Hi Helmuth.
I created the mold by myself, a slightly variation of tractrix.
It´s not good to praise ones own creations, but these are the most coloration-free horn I´ve ever heard.
I use BMS4590, I´ve also heard them with Radians.
You are always welcome to come by and let your ears decide, wether they have big colorations or not.
Michael
I created the mold by myself, a slightly variation of tractrix.
It´s not good to praise ones own creations, but these are the most coloration-free horn I´ve ever heard.
I use BMS4590, I´ve also heard them with Radians.
You are always welcome to come by and let your ears decide, wether they have big colorations or not.
Michael
This invitation is accepted Michael,
From Winterswijk to Dortmund is about 100km nice trip with the Guzzi.
Here is another professional audio engineer active how has used a low density polyurethane foam damper within the throat off the horn.
The wave's how want to travel a long way through the horn meets much damping foam the waves how travel straight through the horn have the smallest amount of damping.
So the foam HOM damper discriminates the distortion waves.
Theory off Dr Geddes.Summa
Regards, Helmuth
From Winterswijk to Dortmund is about 100km nice trip with the Guzzi.
Here is another professional audio engineer active how has used a low density polyurethane foam damper within the throat off the horn.
The wave's how want to travel a long way through the horn meets much damping foam the waves how travel straight through the horn have the smallest amount of damping.
So the foam HOM damper discriminates the distortion waves.
An externally hosted image should be here but it was not working when we last tested it.
Theory off Dr Geddes.Summa
Regards, Helmuth
Hi Dr Geddes,
When I read your plea.
I see you want to distract the attention off the distortion discussion towards a room acoustics- directivity story.
You are making some good points by saying the room reflections (acoustics) are also needed to have good sound reproduction. And you have to use them the right way.
The foam plug seems to me a great simple solution to a big horn problem.
What is the claim that 25% THD isn't audible, what are the test pre-sets to not hear 25% distortion?
And you are making some big steps passed many relevant parts in the audio chain to bent the discussion IMO to the room acoustics.
And the Summa design I like only the design can't produce frequency's down to 50 Hz I guess with a closed box and 95dB/Wmtr. Am I right.
Regards Helmuth
When I read your plea.
I see you want to distract the attention off the distortion discussion towards a room acoustics- directivity story.
You are making some good points by saying the room reflections (acoustics) are also needed to have good sound reproduction. And you have to use them the right way.
The foam plug seems to me a great simple solution to a big horn problem.
What is the claim that 25% THD isn't audible, what are the test pre-sets to not hear 25% distortion?
And you are making some big steps passed many relevant parts in the audio chain to bent the discussion IMO to the room acoustics.
And the Summa design I like only the design can't produce frequency's down to 50 Hz I guess with a closed box and 95dB/Wmtr. Am I right.
Regards Helmuth
Helmuth
A few years ago B&C, my wife and I published a study on the audiblity of distortion in compression drivers. You should read the AES paper for details, but the bottom line was that 25 listeners failed to detect 25% THD in a blind test. They could detect frequency response differences in drivers, but not non-linear distortion. Many people have found these results disturbing, but they are completely in line with our studies on nonlinearity that we did back in 2002 (also on my web site). THD is a meaningless number as far as audibilty is concerned. There ARE audible nonlinear distortions, but THD or IMD don't identify them.
"you are making some big steps passed many relevant parts in the audio chain "
In my 40 years in audio I have found that only the loudspeakers and rooms really matter. Sure there are bad electronics, but there are also inaudible electronics. I use the later. When that is done the loudspeaker and room are the sole remaining problems and those can never be "inaudible". So in essence, the electronics problem is done, but the loudspeaker/room problem is only just being properly dealt with.
"the design can't produce frequency's down to 50 Hz "
I never design rooms with only main speakers, I always use multiple distributed subs. In this case, how low the mains go is not really that important. The lowest octave is handled by the subs, where there are usually three, all overlaping with the mains - no crossovers. This entire concept has been well hashed out elsewhere.
The sensitivity is about 96 dB free field.
A few years ago B&C, my wife and I published a study on the audiblity of distortion in compression drivers. You should read the AES paper for details, but the bottom line was that 25 listeners failed to detect 25% THD in a blind test. They could detect frequency response differences in drivers, but not non-linear distortion. Many people have found these results disturbing, but they are completely in line with our studies on nonlinearity that we did back in 2002 (also on my web site). THD is a meaningless number as far as audibilty is concerned. There ARE audible nonlinear distortions, but THD or IMD don't identify them.
"you are making some big steps passed many relevant parts in the audio chain "
In my 40 years in audio I have found that only the loudspeakers and rooms really matter. Sure there are bad electronics, but there are also inaudible electronics. I use the later. When that is done the loudspeaker and room are the sole remaining problems and those can never be "inaudible". So in essence, the electronics problem is done, but the loudspeaker/room problem is only just being properly dealt with.
"the design can't produce frequency's down to 50 Hz "
I never design rooms with only main speakers, I always use multiple distributed subs. In this case, how low the mains go is not really that important. The lowest octave is handled by the subs, where there are usually three, all overlaping with the mains - no crossovers. This entire concept has been well hashed out elsewhere.
The sensitivity is about 96 dB free field.
Hi Gedlee,
thanks for the quick reply.
On the Dutch forum some guy's are mad about proving they are right. And the tool to do this is ABX so sample A and B and blind you have to say who is X.
Now they mathematical calculated how many trails are needed to become a usable result to say it audioble or not.
There for you need minimal of 67 test trails and 49 right answers. I do not know the details of your test but the test trail it is a knowledge at its self.
I have to verify if I understand Non linear distortion right. I think this is variation of SPL over the frequency range.
What are non linear distortions. Intermodulation distortion of an amplifier for example. Or horn colouration.
Regards, Helmuth
thanks for the quick reply.
On the Dutch forum some guy's are mad about proving they are right. And the tool to do this is ABX so sample A and B and blind you have to say who is X.
Now they mathematical calculated how many trails are needed to become a usable result to say it audioble or not.
There for you need minimal of 67 test trails and 49 right answers. I do not know the details of your test but the test trail it is a knowledge at its self.
I have to verify if I understand Non linear distortion right. I think this is variation of SPL over the frequency range.
What are non linear distortions. Intermodulation distortion of an amplifier for example. Or horn colouration.
Regards, Helmuth
The test was done statistically correct using ANOVA. You can read this is the AES paper. The results wre all statistically valid because each subject did several trials and there were 25 subjects.
You should review the difference between linear and non-linear distortions. SPL variations with frequency are linear. Those are clearly audible. Nonlinear distortions vary with level and create new signals not present at the input - such as harmonics if the input is a sine wave. This are hardly ever audible with real world signals, but can be for SOME types of nonlinearity. But THD and IMD are completely uncorrelated with the subjective impression. They are meaningless measures of nonlinearity.
ABX testing is more than 20 years old and critical to making any statements about audibility. I'm surprised that you are just learning about this.
The most unreliable subject in the world is a human - this has been shown time and time again. Without some controls their evaluations are without vailidity - totaly biased subjectively by the listener.
You should review the difference between linear and non-linear distortions. SPL variations with frequency are linear. Those are clearly audible. Nonlinear distortions vary with level and create new signals not present at the input - such as harmonics if the input is a sine wave. This are hardly ever audible with real world signals, but can be for SOME types of nonlinearity. But THD and IMD are completely uncorrelated with the subjective impression. They are meaningless measures of nonlinearity.
ABX testing is more than 20 years old and critical to making any statements about audibility. I'm surprised that you are just learning about this.
The most unreliable subject in the world is a human - this has been shown time and time again. Without some controls their evaluations are without vailidity - totaly biased subjectively by the listener.
Hi Earl,
Couldn't agree more (your points in posts 45 and 47). It's amazing how many people still swear blind that strange tweako changes to audio gear gives them huge improvements - despite some of the tricks not even being based on the laws of physics.
Also, there are still quite a few who decry ABX testing as unsuitable for audio. To be honest - I think there is a valid point that putting someone under 'test conditions' can stress them, and alter their perception. The problem is that I never see valid alternatives being put forward.
On your comments about non-linear distortion - what types have you found most audible? John Krute (Zaph Audio) has some writings about this on his website, and it's always interesting to see if findings correlate. BTW Could you give me a reference to your AES paper, as it's always good to expand my (limited) knowledge.
On the subject of expanding polyurethane foam - I've just had 4LB of the stuff delivered, though I've only been able to source a 2LB/ft^3 density product so far. However, I'll give it a test and see how I get on. Many thanks for the points about continued expansion. I'll take care. My biggest concern is that the datasheet indicates 20C as the optimum working temp. Not gonna get that in my workshop during an English November!
Couldn't agree more (your points in posts 45 and 47). It's amazing how many people still swear blind that strange tweako changes to audio gear gives them huge improvements - despite some of the tricks not even being based on the laws of physics.
Also, there are still quite a few who decry ABX testing as unsuitable for audio. To be honest - I think there is a valid point that putting someone under 'test conditions' can stress them, and alter their perception. The problem is that I never see valid alternatives being put forward.
On your comments about non-linear distortion - what types have you found most audible? John Krute (Zaph Audio) has some writings about this on his website, and it's always interesting to see if findings correlate. BTW Could you give me a reference to your AES paper, as it's always good to expand my (limited) knowledge.
On the subject of expanding polyurethane foam - I've just had 4LB of the stuff delivered, though I've only been able to source a 2LB/ft^3 density product so far. However, I'll give it a test and see how I get on. Many thanks for the points about continued expansion. I'll take care. My biggest concern is that the datasheet indicates 20C as the optimum working temp. Not gonna get that in my workshop during an English November!
sploo said:
The neighsayers want to throw out blind testing because it gets the wrong results. John Atkinson says this himeself. He discounts ABX tesing because he gets different results than when he listens with full knowledge - hence, the test is flawed. Well something IS flawed thats for sure.
The type of distortion that is most audible is crossover distortion in an amp. This is audible at even very small percentage of THD. Loudspeakers tend not to suffer this kind of distortion (although it is possible) and in general the types of nonlinear distortions that tend to occur in loudspeakers are the least audible. For example 2nd order (2nd harmonic for a sine wave) is almost never audible under any circumstances. Third is only marginally more audible. These two dominate the THD in a loudspeaker while the more audible ones like the 6th, 7th etc. and higher are seldom even measured.
Poly foam is highly sensitive to temp and humidity so the expansion you get on one day may be completely different than on another. Good luck.
Helmuth said:I am reading them right now but have some problem with the English terms.Because I am Dutch.
Thanks for the links. Your English is way better than my Dutch!
gedlee said:
Absolutely. If the test shows your pre(mis?)conceptions are wrong, discount the preconceptions, not the test. The problem is that these arguments always degenerate into an objectivist vs subjectivist rant, and it's impossible to have a serious discussion on the topic.
I'd love to find some common ground on this - i.e. methods of testing that both sides can agree are valid, but I don't hold out much hope.
gedlee said:
Could you explain that in novice terms (crossover distortion in an amp)? My background isn't in electronics, and I thought you got (passive or active) crossovers in speakers, and active crossovers in external boxes - not familiar with a crossover in an amp.
Your points on speakers are understood and appreciated.
gedlee said:
Thanks. I think
. I was aware that temperature would have a major effect from reading the datasheet, so I guess I'll just have to be careful and experiment. It's occurred to me that I could use a thin plastic bag inside the speaker enclosure, and pour the mixture into that. Hopefully, this would give an internal mold 'plug' with the bag stuck to the outside, which wouldn't have stuck to the speaker whilst the poly was curing, and won't stick to the concrete when molding the finished unit. The surface detail on the internal mold isn't critical, just that the volume is correct.
Crossover distortion happens when a class B or class AB amp switches from one set of output devices to the other. The idea is that feedback elliminates (minimizes!) any discontinuities in the "crossover" or switching from one device to the other. But feedback does not elliminate it altogether and even very small residuals can still be quite audible. Class A amps are imune to this type of distortion but tend to have very high 2nd and or third order nonlinearities. But these lower orders are not audible while the crossover problems are. Hence the reason why people like high THD class A amps.
The reason that crossover is so audible is that it is very high order and it occurs at very low signal levels - basically ALL signals are affected. The lower orders only significantly affect large signals, but then the signal masks the artifacts.
But not all amps have crossover distortion, its just that none of the typical measurements of amps show if it does of doesn't. You have to look at the spectrum of a sine wave for very very low signals - all the way down into the noise floor. Then you can see the distortion that is audible. This test deffinately seperates good amps from bad but is never done in practice (or at least never shown).
Another significant point is that loudspeaker nonlinearity will not mask crossover distortion because this type of thing is not very common in a loudspeaker. The loudspeaker will mask almost any type of large signal nonlinearity in the amp however - except clipping, because clipping is very high order. A soft clipped amp would not be detectable under any circumstances (this I have done tests to prove) (assuming the amp has no crossover distortion.).
The reason that crossover is so audible is that it is very high order and it occurs at very low signal levels - basically ALL signals are affected. The lower orders only significantly affect large signals, but then the signal masks the artifacts.
But not all amps have crossover distortion, its just that none of the typical measurements of amps show if it does of doesn't. You have to look at the spectrum of a sine wave for very very low signals - all the way down into the noise floor. Then you can see the distortion that is audible. This test deffinately seperates good amps from bad but is never done in practice (or at least never shown).
Another significant point is that loudspeaker nonlinearity will not mask crossover distortion because this type of thing is not very common in a loudspeaker. The loudspeaker will mask almost any type of large signal nonlinearity in the amp however - except clipping, because clipping is very high order. A soft clipped amp would not be detectable under any circumstances (this I have done tests to prove) (assuming the amp has no crossover distortion.).
Hi Guys, Don't mean to hijack your thread but see you guys have experience with casting concrete. I'm doing to some fiberglass tractrix variants and I need a plug that's going to last and thinking concrete ie in reverse of casting concrete with fiberglass, (well id be doing this initially to create the plug) but taking fiberglass molds from a cast concrete plug.
Has anyone here done this? Being void free and achieving a high gloss polish of the concrete being my concern.
Tia.
Has anyone here done this? Being void free and achieving a high gloss polish of the concrete being my concern.
Tia.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.