Group buy
"As far as a "group buy", I'd only consider this if I got one payment and shipped to one address. I could do $200 each for 20 and $150 each for 100.___________Earl Geddes"
----------------------------------------------------------------------
Count me in for 2 channels in the group buy.
Chris Wait, Rochester NY, USA
"As far as a "group buy", I'd only consider this if I got one payment and shipped to one address. I could do $200 each for 20 and $150 each for 100.___________Earl Geddes"
----------------------------------------------------------------------
Count me in for 2 channels in the group buy.
Chris Wait, Rochester NY, USA
jzagaja said:
Not conical, surely. But yes the side walls do act as a sort of waveguide, but note Patricks comments as he is quite correct. Its the throat that is so sensitive. And , the wavefronts from the ribbons is flat (basically) which is the wrong shape to drive the walls correctly. Glass is also not very rigid. All in all not a very good idea - sorry.
Dr. Geddes,
What woud you recommend for the center channel in a 5:0 home theater that uses 4 full range speakers front and rear with your tweeter waveguide ?
My current center channel uses a DIY coaxial with a Lambda 12 and Seas 27TDFC that is 3-point mounted over the Lambda's phase plug. I think it improved voice location to have some beaming in a center channel.
What woud you recommend for the center channel in a 5:0 home theater that uses 4 full range speakers front and rear with your tweeter waveguide ?
My current center channel uses a DIY coaxial with a Lambda 12 and Seas 27TDFC that is 3-point mounted over the Lambda's phase plug. I think it improved voice location to have some beaming in a center channel.
Why Horn Guys Like Tubes
This has some ming-boggling ramifications - I thought I'd delve into this a little deeper since I've given it a bit of thought.
Back in 2005 I started considering the implications of this, after we discussed while you were demo-ing the Summas at the RMAF.
As I understand it, the fact that the perception of HOMs is level-dependent explains why conventional horn loudspeakers are so intolerable at high volumes. What this means is that the audibility of high order modes goes up dramatically as the volume level inches up. So even if the frequency response at 90db is similar to the response at 80db, the listeners perception of the speaker is dramatically worse at high volumes. This is all-but unique to horns; and it explains why they sound so different than conventional loudspeakers, particularly at high volumes. Ironically, this also explains why horn guys like tubes. The low output of tubes is a natural match for a horn that has HOMs; the lack of power puts a soft limit on the HOMs. (Feel free to comment on these impressions - I didn't realize the implications of this until our discussion a couple of years back.)
Has anyone here listened to the Avant Garde speakers? I believe they are a good subjective example of this effect. The Avant Garde speakers, played at a reasonable volume, are dynamic and almost "delicate." They are very good at revealing nuances in recordings. Unfortunately, as they get louder, I find that they become steely and painful. In other words, as they get louder, they sound a LOT like a traditional horn.
That was the thing that really blew me away about listening to the Summas - as the volume level became louder, their sonic signature remained the same. If I owned a pair, I would keep a DB meter handy, as you could do some serious damage to your hearing because of the levels they can achieve effortlessly.
:: PB ::
gedlee said:
This has some ming-boggling ramifications - I thought I'd delve into this a little deeper since I've given it a bit of thought.
Back in 2005 I started considering the implications of this, after we discussed while you were demo-ing the Summas at the RMAF.
As I understand it, the fact that the perception of HOMs is level-dependent explains why conventional horn loudspeakers are so intolerable at high volumes. What this means is that the audibility of high order modes goes up dramatically as the volume level inches up. So even if the frequency response at 90db is similar to the response at 80db, the listeners perception of the speaker is dramatically worse at high volumes. This is all-but unique to horns; and it explains why they sound so different than conventional loudspeakers, particularly at high volumes. Ironically, this also explains why horn guys like tubes. The low output of tubes is a natural match for a horn that has HOMs; the lack of power puts a soft limit on the HOMs. (Feel free to comment on these impressions - I didn't realize the implications of this until our discussion a couple of years back.)
Has anyone here listened to the Avant Garde speakers? I believe they are a good subjective example of this effect. The Avant Garde speakers, played at a reasonable volume, are dynamic and almost "delicate." They are very good at revealing nuances in recordings. Unfortunately, as they get louder, I find that they become steely and painful. In other words, as they get louder, they sound a LOT like a traditional horn.
That was the thing that really blew me away about listening to the Summas - as the volume level became louder, their sonic signature remained the same. If I owned a pair, I would keep a DB meter handy, as you could do some serious damage to your hearing because of the levels they can achieve effortlessly.
:: PB ::
Re: Why Horn Guys Like Tubes
Patrick
I guess I would agree, but there are other things about tube amps that have their attractions - not just their soft clipping.
We have now installed our speakers in a couple of discos in Bangkok. The result is unanimous that at the extreme volume levels there is simply nothing that compares with them. They can play at volumes which I find completely uncomfortable (and intllerable) and yet they remain "clean". It now appears, from sales orders, that if you don't have our speakers in your club in Bangkok, you won't be "in the game". They will become the defacto standard.
As I have always said there are lots of good speakers for low volume levels, but when I started the Summa design, I could not find any speaker which had good sound quality at theater like sound levels.
Patrick Bateman said:
Patrick
I guess I would agree, but there are other things about tube amps that have their attractions - not just their soft clipping.
We have now installed our speakers in a couple of discos in Bangkok. The result is unanimous that at the extreme volume levels there is simply nothing that compares with them. They can play at volumes which I find completely uncomfortable (and intllerable) and yet they remain "clean". It now appears, from sales orders, that if you don't have our speakers in your club in Bangkok, you won't be "in the game". They will become the defacto standard.
As I have always said there are lots of good speakers for low volume levels, but when I started the Summa design, I could not find any speaker which had good sound quality at theater like sound levels.
LineSource said:
I have always contended that the three front channels really have to be identical. Thats because things pan across them so often. If you look at most movie mixes, the center channel is the most important - most used. The surrounds can, and should be different, but the three main channels need to be identical.
This, of course, poses a real problem with positioning. One in which I have found that only a projection system onto an acoustically transparent screen can solve. I'd use this technique even if I only had a 30" screen.
It may interest some to know that I am going to set up a home theater system in a small appartment here in Bangkok - mine! So I'll document how I do this. A real challenge. I know that I'll use a projector - that much is certain. The screen - a bed sheet, of course!! I'm planning on using three ESP12's for the main channels. In Bangkok, the sun always goes down at 6 PM so there is no daylight savings to worry about (a big problem in the summer in Michigan - it stays light until about 10 PM).
gedlee said:
At the CES a few years back, there was a demo where the vendor had placed the five channels in a perfect circle in the demo room. This is much different than what you typically see in a home theater, where the three front speakers are all on the same horizontal plane.
So this vendor, by using a circle, had the centerl channel back much further than the left and right. And the difference was remarkable; I had no idea that the timing error introduced by having the center channel too close is SO OBNOXIOUS.
I'm using 15inch two-ways in my home theater, and have ditched the center channel for this reason. If I decide to add a center channel, I would certainly build a false wall* so that I could have every channel equidistant.
:: PB ::
* I use a DLP projector also.
I was requested to submit an impulse response of the Summa (ESP15) for someone to do a waterfall on. Here it is. This impulse may be a bit noisey owing to the way the data is taken. I usually average in the frequency domain which mostly negates the noise effects so, if used, the responses should be 1/6 or 1/3 octave averaged.
The sample rate was 50 kHz.
The sample rate was 50 kHz.
Russell Dawkins said:
At least good club sound is available somewhere! Too bad Bangkok is so far away.
Congratulations - sounds like a good beginning. ""
Hold on!!! The sound is still terrible!! But when they use ours it's NOT the speakers fault
Most clubs here burn the ears. At least in our installations you get to actually hear how bad the music is! I get the chance to listen to the systems with music of my choosing and settings so I have at least experienced what is possible. But when we went back and heard a DJ, wow, only young people can take that.
Since Earl was kind enough to post the impulse response for his speaker I took the time to make some response plots. The plot shows the response, 10 Hz to 10k Hz scale, with consecutively shorter windows; 0.04 sec, 0.02 sec, 0.01 sec, 0.005 sec and 0.0025 sec. The insert shows the response for the shortest window to 30K. A 1/2 Blackman window was used. The lower plot shows the impulse response. The response before 0.0065 msec was padded with zeros. The start of the window was at T = 0.0065 sec. No smoothing was applied.
An externally hosted image should be here but it was not working when we last tested it.
Water fall
And here is the water fall from Earl's data. Again a 1/2 Blackman window 4 msec with was used.
And here is the water fall from Earl's data. Again a 1/2 Blackman window 4 msec with was used.
An externally hosted image should be here but it was not working when we last tested it.
Burst decay
Took the liberty to import the IR into Arta to show the Waveleth transfomation options of this program. The Sample rate was set to 44100 Hz so the absolut frequency is wrong. The decay seems very clean except what seems to be some kind of reflection corresponding to approx 100-200 Hz and som kind of resonances at the very top approx. 18kHz.
http://picasaweb.google.dk/morten.henckel/BurstDecay/photo#5091194440161215362
And turned 90 degrees;
http://picasaweb.google.dk/morten.henckel/BurstDecay/photo#5091194448751149970
Morten
Took the liberty to import the IR into Arta to show the Waveleth transfomation options of this program. The Sample rate was set to 44100 Hz so the absolut frequency is wrong. The decay seems very clean except what seems to be some kind of reflection corresponding to approx 100-200 Hz and som kind of resonances at the very top approx. 18kHz.
http://picasaweb.google.dk/morten.henckel/BurstDecay/photo#5091194440161215362
And turned 90 degrees;
http://picasaweb.google.dk/morten.henckel/BurstDecay/photo#5091194448751149970
Morten
Re: Why Horn Guys Like Tubes
Interesting... I've been experimenting with various horn setups in the last few years and I came to the conclusion that a majority of the difference in sound from low to high volume with horn systems was due to uneven power compression in the system- at low volume the system is flat and sounds nice, but once you turn the system up to an appreciable level, the midrange and/or midbass is dissipating enough power to compress a little while the horn is still laughing at the power it is seeing so the response ends up being tilted upward and the system sounds bright or harsh. I think most of us can agree that lowering the output of a midrange or midrange/midbass 1 dB will change the tonality of the system pretty significantly, and tilting the response upward is certainly more offensive than tilting the response downward, as would be the case in using a dome tweeter (unless the midrange/midbass was very inefficient).
FYI, here's the story that lead me to believe this. I used to run a setup that had an Eminence Beta 8 (~95 dB sens) for a midrange mated to a horn (~108 dB sens) at 2k. It sounded pretty good until you cranked it up, but when I started looking at how much power was being dissipated in the 8, I noticed that above ~10w (estimated based on RMS voltage at the terminals) the tonality of the system started changing and it got worse (brighter) as the system was turned up. Then I looked at as many real numbers on power compression for comparable drivers as I could find, and can to the conclusion that I could expect about 1 dB of compression at ~25w and 2 dB at ~50w. The compression numbers were definately inline with my observations, although I certainly never assumed any of my "findings" to be scientifically rigorous.
The next step of course was to design a speaker that had similar properties but that had predicted negligible power compression (and then lean on her), so I put together a system, partially shown earlier in this thread, consisting of a horn, a B&C 12PE32 midrange (~102 dB sens and 2.5" coil), and an Eminence Kappalite 3015 (~99 dB sens and 3" coil). Compared to the Beta 8, the 12PE32 has 7 dB more sensitivity, and has a bigger coil, so I'd estimate it has about 10 dB more thermal compression headroom (output at same compression). Instead of using a ~95 dB senstivity midbass in a large sealed enclosure (previous setup), the Kappalite is in 2.5 cubes ported tuned to 70 Hz, yielding not just a more voltage sensitive setup, but a giant impedance spike over about half of its bandwidth (high efficiency) and over the other half a nominal impedance of about 10 ohms, meaning this thing has about 7 dB more thermal headroom than the previous midbass. The system is driven with a 6 channel amp that has about 35V rails, so with anything having a crest factor resembling that of music, none of the drivers are going to be dissipating more than a handful of watts.
The result? When you turn the system up the tonality doesn't change very much. Most of the difference from low to high volume is gone, although there is still a little left. I did an A/B comparison with a good commercially available dome tweeter setup (a JBL LSR 1/4/12 speaker w/ some EQ), with the two systems level matched and a few trained listeners, and we pretty much agreed that up to and a ways past the point where the dome tweeter crapped out, there was no change in the horn system's sound. At the highest volumes though, there was some "degradation" although I'm not sure that that wasn't due to the amp running out of steam or the music being played back at an unnaturally high volume. Since that time I have slightly modified the system, including the horn, and have lived with it and spent more time tuning it. I am left wondering how significant the level dependence of the perception of HOM can be. I have heard quite a few horn systems that sound shitty when turned up, but they usually either sound the same at low volume or I have good reason to believe that power compression in the midrange/midbass is the culprit. My experience is limited though. I should say that I have not heard the Avante Guard speakers you were talking about though.
This is just something to think about. It is likely that my next setup will be a OS waveguide/15 setup, especially if a group buy on waveguides gets started. Maybe at that point I'll see things differently.
Patrick Bateman said:
Interesting... I've been experimenting with various horn setups in the last few years and I came to the conclusion that a majority of the difference in sound from low to high volume with horn systems was due to uneven power compression in the system- at low volume the system is flat and sounds nice, but once you turn the system up to an appreciable level, the midrange and/or midbass is dissipating enough power to compress a little while the horn is still laughing at the power it is seeing so the response ends up being tilted upward and the system sounds bright or harsh. I think most of us can agree that lowering the output of a midrange or midrange/midbass 1 dB will change the tonality of the system pretty significantly, and tilting the response upward is certainly more offensive than tilting the response downward, as would be the case in using a dome tweeter (unless the midrange/midbass was very inefficient).
FYI, here's the story that lead me to believe this. I used to run a setup that had an Eminence Beta 8 (~95 dB sens) for a midrange mated to a horn (~108 dB sens) at 2k. It sounded pretty good until you cranked it up, but when I started looking at how much power was being dissipated in the 8, I noticed that above ~10w (estimated based on RMS voltage at the terminals) the tonality of the system started changing and it got worse (brighter) as the system was turned up. Then I looked at as many real numbers on power compression for comparable drivers as I could find, and can to the conclusion that I could expect about 1 dB of compression at ~25w and 2 dB at ~50w. The compression numbers were definately inline with my observations, although I certainly never assumed any of my "findings" to be scientifically rigorous.
The next step of course was to design a speaker that had similar properties but that had predicted negligible power compression (and then lean on her), so I put together a system, partially shown earlier in this thread, consisting of a horn, a B&C 12PE32 midrange (~102 dB sens and 2.5" coil), and an Eminence Kappalite 3015 (~99 dB sens and 3" coil). Compared to the Beta 8, the 12PE32 has 7 dB more sensitivity, and has a bigger coil, so I'd estimate it has about 10 dB more thermal compression headroom (output at same compression). Instead of using a ~95 dB senstivity midbass in a large sealed enclosure (previous setup), the Kappalite is in 2.5 cubes ported tuned to 70 Hz, yielding not just a more voltage sensitive setup, but a giant impedance spike over about half of its bandwidth (high efficiency) and over the other half a nominal impedance of about 10 ohms, meaning this thing has about 7 dB more thermal headroom than the previous midbass. The system is driven with a 6 channel amp that has about 35V rails, so with anything having a crest factor resembling that of music, none of the drivers are going to be dissipating more than a handful of watts.
The result? When you turn the system up the tonality doesn't change very much. Most of the difference from low to high volume is gone, although there is still a little left. I did an A/B comparison with a good commercially available dome tweeter setup (a JBL LSR 1/4/12 speaker w/ some EQ), with the two systems level matched and a few trained listeners, and we pretty much agreed that up to and a ways past the point where the dome tweeter crapped out, there was no change in the horn system's sound. At the highest volumes though, there was some "degradation" although I'm not sure that that wasn't due to the amp running out of steam or the music being played back at an unnaturally high volume. Since that time I have slightly modified the system, including the horn, and have lived with it and spent more time tuning it. I am left wondering how significant the level dependence of the perception of HOM can be. I have heard quite a few horn systems that sound shitty when turned up, but they usually either sound the same at low volume or I have good reason to believe that power compression in the midrange/midbass is the culprit. My experience is limited though. I should say that I have not heard the Avante Guard speakers you were talking about though.
This is just something to think about. It is likely that my next setup will be a OS waveguide/15 setup, especially if a group buy on waveguides gets started. Maybe at that point I'll see things differently.
I also worked a little bit with the impulse response file. I imported it into Liberty Instrument's Praxis.
Regardless of all the interesting stuff you can do by transforming the data, the IR itself is interesting and says a lot about the design compromises. Does anyone care to discuss any of this? Perhaps another thread would be the best place for such a discussion.
And you do not need my graphing of the data. John's IR graph looks perfectly good to me. It shows everything you need to see to get some ideas about what is being done with the system. You can see it whether you agree with it or not.
If my wishes counted for anything I would like to see the IR of just the high end horn. I am not, however, asking that anyone supply such an IR file. Just a wish and motivated just by pure curiosity (and a desire not to have to build a horn just to test a hypothesis about how horns "really" function).
I believe this thread was to be about the "wave guide" for a top end horn. Unfortunately, the low frequency portion of the system is masking much of the data on the low end output of the horn. I can make some assumptions about low end horn performance, but I cannot get a valid decay response for the horn.
Best and all that,
Mark
Regardless of all the interesting stuff you can do by transforming the data, the IR itself is interesting and says a lot about the design compromises. Does anyone care to discuss any of this? Perhaps another thread would be the best place for such a discussion.
And you do not need my graphing of the data. John's IR graph looks perfectly good to me. It shows everything you need to see to get some ideas about what is being done with the system. You can see it whether you agree with it or not.
If my wishes counted for anything I would like to see the IR of just the high end horn. I am not, however, asking that anyone supply such an IR file. Just a wish and motivated just by pure curiosity (and a desire not to have to build a horn just to test a hypothesis about how horns "really" function).
I believe this thread was to be about the "wave guide" for a top end horn. Unfortunately, the low frequency portion of the system is masking much of the data on the low end output of the horn. I can make some assumptions about low end horn performance, but I cannot get a valid decay response for the horn.
Best and all that,
Mark
Re: Re: Why Horn Guys Like Tubes
It's a bummer that so few people have done the foam "mod." After listening this way for years, I think I've been able to train myself to hear "HOMs" to some extent. I've extolled it's virtues quite a bit, no need to waste too much bandwidth in this thread. Suffice to say, it makes horns sound very similar to direct radiators, while still offering the virtues that make horns popular. Your basic "best of both worlds" deal.
Rybaudio said:
It's a bummer that so few people have done the foam "mod." After listening this way for years, I think I've been able to train myself to hear "HOMs" to some extent. I've extolled it's virtues quite a bit, no need to waste too much bandwidth in this thread. Suffice to say, it makes horns sound very similar to direct radiators, while still offering the virtues that make horns popular. Your basic "best of both worlds" deal.
Comments on data
First, no attempt ewas made to get data below about 200 Hz that had any meaning, so lets ignore this area. At about 600-700 Hz. the woofer has a suspension resonance that is clearly visible, audibility is still being determined. The compression driver clearly has a resonance at about 15-18 kHz. Don't they all!?
The decay in the 6-7 kHz. region is possibly HOM and possibly a resonance of the mouth diffraction.
I have attached an impulse of just the waveguide and compression driver.
As to the discussion of the perception of horns. The work that we did indicated that power compression in a compression driver is insignificant as far as audibility goes (See our AES paper). Our further work indicates that the HOM should be highly audible and level dependent. Make sure that when you "hypothesize" an explaination for some subjective results that this hypothesis be consistant with existing proven facts.
First, no attempt ewas made to get data below about 200 Hz that had any meaning, so lets ignore this area. At about 600-700 Hz. the woofer has a suspension resonance that is clearly visible, audibility is still being determined. The compression driver clearly has a resonance at about 15-18 kHz. Don't they all!?
The decay in the 6-7 kHz. region is possibly HOM and possibly a resonance of the mouth diffraction.
I have attached an impulse of just the waveguide and compression driver.
As to the discussion of the perception of horns. The work that we did indicated that power compression in a compression driver is insignificant as far as audibility goes (See our AES paper). Our further work indicates that the HOM should be highly audible and level dependent. Make sure that when you "hypothesize" an explaination for some subjective results that this hypothesis be consistant with existing proven facts.