doing.

What else would they bring to the situation that isn't present in the individual situation? Probably measurement tools in order to correlate reported changes in perception with actual measured changes in reflection levels, and a standardised method of gathering data from test subjects. By standardised I mean nothing more than the fact that each subject would have their reports on what they perceived gathered in exactly the same way. The way that information is gathered would be determined in large part by the hypothesis underlying the study and the data analysis would be designed to test whether the data gathered supported the hypothesis or not.

If you're only concerned with whether or not treatment results in a more enjoyable and rewarding listening experience as I am, since I'm in this for my listening pleasure and so are most of the people posting on this board, you can dispense with the measurements. And since I'm not conducting a study and there is only one listening subject—me—when I make my comparisons, I don't need to have a standard method of gathering data to ensure that different people are reporting on the same things. There is only one person reporting and I can compare my perceptions on any individual aspect at any time.

This certainly means that the data I gather, and the conclusions I reach, aren't up to the standard of those one would expect from a peer reviewed paper but then one couldn't reach those standards anyway, regardless of what one did, with only one subject. It means that my data is anecdotal but, contrary to what you seem to think about anecdotal data, that does not automatically mean that it is unreliable. If you simply asked each one of Toole's subjects to describe their impressions at the end of a test session and wrote down those reports, what you would have is nothing more than anecdotal data yet it's precisely the data that Toole analysed to reach his conclusions. The only differences: the subject is asked to respond to specific questions, perhaps even in specific ways like marking a point on a scale or giving a clear yes/no answer, in order to allow data from different subjects to be combined, and the resulting data set from a number of subjects statistically analysed.

If a subject's data does not agree with the conclusions reached after statistical analysis, it does not mean that their reports were unreliable. It simply means that their perceptions did not agree with the overall impressions formed by the group as a whole.

So, when I or Ethan of someone else gives our report on how we view the effects of treating reflections, what we give is a non-standardised report by a single subject. It tells what our experience was and you certainly can't generalise from that. When you have anecdotal reports from a number of people and they all tend to agree in one or more aspects, you're getting reports that you can start to generalise from as the number of reports increases. The problem can be determining exactly what the common points really are. That is not likely to be a problem if you're simply interested in whether or not people who have treated reflections find the results of that treatment beneficial to their listening experience because that aspect of their reports is usually quite clear. What you will have a lot more difficulty determining is which aspect of the changes in sound introduced by the treatment are most significant and contribute most to the subjects' overall impression. For that you really need to gather the data in a standardised manner but what you are gathering is simply each subject's own anecdotal data.

So, to respond to your criticism about method and meaningfulness of results, I say that:

- individuals can approach treatment of reflections on an individual basis in a manner consistent with the basic approach to experimental method;

- their perceptions of the changes that result with treatment are reliable at an individual level and are certainly meaningful for the individual and capable of providing them with practical information on whether or not such treatment suits their particular needs in the situation;

- the reports of many people who have tried such treatment individually can certainly yield enough information to provide data for a reliable conclusion on any point that can clearly be established in each report;

- the reliability of any conclusions reached in that way depends on the number of subjects and the uniformity of their responses. Increases in the number of subjects and in the uniformity of their responses both increase reliability in the same way as they do in more formal studies.

- variables to be considered are going to be room variations, differences in the nature of the treatments applied, and differences in the placement of those treatments in the room. These variables will result in the need for caution in interpreting the data but they will not necessarily negate a conclusion, especially if the data is highly uniform;

- the data obtained and the conclusions reached can never be as detailed and specific as the data and conclusions of a formal research study, but they may certainly be good enough to determine whether or not most people are likely to find treatment of first reflections beneficial to their listening experience and that is a conclusion about preferences for treated vs untreated reflections which does not provide any data at all on the ways in which the presence of absence of first reflections influences perceptions. It just says something on whether or not people like or don't like/enjoy or don't enjoy the results of such treatment.

I think there are more than enough consistent reports to conclude that many people, if not most people, do find treatment of first reflections beneficial and even to justify recommending treatment of first reflections to people with certain preferences for the way they like hearing recorded music presented.



David Aiken

 

RE: Re:If you have a domestic size room, modes are always a problem, posted on January 30, 2008 at 00:52:42
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 >>The first width modes are at 37 and 72 Hz, per equation. When playing the tracks I get strongest response at 36 and 73 Hz, the 73 Hz mode being a bit weaker than the other. My speakers have two separate mute funcions, actuable via remote control, both can be set to any value between 0 and 128 dB attenuation. By default they are set to 10 and 20 dB, respectively. When I compare the strongest response at 36 Hz with the weakest response I find that the difference is more like 10 dB, not 20.<<

>I assume you're saying that the difference between the 2 modes is more like 10 dB.<

No. What I wanted to say is the following:

Below 100 Hz you have easily detectable interference patterns  standing waves when you play pure sine waves. Some of these standing waves are resonant, called modes. At listening position the strongest response I got was at frequencies close to the calculated width modes. The weakest response I got some other standing wave frequency, possibly midway between the two modes. This weakest response was subjectively about 10 dB weaker than the 36 Hz mode, this based on comparison with the level of the 36 Hz mode when attenuated by 10 and 20 dB, respectively.

Of course, to be sure one would have to measure, but I can say with certainty that the mode was only 10 dB (roughly) louder than the normal standing wave, not 20 dB.


>The difference in the ear's sensitivity at 36 Hz and 72 Hz can easily be 10 dB or more depending on SPL. You simply can't reliably judge the difference in level of your 2 modes by ear, you need measurements in order to get a reliable result.<

The 73 Hz mode was a bit weaker than the 36 Hz mode, this also I can say with certainty.

>>The sound waves that hit the membrane of a mike are very different from those that hit the ear drums, because of torso, hair and beard, pinnae, ear canals.<<

>I think those differences are basically irrelevant. Microphones/measuring equipment respond differently to the way the ear/brain system responds and the differences in how they respond are really so great as to make any differences between the actual sound waves impinging on the mic diaphragm and the ear drum irrelevant.<

These differences are anything but irrelevant: a mike is simply not measuring the same thing!!! These differences + the fact that signal processing is very different in human hearing are the reason that you cannot take simple SPL measurements and say, “look, we have a problem here”. Mikes operate in a linear manner, human hearing does not, as the curves of equal loudness prove. Which means that on a response graph you may see a 20 dB difference at 30 and 50 Hz, yet you may actually be on the same equal loudness curve, hence no subjective difference for the ear. If you further look at in-ear SPL measurements, you will see that the response is anything but flat. They look more like the on-axis response of a very bad loudspeaker, some frequencies attenuated, some peaking. Furthermore, no two individuals have the same in-ear response, which may explain why some people hear things which others don’t.


>>When I switch the signal off, the mode is gone within one blink of the eye but then, my speakers don't ring<<

>Well, I think every driver exhibits some ringing but the reason the mode is gone is not because the speaker "doesn't ring" but because the decay time of the mode in the room is short enough for you not to notice the time it takes for the sound to decay to inaudibility after the speaker stops emitting a tone. If we're talking about modal behaviour in a room, we're not interested in how quickly the speaker stops making sound but how long the room's modal behaviour causes the tone to persist after the speaker stops making a sound.<

If you look at the graph you see that all three drivers are completely silent after some 0.8 ms or so (the graph goes down to 100 Hz, the bass driver covers the range up to 580 Hz, there's hence no output at 30 Hz after 0.8 ms), whereas normal loudspeakers are still producing output at 3 ms, maybe more. Which means that these loudspeakers are simply driving the room modes longer.


>As an observation, measuring decay times in terms of eye blinks seems particularly unhelpful since the degree of resolution of the measurement unit is simply too gross for the measurement to be useful or informative.<

When talking room modes one often gets the impression that people think that it takes seconds for a mode to decay. I don’t have (yet) any data on mode decay time but the wink of an eye was merely to indicate that a mode decays rather quickly. FYI, RT60 in domestic furnished rooms has been measured:

Burgess et al., “Reverberation times in British living rooms”, Applied Acoustics 1985, vol. 18, p.369
Diaz, “The reverberation time of furnished rooms in dwellings”, Applied Acoustics 2005, vol. 66, p.945

Burgess measured values below 0.5 seconds, Diaz below 0.6 seconds. In my own room the calculated time is between 0.4 and 0.5 seconds.


>>Standing wave and modes are steady-state phenomena, how long an impulse does it take to excite a mode?<<

>I think that's the wrong question. In my view the important question is how long does it take a mode to decay once it has been activated.<

If modes are not excited because the impulse is too short the it doesn’t matter how long it takes for them to decay. You seem to assume that the simple fact that we can calculate room modes means that these modes will be excited, regardless of circumstances. I do not have any particular bass treatment, apart from that ceiling, yet I don't hear any modes when playing music, apart from the one I mentioned previously and that might as well be a simple standing wave.


Klaus

 

RE: Re:If you have a domestic size room, modes are always a problem, posted on January 30, 2008 at 06:34:32
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 >The 73 Hz mode was a bit weaker than the 36 Hz mode, this also I can say with certainty.<

A bit weaker I'll accept. Your earlier quantification I wouldn't, for the reasons I said.

>These differences are anything but irrelevant: a mike is simply not measuring the same thing!!! These differences + the fact that signal processing is very different in human hearing are the reason that you cannot take simple SPL measurements and say, “look, we have a problem here”. Mikes operate in a linear manner, human hearing does not, as the curves of equal loudness prove. Which means that on a response graph you may see a 20 dB difference at 30 and 50 Hz, yet you may actually be on the same equal loudness curve, hence no subjective difference for the ear. If you further look at in-ear SPL measurements, you will see that the response is anything but flat. They look more like the on-axis response of a very bad loudspeaker, some frequencies attenuated, some peaking. Furthermore, no two individuals have the same in-ear response, which may explain why some people hear things which others don’t.<

Well, let's hit the points:

-microphones don't measure, but they can be attached to something that does;

- the ear does not measure anything either, it responds to changes in air pressure;

- I clearly stated that the ear/brain handles things very differently to the mic/meter;

- I also didn't say at any point that you can take simple SPL measurements and say there is a problem.

- I am clearly aware of the Fletcher Munson curves since I actually referred to them in response to your comment about determining the scale of difference between your 2 peaks by ear.

- I'm also aware of the differences in hearing between different people.

You dishonestly imply that I'm saying things that I didn't say, and that I'm unaware of things that I actually mentioned, and then you mount arguments against things I never said and imply that I'm making mistakes because I'm not aware of things like the Fletcher Munson curves to which I referred.

What I said is that the processing differences in the brain and what they mean for how we perceive sound are so much different to what happens in a microphone/meter situation that in my view they render the changes introduced to the soundwaves in both systems as insignificant in comparison. I still think that's the case. In any event the only things we're really concerned with in this discussion are what the meter or software attached to the microphone tells us and how we perceive the sound after the brain's processing has been done.

>If you look at the graph you see that all three drivers are completely silent after some 0.8 ms or so (the graph goes down to 100 Hz, the bass driver covers the range up to 580 Hz, there's hence no output at 30 Hz after 0.8 ms), whereas normal loudspeakers are still producing output at 3 ms, maybe more. Which means that these loudspeakers are simply driving the room modes longer.<

Well, first I find the graph unreadable at the size it shows in your post so that doesn't help.

As for your conclusion about some speakers driving the room modes longer, there's no doubt at all that output at the relevant frequency from a speaker will drive a mode but once that output stops, modal behaviour will be the same for all rooms provided speaker output was identical in level. If one speaker does not stop emitting sound as quickly as another, we're talking about a difference in speaker behaviour rather than modal behaviour.

Because of that, your rather trite comment that "If you have problems with modal ringing, you should perhaps buy better speakers!" is simply inaccurate. If you've got modal ringing in the room, changing speakers won't fix it. It may fix a different problem, one concerning the accuracy of the speaker's response, but it won't change the modal ringing in the slightest. You've made a statement suggesting that changing speakers will change room response characteristics which it simply won't do.

>When talking room modes one often gets the impression that people think that it takes seconds for a mode to decay. I don’t have (yet) any data on mode decay time but the wink of an eye was merely to indicate that a mode decays rather quickly. FYI, RT60 in domestic furnished rooms has been measured:

Burgess et al., “Reverberation times in British living rooms”, Applied Acoustics 1985, vol. 18, p.369
Diaz, “The reverberation time of furnished rooms in dwellings”, Applied Acoustics 2005, vol. 66, p.945

Burgess measured values below 0.5 seconds, Diaz below 0.6 seconds. In my own room the calculated time is between 0.4 and 0.5 seconds.<

Well, as I pointed out, I have difficulty blinking twice in a second and I was aware of those measured times. I stand by my comment that there's nothing surprising about your claim and that the 'blink of an eye' is a useless measurement scale.

More importantly, however, you made an accusation about the mode disappearing in "a blink of an eye" and implied that this was due to the fact that your speakers don't ring. It isn't. It's due to the decay characteristics of your room. How fast the speaker stops emitting sound doesn't change the room's modal behaviour though, as I said above, output from the speaker certainly excites and maintains the mode while it occurs, but sound emission from the speaker is not modal behaviour, it is speaker behaviour. You can't cure room modal behaviour by changing speakers, unless you want to change to a speaker so challenged in the low frequencies that it has no output at any frequency low enough to excite a room mode.

In any event, the difference between the 0.8 ms in which your speakers stop emitting sound, and the 3.0 ms that you suggest other speakers take is simply insignificant when we're talking room decay times of 400 to 600 ms based on the papers quoted and your measurements of your room. Do you really think a further 2.2 ms of sound before the decay to silence is going to make an appreciable difference when we're talking about sound dying in the room say 452.2 ms after the signal from the amp stops rather than say 450 ms after that event? Get real. The difference in speaker characteristics you're referring to is insignificant in comparison to the actual room decay time and, besides that, you're putting up the 3 ms time for other people's speakers as if you've got evidence for that and you have no data on their speakers or, if you have, you haven't referred to it. You've simply pulled up a figure and applied it to the speakers of the people you're arguing against, implying that their speakers are significantly worse than yours on this parameter. Then you've stated that that small difference, if it actually is present, is so significant that changing speakers will make a noticeable difference in how quickly modes will decay in their rooms. That's rubbish, and dishonest rubbish at that.

Also, by saying here "When talking room modes one often gets the impression that people think that it takes seconds for a mode to decay" you admit to introducing a view that neither Ethan or I had expressed. You simply introduced it, implied that we held it, and then tried to knock us down, a favourite attack of yours and, as I've pointed out above and in previous threads, it's a dishonest approach to the discussion.

>If modes are not excited because the impulse is too short the it doesn’t matter how long it takes for them to decay. You seem to assume that the simple fact that we can calculate room modes means that these modes will be excited, regardless of circumstances. I do not have any particular bass treatment, apart from that ceiling, yet I don't hear any modes when playing music, apart from the one I mentioned previously and that might as well be a simple standing wave.<

Once again you start from a dishonest position. My comment to which you were responding was "In my view the important question is how long does it take a mode to decay ONCE IT HAS BEEN ACTIVATED." I've added the capitalisation for stress. I made no suggestions whatsoever about situations where the mode was not excited. I specifically said that what is important is the decay time once it is activated.

I also said nothing at all to suggest that I think that the fact that we can calculate modes means that they will be excited. In fact I've previously stated in these discussions that it won't be excited unless that frequency, or one close enough to it to be within the mode's bandwidth, is present in the music being reproduced.

And I made no comment about the cause of the modes in your room.

The whole of your response to my statement on this point fails to deal with what I did say, and accuses me of saying things I did not say.

Sorry Klaus, but I've raised your propensity for this dishonest approach in previous discussions and I'm sorry to see it occur here again, but that ends it for me. If you can't be accurate in referring to your opponent's views and statements, then there's simply no point in continuing the discussion. You're no longer responding to what I've said and you're actually misrepresenting what I said, and I'm not going to waste any more time responding to that approach. I've dealt with it in depth here so other readers can see your tactics for what they are but this ends the discussion for me. You're back to what I regard as dirty tricks and I'm not not going to go along with that.




David Aiken

 

RE: Research and real life, posted on February 6, 2008 at 02:14:22
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 >I don't think you have any real idea of what experimental method really is so I'll ask: have you ever done any research at all in any field? I have.<

So do I! My engineering master thesis was about surfactant solutions (used in enhanced oil recovery) in porous media. Almost one year in the lab! I found out that research work was not my piece of cake.


>If you can apply and remove reflection treatment, and if you can observe whether or not its presence makes a difference, you've got the basics of the scientific experimental method operating...<

The problem is that when treating reflections you have three different yet co-existing changes: loudness, timbre, spatial impression. Yes you will hear an overall difference, but you probably would not know what is the cause without a stringent protocol.

>So, when I or Ethan of someone else gives our report on how we view the effects of treating reflections, what we give is a non-standardised report by a single subject. It tells what our experience was and you certainly can't generalise from that.<

You or Ethan seem to assume that reflection treatment is beneficial, regardless of the initial conditions. Ethan goes even further by using inappropriate and hence misleading evidence to prove the detrimental effect of reflections.

I’m currently doing some simple experiments to see the effects of reflections on stereo image , when finished, I’ll post the results.


Klaus

 

My experiments, posted on February 6, 2008 at 05:23:27
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 I made some experiments using 2 tracks especially recorded for stereo imaging from EBU SQAM (Verdi) and Chesky Jazz Sampler (male speech). The Chesky track positions the talker centred between loudspeakers, midway between centre position and loudspeaker, directly at the loudspeaker, and off-stage. The Verdi track has three soloists and two small groups of male and female singers, respectively.

1. Genelec 2-way desktop monitors in a small room, desk midway between side walls, nearfield listening. I added lateral reflections by placing wooden room doors at first reflection points. No change for the centre, midway, at-loudspeaker positions. The off-stage position provided the best clue in that on the left side the talker was off-stage with the panels in place, almost at the loudspeaker without the panels. However, the room is not entirely symmetric in that there are no early reflections from the right but possibly reflections off a sloped ceiling on the left.

It should be noted that an exact location of the talker was not possible anyway, the voice came clearly from off-stage but not from a location I could put my finger on. No detectable changes with the Verdi track.

2. Moved the Genelec setup (loudspeakers on table well off the side walls) to large living room downstairs. When comparing the location of the Chesky talker with vs without room doors placed at reflection points, no difference. With the Verdi track, the positions of soloists and the two choirs were more precise, but no difference with vs without doors.

But then, in both cases the base distance of the Genelec was about 1.10m, listening distance was about 1.20 so possible changes in position were perhaps too small to be detectable. So I used my main system (base distance about 2.60m, listening distance about 3.50m). In this case I put absorbers (15 cm/6 inch bed mattresses) at the lateral reflection points.

Chesky: with the absorbers in place the left off-stage position seemed to change a bit in that the voice came more from the speaker than without absorber, hence a degradation. I noticed one detail that was not detectable with the small Genelec. When starting to speak off-stage on the right side, the talker breathes in before pronouncing his text. This breathing-in sound starts somewhere between centre and left speaker to then quickly move to the right off-stage position. No change of location detectable for this detail with absorbers in place.

Verdi: alto is actually at centre position, bass about midway to the right, baritone between bass and right speaker. With absorbers in place no detectable change in absolute or relative positions.

I then played a track from “Jazz from the Pawnshop (live recording, original Proprius LP). I didn’t get the impression that the absorbers had any effect, be it spaciousness, be it source localisation.


Klaus

 

RE: Research and real life, posted on February 6, 2008 at 13:58:03
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 I'm only responding to this because there are no misrepresentations of anything I've said in it apart from your statement that I or Ethan "seem to assume that reflection treatment is beneficial, regardless of the initial conditions". In fact, if you look back to one of my early responses to the original poster in this thread, I actually said that some people don't like the effects of treatment and that the original poster should go with their own preferences when it came to the treatment of first reflections. I have never stated that it is universally beneficial for everyone but, when it comes to what I personally prefer, then most certainly I do find it beneficial regardless of any initial conditions I have ever started from. I have never assumed that it would be the same for everyone, or recommended that everyone treat first reflections. My recommendation is always to treat if the result of treatment is greater enjoyment of music on your part.

>The problem is that when treating reflections you have three different yet co-existing changes: loudness, timbre, spatial impression. Yes you will hear an overall difference, but you probably would not know what is the cause without a stringent protocol.<

Yes, several things change but is that a problem? Do we need to know what the specific changes are? You can't treat reflections and change only one thing, and the question is really do you prefer the sound with or without the treatment. I prefer it with treatment and in my experience most people seem to prefer it with treatment also. And we do know the cause of the change without needing a stringent protocol: the cause is the treatment. What we don't know without a stringent protocol is what is the 'ranking order' for the effect of those changes for listeners and whether everyone has the same 'ranking order' or whether some people are more 'sensitive' to changes in one of those things than are other people. We also don't know how measurable differences in the type of treatment affect the scale of change in those things and whether it is possible to design treatments that favour change in one thing over the others.

Re your experiments:

Your first experiment with the Genelecs is problematic since you point to some room asymmetries in the situation without the panels in place. Also, from my own experience in asymmetric rooms with strong reflections on one side and not on the other, I've never achieved results I was happy with when I tried to add a reflective surface to the side without reflections. I have no idea of why that should be so, and I have to admit that I didn't try placing a similar surface at the side where I already had reflections, but after several experiments I simply gave up on trying to augment reflectivity as a solution to asymmetry. Absorption on the strongly reflective side alone always produced a better L/R balance for me and absorption on both sides an even better L/R balance.

It's also hard to comment on your experiments with the mattresses since the covering material may have been quite reflective at high to mid frequencies. The comment about hearing the in breath with the treatment but not without it is interesting and flags something that we haven't discussed—the ability to hear details that have previously been masked when treatment is added to a room.

Still, at the end of the day we're talking about a preference and it certainly is a preference. Some people certainly don't want the gains in soundstage and imaging that I hear in a treated room and that's fine. As I have pointed out frequently, the 'right' presentation is the one that satisfies the owner of the system since they have put together their system for their listening enjoyment. It makes no sense at all for them to set things up in a way which diminishes their enjoyment, whether that way be the addition of treatment or the lack of treatment.

Some final comment on your experiments. You've got strongly held views based on your reading and you've done brief experiments with 2 tracks. You're also strongly familiar with the sound of your rooms without treatment. While you have heard some differences, you probably haven't heard all of the differences that can be heard. I think for that you would need to listen to a wider range of material over a longer length of time, and also listen in a more normal frame of mind than when you're critically listening for differences. I often notice some things more strongly when I've almost been asleep and I'm coming out of that state to a more aware/alert state but I'm still deeply relaxed. That's not a state I can put myself in voluntarily but some of my strongest spatial impressions from music have come at those times. Finally, it's hard to change ingrained preferences. My parents always drank tea, never coffee, so I didn't taste coffee until I was in my late teens and I initially did not like it. It took a long time and a lot of exposure to coffee for it to become my preference over tea. Your preferences aren't going to change overnight, if they're going to change at all. What a treated room provides in 'benefits' to reproduced music may not be benefits to you and may conceal some of the things you like most in the sound you hear so continued exposure to listening in a treated room may reveal more differences that treatment results in, but may also not change your preferences for the sort of sound you prefer.

So I think your experiments are interesting. The results are unclear at one level, given the room asymmetry in the first experiment and questions about the absorptivity of the mattresses in the second but those are the kind of individual differences that exist when anyone tries treatment on their own. I found your comment on hearing the in breath more interesting than anything else because it is not covered in your "loudness, timbre, spatial impression" list of effects yet a common experience that people have when hearing familiar music in a treated room is the hearing of detail that had previously been hidden so there's a fourth effect for you to consider and one that has not received as much attention in the literature as the effects you mention, despite the fact that it is often reported by people who try treating their rooms. You've just managed to confirm one aspect of many anecdotal reports.







David Aiken

 

RE: Research and real life, posted on February 7, 2008 at 10:37:58
Ethan Winer
Manufacturer

Posts: 1709
Location: New Milford, CT USA
Joined: December 3, 2003
 > It took a long time and a lot of exposure to coffee for it to become my preference over tea. <

This is an excellent point David, and I often explain to people that a well-treated room can be an acquired taste. Especially people who are used to excess ambience and believe (wrongly IMO) that their own room's ambience is important for lifelike reproduction of music. But once you get used to hearing clearly all of the music and its inherent ambience as captured in the recording, most people will never go back.

--Ethan

 

RE: Research and real life, posted on February 11, 2008 at 11:59:49
jimgrocecfp@hotmail.com
 Well, its been a couple of weeks since I've revisited this thread, and thanks to all who have contributed to the debate. However, as usual I am left more confused than I was at the beginning. The differences in opinion on how to go about treating a room or whether or not to bother with it at all cause me to do nothing. Though I'm not technically trained on this issue, I have played guitar for years, solo and with other players, in good acoustics and bad, and have a very good ear for what sounds good and what doesn't.
Since my original post I have tried make-shift 1st reflection treatments (pillows stuck on guitar stands) and find that this makes for a slight improvement. Enough improvement to spend money on a product? Not sure. Again thanks for the replies.
Jim

 

RE: Research and real life, posted on February 11, 2008 at 13:54:43
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 Jim,

I'm sorry that things got diverted by what is actually long standing and probably not resolvable disagreement between Klaus and I, and also between Klaus and Ethan, about room treatment and especially treatment of first reflection points. Ethan and I are both pro room treatment and treatment of first reflections, and so is F. Alton Everest if you read his "Master Handbook of Acoustics" which I think should be essential reading for anyone considering room treatment or interested in room acoustics in listening rooms. Klaus on the other hand takes what I consider a fairly narrow view of the research he's read, especially that of Floyd Toole, and simply tends to state that room treatment is unnecessary whenever he enters a thread.

Ethan is extremely pro room treatment as you would expect from someone who makes and sells acoustic treatment products. While I'm more hesitant than he is to recommend treatment of first reflection points to everyone, since I have come across people who don't like the result and have come to accept that there's an element of preference involved, my personal position when it comes to my own room is completely in line with Ethan's and I think using absorption at first reflection points yields quite noticeable benefits in terms of imaging and clarity. I think the reason you only noticed a slight improvement with your experiments using pillows on guitar stands is simply that pillows aren't a particularly broad band absorber and they are also probably a little too small in area for effective treatment of a reflection point. I think you would be much more impressed with the results of panels which work over a much wider range and which are considerably larger, say around the 2' by 4' size as a minimum and preferably closer to 3' by 4' or even 4' by 4' in size.

If you look over posts here over the last few years, I think you'll find that with only a couple of exceptions, Klaus being probably the strongest of those, most people posting here are in favour of room treatment. Given that this board was established for discussing the practical aspects of room treatment that's really not surprising and the distribution of opinions represented here will definitely be strongly skewed in favour of treatment. In my experience the opinions of most people who have listened to music in treated rooms is also in favour of treatment but some people aren't prepared/willing to treat their own rooms because their room is a living room and there are certainly strong visual impacts from room treatment. Most acoustic treatment products are not attractive additions to a living room, and there are also often placement restrictions on where things can go in a living room given the existing furniture and the functions the room is serving apart from housing an audio system.

If I were asked for a priority order for room treatment, what I would say is:

- bass trapping is the highest priority and I think every room benefits from bass trapping

- if you're interested in improving imaging/soundstage and tonal accuracy, I think treatment of first reflection points is the second priority, especially in regard to side walls and the wall behind the speakers. I think ceiling treatment is probably a good idea, especially for lower ceilings but lower ceilings also give you less space in which to do it, but I don't have ceiling treatment in my room because I don't want to have to drill permanent mountings into the ceiling and I've yet to come across a way of sticking any acoustic treatment to a ceiling that will both hold up long term and not mark the ceiling. Every trick I've tried for holding foam tiles to the ceiling without marking or damaging the ceiling has eventually resulted in the tiles falling off, and anything other than foam tiles is even heavier and requires attachment techniques I don't want to use. On the other hand, if you find imaging and a clearly defined soundstage distracting as some people certainly seem to do, then I would expect that you might well prefer the sound of a room in which first reflection points are untreated.

- finally, after bass traps and treating first reflection points, I'd consider adding diffusion to the area behind the listening position if that's possible but diffusion requires space in which to develop and you really need around a minimum of 6' or so behind you if diffusion is going to be really worth while.

Klaus will almost certainly disagree with me on any recommendation for treatment. Ethan on the other hand would probably recommend more bass trapping than I would employ on the basis that bass trapping in corners always smooths the room response and that more trapping smooths things even more. He's right but I think my threshold for accepting the visual impact of a lot of bass traps is lower than his. I'm happy just to treat the 4 room corners though I do agree that treating the wall/ceiling and wall/floor junctions as well would certainly be more beneficial than just treating the wall/wall junctions as I do. That puts me in the middle between Klaus and Ethan: I agree with Ethan about the benefits of treatment and that more is almost always better than less, but I tend to be more concerned about the visual impact than I think he is.

I also think that the only person who has to be satisfied is the person who owns the system because they're the person who spends more time listening to it and the point of this hobby is musical enjoyment. If acoustic treatment doesn't add to your enjoyment, then don't do it and most definitely don't do it if it detracts from your enjoyment. The only other thing that needs to be said is that if you're going to add absorption to a room, ensure that the absorption is broad band and covers as much of the audible frequency spectrum as possible. The worst thing you can do is to use narrow band absorption, especially if it's basically only absorption of the high frequencies with little or no mid-range and bass absorption. The resulting sound will simply be tonally unacceptable with a dull top end and a poorly defined bottom end, possibly with room decay times in the bass frequency that destroy the clarity of legato playing by the musicians. The basic rule, if you're going to do any acoustic treatment at all, is to do a good job on the treatment that you do implement. If you can't do a good job, don't do it at all because the results may well be worse with a bad job than they are with no treatment at all.

So definitely keep reading things here, and reading Everest's book if you can, and keep thinking about the topic. Try what you can when you can, and definitely don't forget that the point of everything is always musical enjoyment. If you're not getting that you're doing something wrong. You can certainly enjoy music in an untreated room but I think you will get more enjoyment in a treated room. Best wishes in your quest for better sound.




David Aiken

 

RE: Research and real life, posted on February 12, 2008 at 06:14:12
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 >Your first experiment with the Genelecs is problematic since you point to some room asymmetries in the situation without the panels in place.<

Actually, the sloped ceiling (which in fact is the roof, the room is on 2nd floor, just below the attic) covers only about 50% of the room, when making the test I was sitting at the room’s midpoint, the sloped part of the ceiling was on my left and towards my back. It may have been possible that this asymmetry had an effect, so I don’t draw any conclusions from this scenario. But then, also on my main system, the left off-stage illusion was less convincing than the right. It may be the recording, it may be the CD-players (Philips walkman for Genelec, Tascam CD-recorder for main system).


>It's also hard to comment on your experiments with the mattresses since the covering material may have been quite reflective at high to mid frequencies.<

The covering is made from cotton, but yes, I don’t know the absorption characteristics, so I won’t draw any conclusions other than that mattresses don’t have an effect on phantom source location.

>The comment about hearing the in breath with the treatment but not without it is interesting and flags something that we haven't discussed—the ability to hear details that have previously been masked when treatment is added to a room.<

I wrote:, “This breathing-in sound starts somewhere between centre and left speaker to then quickly move to the right off-stage position. No change of location detectable for this detail with absorbers in place.”

I don’t quite see how this could be understood as that detail being masked and becoming apparent when treatment is added, but perhaps I wasn’t clear enough: the in breath was clearly audible with and without the mattresses, in both cases it started somewhere between left speaker and centre, then moved to the right off-stage position where the talker then pronounced his text. During the test I did not notice it with the small Genelec. I just checked it out, now that I know it’s there, it’s clearly audible on the Genelec.

>Some final comment on your experiments. You've got strongly held views based on your reading and you've done brief experiments with 2 tracks. You're also strongly familiar with the sound of your rooms without treatment. While you have heard some differences, you probably haven't heard all of the differences that can be heard.<

These two tracks were designed/intended to examine stereo image, so I focussed on this aspect only. If treatment changes sound stage I guess it would be detectable on such tracks. My mattresses didn’t. In a next step I’ll try the inner foam core alone.

>Your preferences aren't going to change overnight, if they're going to change at all.<

Of course, but in this case I did not prefer anything about something else because there wasn’t even a change, at least not in stereo image.


>I found your comment on hearing the in breath more interesting than anything else because it is not covered in your "loudness, timbre, spatial impression" list of effects yet a common experience that people have when hearing familiar music in a treated room is the hearing of detail that had previously been hidden so there's a fourth effect for you to consider and one that has not received as much attention in the literature as the effects you mention, despite the fact that it is often reported by people who try treating their rooms. You've just managed to confirm one aspect of many anecdotal reports.<


Note that the effect did not change with the absorbers in place. I suppose that for each of the positions (centre, midway between centre and loudspeaker, at loudspeaker, off-stage) the talker stood right in front of the microphone and the Chesky engineer obtained the effects afterwards by modifying interaural level/time differences via the mixing console. In his book about recording studio design Phil Newell presents anecdotes where a sound engineer hears certain spatial effects and Newell doesn’t. I’m speculating now, but maybe the Chesky engineer was simply not aware of the breathing-in sound which sweeps across the stage, maybe the circuits used for the off-stage effect did not cover the frequency range of the breathing sound, in any case it is strange that the talker is breathing in at a left position when half a second later he is talking from the right side.

As for the preference issue, in a 1990 paper Toole mentions “timbral corruption” which means that the off-axis response is very different from the on-axis response in which case “timbral considerations alone demand absorption”. In some Stereophile reviews you read comments such as:

“Higher in frequency (and measured using a calibrated B&K 4006 microphone), the Dynaudio [Contour 3.0] has a slight but broad peak in the upper midrange. Depending on the music, this might also be perceived as slight lacks of energy in the lower midrange and presence region. However, as can be seen from the graph showing the speaker's lateral dispersion (fig.3), the off-axis behaviour goes a long way toward compensating for the on-axis aberrations.”

This looks to me like a case of timbral corruption, and instead of critical comments this is considered to be just fine and well thought. Timbral corruption: a possible common factor in audiophiles’ anecdotal evidence re: benefits of reflection treatment? In cases of timbral corruption I understand the preference for the sound with reflection treatment applied, but in this case there is a problem that makes treatment necessary.


Klaus

 

RE: Research and real life, posted on February 12, 2008 at 16:01:16
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 I misread part of your post regarding that in-breath. What you actually said, starting one sentence earlier than the quote in your reply, was "I noticed one detail that was not detectable with the small Genelec. When starting to speak off-stage on the right side…". I somehow skipped over the phrase "with the small Genelec" and read the comment as referring to something you hadn't noticed on the speakers you were using downstairs. On rereading your whole post you started out with the small Genelecs upstairs, said you moved them downstairs, and then don't mention changing to any other speaker but it seems you did. I certainly missed a phrase in your post which makes my comments wrong, but you seem to have omitted mentioning a change from the Genelecs to a different speaker which makes your post difficult to understand. I suspect the reason I thought you were still listening to the original set of speakers is that you hadn't mentioned changing from them to something else so I interpreted that phrase as meaning something you hadn't previously noticed.

I'm not certain what Toole means by "timbral corruption" but I wouldn't have thought a change in dispersion at the crossover point was what he meant, especially when it is obvious that the change in dispersion is being used to fill in a dip in the on-axis response. In the case you mention, taking Toole's advice and absorbing too much reflected sound is going to actually make the speaker sound worse. If you're going to use absorption with a speaker that behaves that way, then you're going to want to apply absorption carefully, either by tailoring the absorption spectrum so that it doesn't absorb too much of the 'fill in' bandwidth or by limiting the absorption area so that a major part of the room's reflective area remains operational.

So, in the case of the speaker you referred to, I would disagree that the off axis response makes treatment necessary since that off axis behaviour is apparently intended to contribute to the listener's perception of the speaker's tonal balance.

Anyway, see the response from the original poster. This discussion has outlived its welcome so let's call it quits.

I think if we're going to debate technical points here in future we're better off starting a separate thread rather than cluttering up the responses to practical questions with theoretical disccusion. That doesn't mean you shouldn't object to a practical response if you feel the need to, but I do suggest that you simply do a post saying that you object and providing a link to a new post detailing the theoretical issues so that the discussion of them can be considered in a separate thread that readers can follow through on if they so wish.



David Aiken

 

Main and 2nd system, posted on February 14, 2008 at 04:43:40
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 In my description of the experiments I mentioned that I used my main system for the third part of the test, thinking that that was clear enough. Both systems are listed in inmates' systems.

What Toole means by timbral corruption is that the speakers have an off-axis response that is very different from the on-axis response. It would appear that there are designers who focus on on-axis and neglect off-axis. This is the reason why you need to see individual graphs of both responses, an in-room measurement, like some folks around here seem to believe, will not provide the necessary information.

I agree that the Dynaudio Contour 3.0 is not a worst case scenario, I have seen much worse in Stereophile's pages. However, a well engineered speaker should NOT show any on-axis aberrations that need compensation.

FYI, I finally got one of Toole's papers where he briefly describes the (stereo) continuation of his and Olive's experiments re: Detection of reflections in typical rooms. I'll post on that later in a separate post.


Klaus

Edit: I've also made a final experiment with 5.5" mattress PE-foam cores, which should be effective down to about 600 Hz, no detectable effect with the two stereo image tracks.


2nd edit: I'm still working on the literature review re: early reflections. Once it's finished, I'll post it here so I can include a link in any future post on this controversial topic.

 

"However, a well engineered speaker should NOT show any on-axis aberrations that need compensation", posted on February 14, 2008 at 16:34:33
David Aiken
Audiophile

Posts: 5858
Location: Brisbane
Joined: September 25, 1999
 I disagree.

A well designed speaker should not show any aberrations which need correction when set up and listened to in accordance with the designer's recommendations.

Designers choose different drivers and crossover circuits for specific reasons and no driver or crossover is perfect. The designer can also choose to use the room in various ways such as corner loading or using reflections from the off-axis response in order to help them achieve that aim. At the end of the day what counts is what the listener hears when the speaker is properly set up, and the designer is responsible for choosing components and assembling them into a speaker which delivers tonally accurate sound amongst other things when it is set up and used in the way in which it was intended to be used.





David Aiken

 

RE: "However, a well engineered speaker should NOT show any on-axis aberrations that need compensation", posted on February 18, 2008 at 06:59:37
KlausR.
Audiophile

Posts: 2154
Joined: November 17, 2004
 1. It is the direct sound, and hence on-axis response, which triggers perceptual processes like the precedence effect. It is the direct sound that serves as reference for the perception of later sounds.

2. It is tempting to think that, for instance, an on-axis trough of 3 dB spanning 1 octave can be compensated by a 3 dB plateau off-axis spanning the same octave. However, unlike a simple microphone human hearing is highly directional, meaning that it does not simply add up the amplitudes of the trough and the peak to arrive at a linear response! Such compensation would work if in-ear response was independent of direction of incident, which is isn't. If you go and measure in-room you will, of course, obtain a flat curve because the microphone simply adds up the two amplitudes in a linear manner. For the average reader and audio reviewer this apparently is sufficient proof that the compensation works. Manufacturers have an agenda, so they can't be entirely trusted. That's why the audio reviewer should have some knowledge in the field of physiology of hearing to be able to recognize such incorrect stuff and inform the reader accordingly.

Clearly, loudspeaker designer and audio reviewers who think that such compensation can actually work as intended have not done their homework!

Shaw, “Earcanal pressure generated by a free sound field”, J. of Acoust. Soc. of America 1965, vol. 39, no.3, p.465
Muncey et al., "The listener and the room”, J. of Sound and Vibration 1964, vol. 1, no.2, p.141


Klaus

 

Page processed in 0.027 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]