|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
173.33.186.185
From most of the systems that I saw online, most of the Giant Horn Speakers system didn't do anything regarding the Time Alignment issue. All the drivers are not time aligned at all. One driver here and the other over there. The worst offenders seem to be those ultra expensive big horn speakers systems from Goto or ALE.Is time alignment in horn speakers important at all ?
Thanks....
Edits: 03/06/12Follow Ups:
I am starting on my second set of speakers. The first was Bob Crites Cornscala D. The idea is that the 2 way speaker with a 500 hz xover is going to sidestep the time alignment problem. It does seem to work but 500 hz is still a middle C. So there is lots of music below 500 hz.
The design that I am starting is going to use to 15 woofers and the same Faital 140 and 142 HF driver/horn. I going to try a 900 hz xover on this one. So perhaps the horn should be moved forward? My plans are to make the boxes more of a test - quick construction - and make the horns so I can experiment.
If you want to see my first speakers go to the trader and no I am not trying to sell you speakers!
fortcollinsaudio.com
Just for the sake of accuracy, 500 Hz is closer to a "B" (493 Hz) in the standard treble clef staff. The "C" just above it is 523 Hz, but, you referred to it as "Middle C". "Middle C", as used in music parlance, is the "C" just below the treble staff, and is 261 Hz, not the one at 523 Hz.
(All stated frequencies ignor the fractional component.)
Carry on.
:)
DonIf you want to experiment with time delay, follow the testing I used for my 605As.
You will need a microphone, preamp, and a O'scope. I included the test signal on my web site.Don
http://web.mac.com/donaldpatten
Edits: 03/13/12
I had thought that a correcting xover was too hard to pull off. Looks like its an option from your site. Thanks for the link. This is a smart group.Inmate51 - Thanks for correcting my C/middle C mistake. I will be more carefull. I am really just a woodworker and .net programmer.
Don Walker
Edits: 03/14/12
here is a good graph for frequency reference:
There's going to be a lot of disappointed professional trumpet players when they find out they're not supposed to play above C6. ;)
Wayne Bergeron, Malcolm McNab, Jon Faddis, Eric Miyashiro, and Doc Severinsen come to mind, among others. Heck, even little old me can hit an E6.
But I'm just jerkin' your chain... there are as many versions of these charts as there are musicians in an orchestra. Many of them do a decent job of showing the approximate range of instruments, but it's important to remember that it is approximate, and just for general guideline purposes.
:)
Hi Don,
We see that you are new here, so welcome, and thanks for registering as a member of trade. Unfortunately, your moniker violates one of our terms of use:
"No member of the trade shall use his/her business name, postal address, e-mail address, telephone number or URL as part of his/her moniker or as part of any post, except as permitted in rule no. 5. This Rule includes magazine and e-zine writers and editors. Trade members may use the "info" link to make this information available and may use their company email addresses for registration, but must use a different moniker/username than the default email address for their posts."
Please revise your registration accordingly.
And, linking to your site in your posts is a no-no, as well.
Thanks!
Fax mentis incendium gloria cultum, et cetera, et cetera...
Memo bis punitor delicatum! It's all there, black and white,
clear as crystal! Blah, blah, and so on and so forth ...
Time alignment is VERY important IMO. I'm using AG Omega Duos. About a year ago I got wooden blocks built for the tweeter cylinders which I have put on the bass speaker allowing me the move the tweeters back in small increments. I ended up moving the tweeter back around 20 cm. Strangely 0,5 cm matters and is easily audible when you get close to the right point! A friend who owns the same speaker recently came over. He listened first with the tweeter aligned, then I hit pause and moved the tweeters back to it's original position and hit play. After a few seconds he said "oh shit". The difference is quite dramatic!
Using digital delay for the horns to align with my corner subs crossed over at 200 hz. Interestingly even at 200 hz a few cm matters and can easily be heard!
In my own system, which was all an experiment, I was trying to get coherent wide range without any digital time correction since the system was intended for analog playback.
Based on some very old test data Dr. Edgar showed me of human sensitivity to time delay at varying frequencies I made sure that differences between bass/midbass and midbass/midrange fell within the imperceptible range. The tweeter and mid are physically aligned with a first order crossover.
In order to do this the bass horns were designed with the mouths flaring to compensate for the volume of the 3-way main arrays to sit in their mouths. There is ~8ms difference between the bass drivers and the midbass drivers. According to the test data this is not perceptible at 65Hz, which is the crossover point.
I forgot the exact difference at the midbass/mid point, but I got it within the tolerances needed. The human ear is most critical of time differences in the midrange and can detect less variation at either ende of the audible spectrum: it graphs to a bell-shaped curve.
It isn't pretty, but those who have heard my system are generally impressed with it's imaging and detail. In addition to considering these things before building I got really lucky too. I had always intended to rebuild it to look better... funny how having kids puts some projects on the back burner.
eso
They were a carnival of American decay on parade, and they had no idea of the atrocity they had inflicted upon themselves.“ Henry Chinaski
I am also trying to stay all analog at this time. This alignment took me a while to a) wrap my head around what I was hearing and, b) how to setup my horns consistently. The mid-bass/mid-range, obviously, seems to be the most critical and I've taken to using a mic at the listening position and a pc-based o-scope to get the drivers aligned very close at the crossover frequency, around 600hz. This has worked well for me so I've also been attempting to do some alignment between right and left side using this technique. My room is asymmetrical, l-shaped, so getting things close with a 350Hz tone in the mid-bass seems to give me better right/left alignment in my room. I may not be spot on but I'm getting a single, solid, woman's voice on opera passages. Used to sound a bit like a herd of sheep. :-)
In my experience (so yes, it is purely anecdotal) having the tweeter coil x inches in front of the mid or woofer (for 2ways) coil is worse than having it behind by the same distance.
I have no idea why exactly that would be the case but it is true for my ears.
"I have no idea why exactly that would be the case but it is true for my ears."
Your ears are not lying to you either.
The signal from a crossover naturally has the hf emerge first, then the mids and last the lowest frequency range.
To make a multi-way speaker appear on axis to have a single arrival time over a broad frequency range, requires DSP trickery or like we do, off set the drivers front to back in a complementary way which compensates for the inherent phase shift / time delay normal crossovers produce.
Good to know that my ears were not lying, I find it hard to trust them on their own (as should everybody IMO).
Last time that happened was when I adjusted the the phase response on my active xovers by ear.
I then measured what it should be and it turned out that I got it pretty much right by ear.
Congratulation !
This is a surprisingly hard area to consider for a number of reasons, including that one is ultimately judging a speakers accuracy based on how one imagines the recording should sound without knowing what the original event sounded like or what the artists had in mind and NOT the accuracy with which the speaker reproduces an input signal.
At least when one makes their own recordings, you have a good idea what it sounded like live.
Ignoring the most discussed aspects initially;
A difficult part about dealing with sound is that you can’t see it and most people have no idea that one is dealing with something which also has such a wide range of sizes depending on the frequency.
If one thinks of a sound wave as Russian doll (the dolls which stack one within another), one can think of 20 KHz as a 5/8 inch tall doll (the wavelength is about 5/8 inch) while at 20Hz, the doll is 1000 times larger.
A sound wave also have physical volume and so if one lowers a frequency an octave, then the volume the wave occupies is cubed.
This is why if one lowers the cutoff on a woofer an octave with the same cabinet volume, one lowers the maximum efficiency by -9dB or to keep the same sensitivity, one needs to cube the volume.
Once one ventures beyond the acoustic power of a headphone driver, one finds you really can’t accurately reproduce the entire spectrum with one single driver, the span has contradicting requirements. The obvious answer is to accept less than the entire bandwidth and other compromises or use multiple drivers each optimized for the frequency range it will operate in.
Many think of this as something like you use high pass and low pass filters and the acoustic signals from the speakers add like you were adding a signal through resistors etc.
Actually loudspeakers can add like resistors, coherently, BUT this only happens when the sources are less than a quarter wavelength apart (like how subwoofers add coherently if you put two identical ones close together). As soon as the spacing is larger than about 1/3 to 1/2 wavelength apart, the two sources radiate as two independent sources and now they produce an interference patter which is a series of lobes and nulls in it’s dispersion pattern.
The summation depends on where on is and in this case, crossover optimization has to include making sure that in the polar patterns, one doesn’t have a null on axis in the crossover region.. The other lobes and nulls are ignored generally as they can't be dealt with but still excite the room and detract from the direct to reflected sound ratio.
One can’t measure stereo imaging but you can hear if a system has it of not, in a addition to all the other flaws a speaker can have, where the sound goes has a powerful effect on if you have good imaging or not.
By good, to me that means if you play a mono signal, one gets a solid phantom image in the center ideally convincing enough to sound like a speaker were in that location. By not good, I mean the same signal produces a impression the source were wide or at worst, a center phantom and an obvious source on the right and left.
Here, horns (or other large directive sources) can have a big advantage as some of what governs this is how strong the side wall reflections you have at the listening position. If one has a normal cone/dome speaker system the dispersion pattern puts a large amount of energy outside the listening position and that energy is in reflected sound which competes with the actual direct signal.
If one sets that kind of speaker system up outdoors (where one has no room effects) , the difference in stereo imaging can be stunning.
To answer the gist of your question;
In general loudspeakers have so many audible / measurable flaws that there is no one universal “best way”.
The ability to preserve the time aspect of the input signal is possible but only with a single source, with a multi way system with sources radiating from different points in time and space, the best one can do with DSP is a correction that can be 100% right but only in one small spot.
There is also an argument that when one simulates the phase rotation crossovers produce or simulating the “all pass” phase rotation speakers cause, that one can only hear that under certain conditions.
On the other hand, listening to that simulation on headphones as is done is not the same as hearing that result produced by two separate sources.
To a degree we can hear the interference pattern a speaker radiates.
A speaker that radiates as a single simple point source can do something interesting. If you have heard an ESL-63 playing a quiet voice, you may have noticed the voice really sounds like its coming from behind the speaker. The speaker does not radiate an inference pattern and so it’s physical identify is more hidden (the physical depth location less easily localized by ear).
In a stereo pair in the test above(sans significant reflected sound) these speaker produce a strong mono phantom image and very little R and L image That speaker radiates a portion of a spherical segment by using a series of concentric rings of electrostatic drivers.
That speaker produces the signal much closer to one point in space and time as well so it can reproduce a square wave over some bandwidth.
About 15 years ago I started working on a Horn based speaker system for commercial sound which also does this but a different way.
The current versions do not exhibit phase shift due to crossovers, most of them appear to be a single driver acoustically.
In commercial sound, all of the problems are greater, one needs much more acoustic power, that means even more sources which can interfere.
At the size of a football stadium, the sound quality is limited to what the large line arrays can produce most of which radiate a very complex interference pattern which is very audible if the wind blows.
In a way, this is where the difference between an interference pattern and a simple source can be very audible. With a point source, the spectral balance does not change with distance, normally with an array; one has a frequency response which changes continuously with distance and position right to left.
Here are a couple videos of “giant horns”, that sort of demonstrates what a large real single source sounds like. Try them with headphones a few are HD.
The first J1 cabinet (two years ago), this cabinet has 3 2way compression drivers, 6 6.5 inch horn loaded mid drivers and 6, horn loaded 18 inch woofers.
http://www.youtube.com/watch?v=pk54IFD4znw
The first outdoor demo of the J3 cabinet in December.
A fellow who attended took the video. One can set the volume to scale at 1:30 when the operator walks up to the guy next to the camera. As these are for large scale hifi sound, at about 2:30 he pans out to the field where the people were.
This cabinet has 4 two way compression drivers, 8, 6.5 inch horn loaded mid drivers and 6, 15 inch horn loaded woofers radiating as one acoustic source, all on one 60 by 40 degree horn.
http://www.youtube.com/watch?v=5MOG_sPejGA&feature=related
Sometimes even the players and management comment on the difference.
http://www.facebook.com/photo.php?fbid=298426253526848&set=a.185068854862589.34485.126113687424773&type=1&theater
Best,
Tom Danley
Danley Sound Labs
http://www.facebook.com/DanleySoundLabs?ref=ts
http://www.danleysoundlabs.com/
In a stereo pair in the test above(sans significant reflected sound) these speaker produce a strong mono phantom image and very little R and L image That speaker radiates a portion of a spherical segment by using a series of concentric rings of electrostatic drivers.
Are you, per chance, referring to Peter Walker's Quad ESL-63's????
"Are you, per chance, referring to Peter Walker's Quad ESL-63's????"
Yes, my old boss asked me to "fix" his (removing the spark gaps which protected the speaker). I had built electrostatic speakers before so he asked me to try. I had never seen a speaker like that though (concentric rings) and that was (a dozen or more years later) the inspiration for the full range Unity and then Synergy horns (some of which are in the videos).
Here, instead of rings of full range sources, the hf, then mid, then lows are produced sequentially within the horn, at dimensions appropriate for the frequency range such that the sum appears to been from a single driver.
Best,
Tom
Personal Note to cids and coytee: Having admired Kloss' work for many years and corresponded with him years ago I suspect his giant multiway horn system sounds very good indeed. There are many paths to good sound. I wouldn't bet against an old pro like Kloss without extremely good reason.
I believe attention to time and phase alignment is important in all speaker systems which divide the spectrun among different drivers and especially so with large horn systems because of the large distances between the acoustic centers of the various drivers. Some inmates will consider me an apostate heretic, but I use digital a sound processor/crossover, the DEQX HDP-3, to correct the time and phase relationships of my DIY horn system. The physical distance separating my corner horn sub's drivers and the Oris 150s they cross over to is well over sixteen feet, but with the DEQX I can correct that to a time difference as though they were less than 1/8" apart. I know there are trade-offs, but when time and phase alignment alignment is corrected digitally that frees you up to place the various horns (within limits) for the cleanest wave launch and least destructive room interaction.
Just because many of these horn systems are large and thus difficult to integrate due to shear siZe. How would you know from a picture if the networks not optimized for time and phase? You dont thus I disagree since most I know running such are well aware of the weakness of integration and alignment of horns and due address it some how. So the answer to your question is yes integration alignment in time and phase correct is worth pursuing. Also keep in mind unless point source time alignments only in one small spot hopefully where your going to sit. Many forget this.
'How would you know from a picture if the networks not optimized for time and phase?'
The owners mentioned how their system were set-up. :)
I believed a lot of those systems ignoring time alignment issue simply because they cannot hear it. :)
I'm not technical enough to know if mine are as mismatched as some of these or perhaps even more since my bass unit is folded.
That said, I originally attributed the unity of their sound to them being signal aligned. I was essentially bi*** slapped and told it was not primarily because of that but rather because I was listening to 2-way horns from previous 3-way horns.
I also use an active crossover with delay.
I was told to go to my active & change the delay to shove the two horns into a large disparity of misalignment. I ended up doing this.
When I went to the extremes, it was literally an echo so it was a fun experiment. That said, there was a range (and I have no recollection as to the values) where I personally could not tell much difference in the sound BUT, the important thing to me is, the coherency of them was "there" throughout most of the delay changing process. (until I got stupid with the numbers and could hear double-taps)
What that told me was in a practical sense, it's FAR more important to lose a crossover point if at all possible. In my system, adjusting for delay is more for fine tuning than a real fix for anything. (I use 2271 Us).
Personally, I tend to cringe when I see these impressive looking 3-4-5 way mega-horn systems as I have absolutely proved (to myself) that the more crossover points you can lose, the inherently better sound you will have. (presuming good horns to begin with)
Expecting a single driver to cover, satisfactorily, the entire human earing range is pure pure wishful thinking !
On the other hand, Multi-way Giant Horn Speakers can attain blissfulness but only through dealing with numerous parameters ; the tuffest ones being, undoubtedly : phasing & spatial distribution between the various drivers.
I have made the same trip as you from 3-ways to 2-ways (with bi-amping) and
my ears tell me every day that the move was fully justified.
Hornloco
" I tend to cringe when I see these impressive looking 3-4-5 way mega-horn systems as I have absolutely proved (to myself) that the more crossover points you can lose, the inherently better sound you will have. "
This is where time alignement comes in: If you have a perfectly time aligned, (read: in phase) sound system it doesnt matter how many drivers are playing or how many Crossover points you have, it will sound good.
When a driver is out of phase or mis aligned time wise, it makes a mess out of everything, when that happens it is wise to reduce the amount of drivers to minimize the clutter. But dont blame misalignment shortcomings on multiway speakers. The inability of someone to set a system correctly doesnt mean that the drivers are bad or the system is wrong!
Actually when perfectly time aligned, having a couple or more drivers playing the same frequency makes the sound fuller, more detailed, more realistic and much more interesting.
I have taken my currently signal aligned speakers and pushed the delay backwards and forward. I could tell a difference yes....BUT.... a MORE significant improvement in cohesiveness was made by dropping a crossover point.
Isn't one issue of having two drivers overlap, comb filtering? If you can blend those two drivers into a single driver, won't that by definition, clean up more problems and not cause any? (presuming said driver is capable of the range when blended)
Again... anyone can hear, like, dislike what they want. Matters not to me. As I stated above, I've already proved it to my ears by doing the experiments and actually hearing the differences.
Anyone elses milage might vary of course.
:)
'Personally, I tend to cringe when I see these impressive looking 3-4-5 way mega-horn systems as I have absolutely proved (to myself) that the more crossover points you can lose, the inherently better sound you will have. (presuming good horns to begin with)'
Yes, I have the same feeling.
Thanks for sharing your experiences.
On multiple crossover points with horn drivers. For horns by nature are narrow bandwidth. And any horn system using all front horns will be 4-5 way its just how it goes. Sure one can use a full-range driver in large horn bell toss a ported cab under or EQ comp mid and say done. But to me horns sound best with compression drivers. And the best matching bass to a horn system is a horn system. If you compromise extension you can pull off 2-3 way. But if you want truly full range front horns. You end up with horn tweeter, mid, mid bass, bass and sub bass all horns. And these will not be small.
Edits: 03/07/12 03/07/12
You can do first order passive networks with no EQ with the right multiple horn combo.......been there done that (5-way for me), or you use a digital Xover and EQ the heck out of the right two-way and get great sound also (done that too).
Are you familar with the Klipsch Jubilee?
2-way folded bass horn. Stands about 5 1/2' tall, maybe 2 1/2' wide and 2' deep ("V" shaped to fit into a corner like a Klipschorn)
Won't go to the deepest depths, no arguments on that however, for general home listening, will not get its tail kicked by too many speakers out there (if any)
Two 12" compression drivers in the bass bin and in stock form, a 2" compression driver for the mid/hf horn. Some have opted for a TAD-4002 driver to replace the stock driver. A nice upgrade in sound.
"Is time alignment in horn speakers important at all ?"
Depends on who you talk to, some think it is and some think it isn't. Some think it is up to a certain point. I've heard some excellent systems that disregarded the issue and I've heard some bad systems that paid great attention to it. And vice-versa.
.
Then again I see a lot of them that are, and a lot may be electronically aligned with time delay to the smaller horns.
Yes, I believe it is important.
.
i have posted a test signal on my web site. it is after the schematic , click on the little triangles.
http://web.mac.com/donaldpatten
Yes you can hear it. But does it bother you.
Don
:)
I believed this fellow is using some kind of digital time delay unit.
Edits: 03/06/12 03/06/12 03/06/12
Hello cids,
The owner of that system was the importer of BSS here in France. From memory he used 2 or 3 x FDS388 Omnidrive in that system.
Best regards from Paris, France
Jean-Michel Le Cléac'h
'The owner of that system was the importer of BSS here in France. From memory he used 2 or 3 x FDS388 Omnidrive in that system.'What is FDS388 ? Is it an electronic x-over + digital delay unit ?
Edits: 03/07/12
It's a digital loudspeaker management system, very much like the Behringer DCX2496 only better.
And read about time and phase alignment, which seems to address these things in his novel Synergy design. Read up all you can about it, I think, IMHO, Tom's approach is the "wave" of the future in High Efficiency speakers.
Jeff Medwin
~!
The Mind has No Firewall~ U.S. Army War College.
Just to make it all seem more impossible. I want my ht system to be surround. I cant watch surround movies on my stereo because the vocals are soft like they should be coming out of the center Chanel. Perhaps I am missing something simple. But it seems my Samsung blue ray player or my ps3 doesn't seem to mix them down properly. So I may as well go to a digital crossover if I am going to use a surround processor. Integrating a center Chanel should be a head ache but considering multichannel is the future shouldn't all of us be be building center Chanel's? Just think of all the cool single speakers, drivers, tubes, caps you have you can finally get to use.
I hope this mini dsp is decent as its about my price point. I hope they do a surround processor soon.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: