|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
68.212.139.159
It would seem ideal to have a speaker measure flat within a fairly large range (100 to 10khz or better). Assume on axis in free space measuring. But when I look at graphs, either posted here or at speaker web sites, the graphs are nowhere near this.So what is reasonable to expect? I mean if I can put something together and have it stay within 3dB should I consider myself very lucky or a dismal failure? How good is good enough and how bad is too bad? I am looking for general guide lines.
Which would be more desirable, a speaker that was very flat on axis but not so great off axis or a speaker that wasn't as flat on axis but nearly the same off axis?
And last but not least, assuming I can get something to measure well in free space, what is a tolerable in-room response? I am sure many of us can't afford the luxury of a well treated, dedicated listening room. Should I be flipping out over a 6dB (how about 12dB?) peak or dip? How bad before one should consider a digital EQ system as "required"?
Russ
P.S. Is it smarter to intentionally have a non-flat measuring speaker (in free air) to arrive at a better in-room sounding speaker? Perhaps rising low bass and falling high treble?
Follow Ups:
I built speakers as a hobby for many years before getting serious about "what really mattered", and putting in some time at the local university library. Cutting to the chase, the conclusion I drew was that the reverberant field matters much more than I had previously appreciated. Indeed, it now seems to me that 95% of the speakers on the market are designed without any serious attempt to get the reverberant field right.Rather than saying this or that frequency response matters most (and unless you have a polar map, each is just a snapshot as AJinFLA says), I'm going to break it down over time: You want a good clean smooth first arrival sound; then you want minimal early reflections and no strong, distinct ones; and finally you want a lot of diffuse, late-arriving reflections that pretty much have the same spectral balance as the first arrival sound after being modified by the room's acoustics. This can be accomplished in several different ways, which is beyond the scope of this post.
The reason I break it down into three time zones - first arrival, early reflections, and late reflections - is this corresponds well with how the ears process sound.
If you wanted to shoot me an e-mail along with your phone number, I'd be happy to talk with you about DIY speaker topics. I'm no expert by any stretch, but can sure talk a good game. ;-)
Finally, let me mention that the absolute best deal in all of high end audio is Madisound's crossover design service. If you buy the drivers from them, for a very very reasonable fee they'll design you a better crossover than you could yourself unless you spend a lot of money on test equipment and a lot of time (weeks perhaps) fine-tuning the design. It's not mentioned on their website but is in their catalog.
Best of luck to you,
Work blocks it. My email is Rbrunt(at)mercymiami(dot)org. Cell phone is 305-370-2226. Best time for a call is 4-8pm EST if that works for you (anytime on weekends is fine).Thanks for your kind effort and I'll try to keep you from regretting it:)
Since I live mins from madisound, I can drop bye, vist my friends, pick through the drivers,see all the new stuff things that never make the website,find out whats coming to market before others , test whatever I need and pick there brains.Plus I got a cool MJ 1 sound system the MJ stands for Michael Jackson and the NXT panels have great big pictures of the beast on them[MJ].I know you want it maybe we can work out a trade;) I have been a madisound customer for over 20 years now darn Iam old...Larry had all his hair and looked like a biker when I 1st started to shop.
Hey Kloss,I first shopped at Madisound when they were downtown, near the library, a few blocks off the 'Square'. My favorite Madisound product from those days (early 70's) was a folded transmission line using an 8-inch Philips woofer and 1 inch dome tweeter. Not particularly efficient, but relatively smooth and very wide range - nothing commercial touched them at anywhere near their price (about $150/pr as I recall). They handled mainly Philips and Audax. Larry had shoulder length hair. I left the Madison area about 20 years ago.
Is this the design service of Madisound that you're talking about?
Thanks, Ben! I looked for it but couldn't find it on their website.
No problem.
Is this a passive set-up of active of a combination of both?? I here about changes, it's real easy to interpret a change and get why wrong without a baseline of some type. An RTA will certainly help you. Passive crossovers are not easiest things to get right so don't get discouraged if it takes you a couple of trys. I like to run my set-ups reasonable flat on axis. It's good if you power response tracks your on axis. If it doesn't any EQ you use to correct the in room response, sum of on axis and off, will only mess up your on axis response. As far as +/- 3 that's fine even 6 can be fine depends on where those peaks and dips are and how wide they are octave to octave.
Your speaker radiates power in a 3D sound field with many FR's.
An axial measurement at distance X is only a snapshot.
Read ALL the Harman papers http://www.harman.com/about_harman/technology_leadership.aspx (Toole's handiwork), including the link below.
You are most certainly on the right track reading the Linkwitz website. It's enormous and contains a treasure trove of information.
Other must reads for DIYers (like myself) are: http://musicanddesign.com/ and http://www.gedlee.com/audio.htm
Read. Learn. Measure/Build. Enjoy..cheers,
AJ
The threshold for disproving something is higher than the threshold for saying it, which is a recipe for the accumulation of bullshit - Softky
Hi Russ,"Which would be more desirable, a speaker that was very flat on axis but not so great off axis or a speaker that wasn't as flat on axis but nearly the same off axis?
And last but not least, assuming I can get something to measure well in free space, what is a tolerable in-room response? I am sure many of us can't afford the luxury of a well treated, dedicated listening room. Should I be flipping out over a 6dB (how about 12dB?) peak or dip? How bad before one should consider a digital EQ system as "required"?"You've asked some interesting questions. Before trying to answer your questions, I'd suggest that you read this writeup (perhaps you already did) about Linkwitz's experience with adding a reari-firing tweeter and his excitement about the "wall effect."
http://www.linkwitzlab.com/orion++.htm
His experience somewhat confirms or re-enforces the suspecion I have had for a long time that room reflections play a critical part in our perception of soundstage, imaging, and "realism." I derived from my understanding of Linkwitz's experience these propositions:
(1) Neither flat on-axis but non-flat off-axis nor non-flat on-axis but uniform off-axis is good enough,
(2) traditional room treatment might be ineffective or even counterproductive.By the way, I put together a pair of speakers that are a simplified variation on the theme of his PLUTO design, and they sound quite interesting.
Regards,
Kurt
suitable drivers for open baffle bass. Word has it you are da man!Now I can't consider 8 drivers:) But I might could make room for a pair per side. What do you like the looks of? For my skill level it would be better to pay more and have a driver that required less from the crossover/eq network. With that in mind the augie from hawthorne audio has caught my eye. It looks like it might be better for me than the dayton IB's. I need the woofers to run up to 250 hertz and don't need 20 hertz performance.
Russ
P.S. The linkwitzlab site has a lot to digest doesn't it? Slowly it is sinking in but speakers are multi-discipline and a lot tougher than I first appreciated.
Hi Russ,Not sure I am "da man." I just happened to have slapped together the biggest OB bass in the suburb of Chicago at one time. Now that I've moved to CA and dismantled the OB bass in connection with the move, I've lost my bragging rights. Nevertheless, I'm happy to discuss about OB.
I've not used or heard the Augie, so I cannot comment on that particular driver. Nevertheless, genearlly I'd rather use a lower-Q bass driver on OB, instead of a driver with a high Q. You can always use an EQ (even a cheap analog 32-band prosound EQ will work) to get the bass to your satisfaction. That's what I did with the eight-18"-woofer OB bass. The 18" drivers I used have a Qts that is more suitable for a vented box. A little bit of EQ (I used a digital Behringer EQ), no more than +6dB max, was enough to get good bass down to 30Hz. The use of an EQ significantly increases the flexibility in the selection of woofers for OB, allowing one to put more emphasis on paraments such as Fs, Xmax, and distortion in choosing the drivers.
A high-Q driver allows the use of a narrower baffle, but it depends on the particular baffle width you want to use and can be too much of a good thing. In this regards,if you read Dick Olsher's writeup about the Feastrex with Augie project, http://www.enjoythemusic.com/magazine/equipment/0107/diy_loudspeaker_project.htm
you'll see that he used a parallel resistor to tame the resonance bump of the Augie. He speculated that the right range of Qts for OB woofers might be 0.6-0.7 (this should be baffle-width dependent). The Q of the Dayton IB woofer is in that range. Thus, you may not need any EQ if you put the Dayton driver on that baffle. Another important parameter is the Xmax. The Dayton IB driver has a Xmax that is twice that of the Augie. For these reasons, unless someone can show me that the Augie has significantly lower distortion figures, I'd pick the Dayton IB woofer over the Augie for my OB projects.As to the question of which woofer is good for 250Hz, I think that they both should have no problem going that high. The dip in the frequency response of the Dayton driver around 800Hz should not be a problem, especially if you use a 4th-order crossover.
Before I left Chicago, I was playing with the Fostex FE206esr drivers on OB with two 15” woofers on each 3’-wide baffle. I got the woofers on clearance sale for $25 each. They have a Qts slightly lower than 0.4. I used the Behringer CX2300 active crossover and a JVC digital receiver to drive the system. If I recall correctly, the frequency of the crossover was set around 250Hz. Those cheap woofers had no problem at that crossover frequency. The OB bass sounded very satisfactory even without any EQ. The main problem I had with that system was that the FE206esr sounded too bright even when listened to off-axis. Nevertheless, the system sounded very impressive in terms of dynamics and impact. I had a good time playing the Planet Drums CD through those speakers. To my ears, they were as dynamic or better, and with deeper and better bass, than some of the horn speakers I had owned.
Overall, I think most decent woofers can be used on OB, with a bit EQ if necessary. The Qts should not be a dominating factor in the driver selection, unless it is ridiculously low (like those drivers requiring horn-loading) or high (like those cheap Goldwood woofers).
Best,
Head on over to the 18sound website and see if something looks promising (uh...please:). The stuff is impressive looking in person and I can get a nice price so if something fits the bill?? I ruled them out because I thought I needed higher Qts and lower Fs.I just picked up a tdm crossover to fart around with so it's woofer decision time:)
Russ
Hi Russ,I took a look at the specs of the 18Sound woofers. Those woofers look nice, and I especially like those with neodymium magnets. It is clear that they are designed with prosound applications in mind, with emphasis on efficiency. Of those drivers, I'd pick the 15NLW9500, as its Fs is in the 30's, which is reasonably low, and its Qts is 0.32, higher than most other woofers offered by 18Sound. It looks pretty too. I won't mind having a pair in my collection if the price is right. Another woofer that looks interesting is the 18w1300, for the same reasons except perhaps the look. I'd be reluctant to use woofers with a significantly lower Qts, since it might require a heroic amount of EQ to bring the bass level back up.
Keep us posted on the progress of your OB project.
Best,
Kurt
Thanks Kurt that was nice of you to take the time. They have a lot of models to wade through!It looks like those drivers are about the same cost. I found myself liking the 18" better. But I did notice the 15" neo has a much lower Vas and some think that is important for open baffle bass. But some think the Le on the Dayton IB's is a problem. I don't confess to knowing why this should be an issue.
I am just trying to be careful and make a wise decision. All things equal, if I can buy something that would work later on in a bass reflex box too I’d prefer that. As you can tell my preference are for pro audio drivers.
I’ll have to look at the Pluto again.
Hi Kurt,Thanks for your woofer insights! I've been looking at specs for my own open baffle project, and my simulation showed the same thing--the Dayton IB woofer going deeper and flatter, baffle effects and all. Add in the lower price, and they look pretty appealing. The Augie has higher efficiency, so might be better for use with a passive XO.
I'm pondering a OB FR driver (Fostex FX120), actively XO ed to dual OB bass drivers (Dayton IB Woofers) somewhere around 200HZ. Given the conflicting goals of unburdening the FR (XO frequency up) and not crossing over in the vocal range (XO frequency down), I'm also thinking a high-order crossover might be just the ticket. I'm thinking the Behringer digital XO would let me play with that.
Another bonus is removing the bass requirement from the FR allows me to pick a smaller whizzerless driver. I'm just guessing, but I'm thinking the sharp changes in response curves for whizzer designs represent resonances and interference with the whizzer, which might represent distortion.
So thanks for your insights, and any more that you or others might have.
Best,
Hi George,In my OB projects, I always tried to keep the crossover frequency below 200Hz or so, if the FR or wide-band driver could go that low on the given baffle. It is difficult to set the crossover point much lower than that, because most FR or WB drivers on a baffle of a reasonable size would start to roll off around there.
Have you bought the FX120 drivers yet? I have a pair of FX120's and I like them. Another small wizzerless FR driver for OB application that I can recommend is the TangBand W4-1320SB (the "bamboo-fiber-cone" driver). Just this morning I was listening to a pair on small OB's augmented in the bass with a subwoofer, and I was again impressed by how good the sound was from those little inexpensive drivers. Compared to the FX120, the TB driver has crispier highs. I'm going to try the FX120's on the same baffles, and I'll let you know how they compare to the TB bamboo drivers in that configuration.
I have to say, however, that my thinking about OB has changed somewhat, now that I've tried a clone of Linkwitz's PLUTO speakers. As good as the OB sound of the TB bamboo drivers is in terms of soundstage and imaging, which is about as good as what I've gotten from all the FR drivers I've tried, it is no match to the sound of the PLUTO-clone speakers in those respects. Even my wife, who is not an audiophile, said that she preferred the PLUTO-clone speakers to the TangBand OB speakers (both were put in the same locations and used the same subwoofer to add bass), because she could clearly tell where the instruments were playing when listening to the PLUTO-clone, while the TB OB speakers did not give such a clear picture.
I am not saying that I’m giving up on OB or that omni-directional is the way to go. Nevertheless, I now think that it takes more than putting a good FR driver on a baffle to get great sound. I’ve ordered some new drivers so that I can do some experiments to verify one idea I have.
In this regards, if you have the time, I’d encourage you to also make clones of the PLUTO speakers. They will serve as a useful benchmark/reference for gauging all other speakers in terms of soundstage and imaging.
Best,
Hi Kurt,I'd like to continue our discussion, if you have the time. I feel bad about hijacking this thread, so if you have an idea of where this would be more appropriate, I'd be glad to move there.
"I have to say, however, that my thinking about OB has changed somewhat, now that I've tried a clone of Linkwitz's PLUTO speakers. As good as the OB sound of the TB bamboo drivers is in terms of soundstage and imaging, which is about as good as what I've gotten from all the FR drivers I've tried, it is no match to the sound of the PLUTO-clone speakers in those respects. Even my wife, who is not an audiophile, said that she preferred the PLUTO-clone speakers to the TangBand OB speakers (both were put in the same locations and used the same subwoofer to add bass), because she could clearly tell where the instruments were playing when listening to the PLUTO-clone, while the TB OB speakers did not give such a clear picture."
I find this very interesting, and have been puzzling over it. Particularly because Seigfried Linkwitz gushed over the improvement in his Orions when he added a rear-firing tweeter, changing them from conventional front-radiating to dipole in the treble.
One thought--how close to you have your speakers to the walls? SL talks about the necessity of having the speakers at least 3' away, so that reflections are delayed enough that the brain processes them as ambience, and not smeared imaging.
If your speakers are close to the walls, that might explain why you find a front-radiating speaker to have better imaging than an open baffle. I'm guessing they are not that close, and I'll have to ponder more. That's not a bad thing.
I'll also be curious to learn your comparisioin of the Fostex with the TB. I'm thinking that treble peak on the TB might be a resonance, which pushes me towards the Fostexes, since I can always boost treble another way.
Thanks, and have fun,
George
Hi George,I doubt that Russ would mind us having a discussion about OB off his thread.
The PLUTO-clones and the OB speakers are both placed about 5.5 feet from the backwall. Neither system is listenable (to my ears) if pushed too close to the backwall. Even at 3 feet from the wall they did not produce a soundstage that is anywhere close to what I used to get in the 1400 SQFT basement of my old house in Illinois. By the way, the PLUTO is supposed to be omni-directional up to 3000Hz, even though the tweeter is front-firing. I've not measured how high the back radiation of the OB speaker goes.
One thing I noticed about the sound from the PLUTO-clones is that there is a bit "hollowness" or "softness" in human voices compared to the sound of OB. Not sure whether this is due to the omni-directionality (perhaps too much reflected sound in the midrange) or mainly a frequency response issue. I got a similar impression, however, when I listened to my "ball" speakers that have nine 4" FR drivers distributed on each multi-faceted enclosure. More investigation is needed.
I compared the TB bamboo drivers with the FX120's using the same baffles yesterday. The FX120 was smooth and full-sounding, but compared with the TB it appeared to lack sparkles in the high and not have as good clarity. The additional energy in the highs given by the TB driver also seemed to produce a better delineated soundstage that is easier to "see into." Perhaps some people would prefer the fuller sound of the FX120, but to my ears the FX120's in the given room and on the given baffles were not as exciting to listen to as the TB drivers. Thus, if I were to pick one, I'd go for the TB driver. I don't think you'd be seriously disappointed by the TB drivers on OB.
Best,
Hi Kurt,It's good to "hear" of your satisfaction with the Tang Band W4-1320SB . I'm still a little leary of that sharp little peak between 15K and 20K, since I suspect sharp changes in frequency response represent non-linear behavior, like cone breakup, interference, or resonance. Do you think that's a reasonable suspicion in general?
Of course I am splitting hair, since the peak is so small, but since I'm planning on equalizing the output, I can always give the FX120 a little more "zing" if need be. These are all just musings on my part, and your experiences are very helpful.
I see you're using passive (speaker-level)1st order crossovers in your Pluto clones, do you use that with your OB setups? I think the Plutos have a XO somewhere in the midrange, which I'd like to avoid. Do you have a guesstimate of where yours is. I'm leaning towards a digital crossover to experiment with, and because I want to try line-level crossovers. I remember hearing the difference between active and passive crossovers with the old Magneplanar Tympani s, and the improvement was dramatic and beautiful.
This is fascinating and fun--there are so many variables, even when trying for a relatively simple solution.
Thanks again for taking the time to share your thoughts and experiences--fun and helpful.
Best,
Hi George,I think it is reasonable to suspect that a sharp change in the response of a mechanica/electrical/acoustic system means that it is going through a resonance or mode transition of some sort.
I don't know how "bad" the peak in the response of the TB driver really is. It doesn't look too bad compared to the peaks and valleys in the repsonse curves of many famous FR drivers, but I agree that it would nicer if it could be avoided.
As to cone break-up modes and resonances, are you thinking that a 5" paper cone like the FX120 will maintain pistonic motion all the way to 20KHz? There are break-up modes, and the question is how well the resonances are controlled. In this regard, I wonder how the response curve of FX120 will look like without smoothing.
I used first-order speaker-level crossover in the PLUTO clones because it was easy to put the speakers together quickly that way. Nevertheless, the first-order crossover was selected also because I thought that it should be relatively benign in terms of phase rotation as compared to a 4th-order active. I was shooting for a crossover frequency around 1200Hz. Even though the crossover was in the midrange, I was hoping that the transition was not be as objectionable as it would be with a 4th-order crossover.
With my OB projects, I normally do not use any speaker-level high-pass filter (i.e. a cap) on the FR driver. In some projects, I run the FR driver wide open, and add bass using a powered subwoofer with a 2nd-order low-pass. In other projects that use separate woofers, I use a 4th order active crossover and biamp the FR and the woofer. When I played with the Audax PR170MO with a tweeter, I did use a cap on the tweeter.
You are right that there are so many variables in designing a speaker system, and the problem is that the effects of some variables are not well understood, let alone well-controlled.
Best,
Hey Kurt,Thanks again for the thoughtful reply. I'm enjoying the back and forth, and appreciate you taking the time with a newbie. Your reasoning seems valid to me--my sympathies. But--doubtless partially due to my inexperience--I want to try a slightly different variation.
In particular, I take your point that any paper cone speaker (or any speaker, I imagine) will have breakup modes, and that one brand's frequency response curve might be smoother than another brand's simply due to measurement techniques.
Granted that, my preference is to err towards smooth rather than crisp. Add to that I will be driving my FR with a SET amp, which has a relatively high output impedence, aka a very low damping factor. Paul Joppa asserts that this causes an SET's output to decrease less than a high-damping amp's when a FR driver's load impedence rises in the bass and treble--boosting them. So I'm thinking the lower-treble speaker might work better with an SET. Any thoughts?
Of course I had my mind set on the FX120, and after a while I might bring myself around. Half the price is a nice incentive, and I will have equalization available. Any thoughts on the 3" Jordans? The specs and buzz look good.
I'm also leaning toward an active crossover for three reasons--I want to hear if it sounds better than a passive, I want ease of experimentation, and I want to unburden my FR's SET amp.
Regarding that--is your dislike of fourth-order crossovers due to their inherent, ideal qualities, or the fact that implementing them with analog/passive components mucks up the signal? My understanding is very shallow here, but I'm under the impression that their theoretical qualities are excellent.
So many thoughts, so many variables. My strategy is to try to keep an open mind. And, as Einstein said "Things should be made as simple as possible, but no simpler."
Thanks again for your experiences and thoughts.
Hi George,It's my pleasure to chat with you about speakers. I think most others at this forum have already forgotten about this thread so perhaps you don't have to keep feeling guilty about hijacking. ;-)
Since you had your mind set on the FX120, go ahead and try it. I know that feeling of wanting a particular toy. ;-)
As to SET, you can treat the output impedance of the amp and the speaker impedance as forming a voltage divider. Thus, when the speaker impedance gets higher either towards the low frequency resonance or in the high range, the speaker gets more of the amp output. For most drivers, this will have the effect of boosting both the bass and the treble. It's like pushing in the "Loud" button on a receiver. In other words, the amp becomes a tone-control, and the effect is speaker dependant. In the case of FX120, I don't think that is an effective way to fix the HF rolloff, as the impedance rise is too gradual to compensate for the rolloff. You'd end up with a rising response in the mid and most of the high and then the rolloff before reaching 20KHz. Perhaps you'd like such a tonal balance, but it is not flat, which I think is what you desire theoretically.
I don't dislike 4th-order crossovers per se. Nevertheless, I am not certain about how the increased phase distortion associated with a high-order crossover might impact the sound so I'd rather keep the crossover simple. Linkwitz, however, thought that the phase distortion of a LR4 crossover is not audible at either 100Hz or 1400Hz. http://www.linkwitzlab.com/phs-dist.htm
Well, I guess we have extended this thread long enough. Time for you to build something! Let us know how the FX120 OB project turns out.
Best,
Hey Kurt,Build something??!!?? Surely you jest! ;)
Actually, I'm building the preamp and SET FR amp now, but won't put the speakers together until i sojorn to MI this spring--I plan to use them to keep me relatively sane in the hovel I am remodeling.
Besides, I find the research delicious foreplay, where anything is possible, before dealing with cold, cruel reality.
Thanks for the Linkwitz link--his site is fascinating. It's interesting that he found phase distortion in his analog 4th-order crossover inaudible. I'm planning on using a digital XO, which I believe will have none of the phase distortion, though naturally it's own compromises. And I'm hoping the sharp slopes will minimize interference between the drivers. The prospect of combining new and old tech in a new way excites me.
The digital XO will also provide equalization, another one of those varibles.
I'll certainly report on my experiences, but that'll be a good while from now.
Feel free to comment if you like, and thanks and good luck!
Best,
Hi Kurt,Thanks for taking the time to write your helpful and informative reply. You've given me some things to ponder. I'd like to go open baffle for ease of construction and experimentation and reduced room-interaction, but I do take your suggestion seriously. In the meantime, I hope I can impose on you for a quick question or two:
What drivers does Linkwitz use in his Pluto, and/or you in your Pluto clone? So far I haven't seen them mentioned--if that's proprietary information, I understand.
No, I haven't bought the FX120 yet, and, not having the speaker to audition, am certainly open to suggestions. In saying the Tang Band W4-1320SB has crispier treble, I'm assuming that is better? Their frequency response chart certainly looks nicely smooth and level, and similar to the FX120. Half the price is nice too.
And could you point me to a description of your setup? I'd like to compare it to what I'm contemplating. Again, if my sieve-like memory has forgotten, please accept my apologies.
Thanks again for your very helpful replies.
Best,
Hi George:I got addicted to the OB approach years ago after putting a pair of Fostex FE83 on cardboard baffles and added a subwoofer for bass. I was floored by the spaciousness and richness of the sound from such a simple setup. I've had numerous OB projects ever since, using various FR or WB drivers, ranging from 69 cents each to over 2 thousand dollars a pair. They all sounded quite different, aside from the common attributes of OB sound. Ease of construction and experimentation is definitely a major factor that makes this approach so interesting to me. Nevertheless, even back then when I tried OB speakers for the first time, I suspected that the room-filling soundstage was due to, rather than in lieu of, room interaction.
Linkwitz's PLUTO uses a 5.25" Peerless woofer and a 2" wide-range driver by Aura. The drivers and construction of the speaker are described here: http://www.linkwitzlab.com/Pluto/construction.htm
When I read about this design, one thing that struck me was that the size of the 5.25" woofer was selected for bass extension so that the speakers could (with some heavy EQing) be produce some bass, but a smaller woofer might be better if omni-directionality is the goal. A smaller woofer in a smaller housing can be omni-directional to a much higher frequency (i.e., the baffle step transition happens at a higher frequency range) at the expense of bass extension. I thouhgt that pushing the baffle step transition up could allow a simpler crossover to be used, and the better omni-directionality might enhance the sound. So, instead of using a 5.25" woofer, I decided to use a 3" driver. Since I happened to have several Fostex FF85K drivers lying around, I decided to use them. Like the original PLUTO, my clones also use ABS plastic pipes for housing. The sealed enclosure for th FF85K is made of a 3" pipe coupling and a 8"-long piece of 3" pipe, with a top wooden plate with a routed hole for the driver and a wooden base plate. The Aura "tweeter" is in a sealed enclosure made of a 1.5" pipe coupling and a 4"-long piece of 1.5" pipe, and an end plate. The Aura driver is pushed into the opening of the 1.5" coupling with a strip of cardboard wrapped around it as a gasket. The tweeter housing is supported on a 1.5"-high piece of wood glued to the top plate of the woofer housing, such that the tweeter faces the listener while the woofer fires upward, just like the original PLUTO design. The crossover uses simple 1st-order elements, with a 15 uF cap in series with the tweeter and a 1.6mH coil in series with the woofer, both drivers being connected in the positive phase. I have to emphasis that the cap and coil values were calculated based on some guesstimates and were intended for a quick-and-dirty project, and the resulting sound could be far from optimal. No other components for level-matching or response shaping are used. Since a 3" driver in a sealed enclosure is not going to give any bass, I use a subwoofer (the same subwoofer I used years ago with the FE83) to fill in the bottom end.
As to the comparison between FX120 and the TB bamboo driver, the TB driver does seem to put out more energy in the highs. In this regard, the TB driver sounds more airy and detailed than most of the FR drivers I've tried. I am about to make new adaptor plates for mounting the FX120's on the baffles, and I'll provide my impression after comparing the two drivers on the same baffle. By the way, those baffle are nothing special. I built them using the material I found in my garage. They are 20" wide, about 35" tall, and have side panels about 8" at the top and 12" at the bottom. The side panels support the baffle and give it a bit tilt up towards the listener. The driver is mounted with its center at about 10" from the baffle top edge and offset from the baffle centerline by 2". I use the subwoofer to add bass.
Best,
RussYou're not really flipping out over 3 dB dips and peaks are you? If so, a bad case of Audiophile Nervosa is diagnosed. Start your treatment by reading J. Gordon Holt's article: "Space...the Final Frontier" from the 3/94 issue of Stereophile. You can almost hear JGH yelling "Iceberg ahead, it's shaped like a giant TV!". The S.S. Stereophile did hit the iceberg, but it did'nt sink. Giant chainsaws were procured from the hold, and big chunks of ice were cut off and stored in the leaking compartments. Ice floats you know. They shoveled more money into the boilers and off she sailed, a little lower in the waterline. Warning: Don't read the next paragraph here untill you finish the article.
Now let's take a look at the receiving end of an ideal pair of perfectly flat speakers: the human ear. Let's subject it to the scrutiny we would apply to a piece of audio equipt. Look at the freq. response: http://members.aol.com/sbench/fm.gif. Turn the graph upside down. At low levels the bass rolls off at 700 Hz at more than 10 dB per octave, with the highs down nearly 10 dB at 10K Hz. There's an 8 dB peak centering around 3500 Hz that increases to over 10 dB at higher levels. The device only has it's most linear region at 100 dB, and some experts will tell you that operating it at this level for extended periods will damage it. This begs the test question: can you fix all this with digital EQ on your ideal flat speakers? I dunno.
Paul
please pardon my behavior Rev. Paulif circumstances permitted maybe a was of money, a nice whore, bottles of booze, good cigars and a long visits to beaches might be better than squinting at RTA and graphs?
FreddyA man in your position should have it all. You can take a laptop with RTA ware to the beach, pack some H. Upman Especiales and a cooler full of Pilsner Urquel. If your companion lacks interest in the RTA graphs, try bribing her with the promise of a nice dinner later. Works for me!
Paul
P.S. Cool it with the Rev. stuff. There is rampant prejudice towards audio polygamy lurking.
Hi Paul:H Upman especiales!!! Brings back memories....as a kid my father used to take me hunting with the owner of that company...Nice guy with sun sensitive skin (discoid lupus). He was totally cleaned out by Fidel in Cuba. The military first took his 57 T bird then his Mercedes (me and the hunting dogs used to ride the backseat). My father ended up buying him a junkie car so they would not take it too. I am traveling now but have residence in Miami too.
ArtemiusPerhaps you can put a word in for me with the Upmans. If they would bring back the Director Royale (not seen for over 10 years), I would happily sit on a beach with Ann Coulter herself and keep my yap shut, with just that Royale in it. The best cigar ever! Just my opinion.
so easy enough to do the cuban cigars and the visit to the beach:) Still need the laptop with RTA and my tastes lean more towards Bass and Guinness.Perhaps I worded my question poorly? I am taking it for granted that most will measure and listen a number of times before they consider a speaker "finished". All I am trying to do is get a handle on how well the speakers should measure. From there, if someone wants to go into how to measure and what to use that would be great.
I have seen the measured stuff in the reviews and such. But I build tube amps and can easily better commercial amp specs simply because my goal isn't to get the most watts for the least bucks. I was hoping the same might be true with speakers. Namely that if I am not trying to get 20-20khz out of a 2-way with a 6" woofer I should be able to get a better looking graph and hopefully better sounding speakers.
I am worried about Berenek's law. Simply put, I don't trust my ears when it comes to speakers. I don't know what to listen for and how to relate the sound to changes that need to be made. So all I can do is lay my cards on the table and ask for advice of how to proceed.
If it matters I have the xt1086 with nd1090 (from 18sound) with audax pr170mo on an open baffle. Bass is, for the moment, 12" in a bass reflex (borrowed stuff so this will be changed). Crossover at 250 and 1,800. Room is all concrete and plaster, so call it lively.
RussIt helps to listen to a lot of stuff, and then you will begin to trust your ears more, after all they will be the final arbiter as Olson once said. I've been to quite a few hi-fi shows, and one thing that amazes me is how bad some of the stuff sounds. Sometimes things on the recommended component list too. You hear complaints about the lousy sound in the hotel rooms and even low voltage at the wall etc, but everybody's got these factors to deal with at the shows to one degree or another. Sometimes there's something obviously wrong with the electronics, and the demo guy has to brave it through. I always feel sorry for the guy. Sometimes the salesman is oblivious to how bad the system sounds and just proceeds to practice his craft anyway, with the Emperor's New Sound approach The measurements have to proceed to some aesthetic end, or they are just measurement for measurement's sake. Sometimes you can change something, and the result just sounds different, not necessarily better. In this case, measurements can help point you in the right direction. But notice the disagreement between JGH in the Stereophile article and John Atkinson. It came down to a point of aesthetic procedure. We can add JGH to list of American enterpreneurs who started a company, got burrned out and sold it, and came back into the business later. JGH now writes for the Absolute Sound. The best advice is to not get too hung up on the measurements, they're just a means to an end and not an end in themselves. This is a hard learned lesson from the THD amp wars, when the amps became so "perfect" that their flaws were considered inaudible. The imperfect human ear won in the end, but it looked like it was going the other way in the 80's. You should proceed by trying to identify what you don't like about your rig using measurments to focus in, and to then formulate a strategy for addressing this. Some solutions will work, many will not. It's a struggle, but fun too. There is a certain pleasure in a nice freq. response graph, but you won't be coming back to it aagain and again like your favorite piece of music.
Paul
P.S. Did you get the roof on the hospital? I coulda blue-tarped it right away ; )
Funny you should ask now:) They ended up torching down some stuff in a small section and recoating everything. We have to let it cure for 60 days and they put some granulated crap on it. The basic roof was sound, just hurricane scrubbed of gravel, so I guess it will be okay.With the speaker stuff...I guess I am just frustrated/overwhelmed. With an amp I can hear things and know what to do....a coupling cap here, a tube there. With the speakers....well I don't know how things interact yet. So it is not trusting my ears so much as not knowing what common speaker problems sound like and the solution.
But the bright news is I feel, and friends agree, that what I have managed so far is a lot better than the commercial stuff I could afford. But now I want to get a little more serious and incorporate measuring....and down the road modeling software. Economically it makes the most sense for me as I simply can't afford to buy a bunch of drivers (and especially crossover parts) and build a bunch of speakers.
But maybe that is flawed thinking and I'd be better served to simply follow some proven designs?
RussYou're on the correct path. The roof coulda turned out lots worse, I could tell you some Chicago roofing stories! And your audio path is illuminated by a dark lantern. A basic speaker test rig does not need to be expensive. John Sheerin recomended a PC RTA freeware program recently, and he uses the ubiquitous Panasonic mic capsule. I use the Mitey Mic variant of this test mic availble from audioxpress, along with the VU meter in a Teac R2R tape deck and some graph paper along with the warble test tones on the Stereophile Test Disc. Test the drivers up close, and compare these measurements with those at the listening position to get a handle on what the room is adding. Problem areas should be obvious. Just don't try to solve room issues with speaker solutions and vice versa. You're on the right path, and have a good handle on the situation with your amp experience. Kurt's your man for OB. He put together several awsome ones! One was a Fostex driver on a cardboard baffle with Kurt's own sub. This rig could have easily mopped the floor with a Spendor rig (over $1000 a pair with sub) I heard at an audiophile salon. His 8 X 18's and ribbons rig was amazing too. So don't forget, this is supposed to be fun!
ok - no more revs - I still think church of Olsonology is good vs the crutch of nothingology
FreddyThat vision of you on the beach lingered a bit in my mind. It drifted to a White Owl and a case of Lone Star (sorry), but here's the good part, I saw Daisy Mae Duke sit'n next to ya! The vision started oscillating between Cathy Bach and Jessica Simpson (could'nt tune in both) and to stop the oscillations I tried to focus on the Fletcher Munson curves from the Steve Bench link, but there she was again! Take a look at it, at 120 dB Daisy is laying flat on the beach, at 0 dB down she's really arching her back. Her feet are cut off in the graph, but check the link for comparison which is also footless. It's a great numonic to remember the frequencies of the F.M. curves.
Forward for science!
Paul
Some of my best friends are nothingologists!
those are interestinghere's a still that looks right for the beach
FreddyThe bass cuts off on the left chanel, but nice stereo pair midrange, nice high extension too. I'd like to see Aki polishing some Japanese heavy iron, an Ongoku say, to see how she stacks up against Jessica for Fletcher-Munson Girl of the Month.
Dynamical Analogous Circuits and beyond!
Paul - perhaps some might think those lower extremities "lean" and lower mids pinched & threadbare, but putting all into perspeotive, everything seems well balanced, no superfluous features, a voluminous & happy chest tone reproduction plus ultra-pretty technicolor top - - -might all this go with GOTO or Feastrex?would Olson approve?
I dunno FreddyThe static Aki representation makes it hard to establish parity with the dynamical format representing Jessica/Daisy. I see high compliance and low Q in the amount of energy being released. We need to establish Qms and Qes to put together a Qts on Aki. Invite her to FreddyiLab for tests!
My speaker building goals, measured at the listening position, and averaging two measurements spaced 8 inches apart (to represent right and left ears):Over 100Hz. you must use pink noise and 1/3 octave smoothing because sine waves without smoothing will reveal horrible 1/4 wavelength (combfilter) cancellations that look bad but are unlikely to be audible except in the upper bass. Averaging four measurements (microphone array) is likely to be more accurate than averaging two measurements.
Below 100Hz. you must use sine waves spaced less than 1/6 octave apart to excite standing waves. You can only get +0dB by using parametric EQ. Otherwise you'll have one or more peaks typically +3 to +6 dB. If you have a lot of flexibility to move your listening seat, you may find a location with no standing wave nulls worse than -6db -- but typically there is least one -6 to -12dB null under 100Hz.
,,
,
,
,
Richard BassNut Greene
Subjective Audiophile 2007
Richard, could I trouble you to give me some advice on what to buy? My budget is limited but I feel good measuring gear is worth the investment. But I really don't know what is adequate for the home hobby type builder.You know, something like go get two behringer ecm8000 mikes and behringer model#?? preamp, TrueRTA level? software, M-Audio model#?? USB sound card, and go buy a 1/3 octive test CD from ??......or whatever.
I need help:) I was using my signal generator. No wonder I was all over the place.
This post is made possible by the generous support of people like you and our sponsors: