You are misunderstanding my use of the term "linear". I do not mean "linear frequency response".
HOMs are a linear phenomena in the sense of a "linear system", not "linear frequency response".
Last edited:
Art, I don't think that the 6 dB range is correct, it was 6 dB steps over a much larger range. And of course no one test can prove everything, but it is incontrovertible evidence that nonlinear distortion is not necessarily a significant factor. It was not in our test.
As to your test, we resolved this long ago, or did you forget. Your test covered a much higher SPL range well beyond what our test was able to do. Perhaps nonlinearity is audible at those extremes, we didn't test that. But the levels that we did test at are well beyond anything that one would listen to in a home, albeit not what one might listen to in an auditorium. We are in different worlds and you need to keep that in mind.
So "prove" maybe not, but I will still side with the large compilation of objective evidence that says that THD is meaningless and that nonlinear distortion in a home setting is not likely to be an issue.
Earl,
Evidently you don't remember your test protocol (or mine), you tested only three levels, 14, 20, and 28 volts, (12.25 to 50 watts assuming 16 ohm drivers) resulting in only a 6 dB range difference, and a 6 dB range of distortion from approximately 10 to 18% at 1 kHz, and 18 to 31% at 6 kHz.
Since those levels of distortion were not compared to a low distortion level (like 2.83) or a reference recording there would be little chance of anyone telling them apart. To confound things more, the source recording (live Talking Heads "burning down the house") has plenty of distortion to start with. No meaningful subjective distortion perception results could be derived from your protocol.
I would agree that large pro woofers and compression drivers distortion at home levels would not generally be unacceptable, though the Klippel tests clearly show very low levels in the range of .5% can be distinguished from undistorted program material by many individuals.
My test http://www.diyaudio.com/forums/multi-way/212240-high-frequency-compression-driver-evaluation.html used levels both lower and higher than yours (including home levels, which your test did not, 28 volts would be around 126 dB at one meter on a typical horn), and also allows the recordings of multiple equalized drivers to be compared the original recording. Being actual recordings all normalized to the same recording level, one can listen at a comfortable level to a very wide SPL range, and determine what level is "too much" for a particular driver.
Art
Attachments
Art I remember three levels, but not what they were.
I don't agree with your critique and apparently neither did the reviewers of the paper. The results stand as per reviewed data that you cannot just wave your hand across just because you don't like the results. How do you know how much "distortion" is in the recording? You are just guessing. And if someone cannot tell the difference between 18% and 31% then just where CAN they tell the difference. Your objections simply do not amount to much.
I assume the discussion on distortion will be moved to a new thread, which I will follow with very much interest.
On the topic of CLD it was mentioned earlier but I would like to clarify. Is Green Glue a good compound for CLD in home speakers? Would mixing with a latex caulk / adhesive benefit anything?
Thanks
Shawn
On the topic of CLD it was mentioned earlier but I would like to clarify. Is Green Glue a good compound for CLD in home speakers? Would mixing with a latex caulk / adhesive benefit anything?
Thanks
Shawn
Earl,
I know you don't remember, which is why I posted your test protocol and the voltage levels you wrote were used in post 506.
One could not tell how much distortion was in the recording compared to the recorded driver's distortion with your test protocol.
Having heard the Talking Heads live recording of "burning down the house" through speakers with low distortion, it has obvious distortion typical of live rock recordings of the era. As a guess, the tube compressors used on David Byrne's vocals probably are adding more than 10% THD in general, and more on peaks.
Saw David Byrne's show just a few years back, he's still using the same old studio processing of that era live, rock & roll has always been distortion based, giving the signature harmonic arrangements of various processing and musical instrument amplification, and tape saturation to "enhance" drum harmonics.
Anyway, we are straying off topic in this thread, talking about distortion perception rather than problems generated by cabinet material and the "best" materials to eliminate those problems.
Art
Last edited:
Horns are mostly linear devices, but at extremely high SPLs can yield some very low order nonlinearities. These would be mostly inaudible and non-existent in any home system.
Internal diffraction in a horn is linear, meaning that it is always there in equal amount at any SPL level. BUT, we tend to become more sensitive to this type of linear aberration as the SPL rises (why this occurs is not clear, but it clearly happens as even Prof Moore noted in his famous papers.) Hence, what happens is that as the SPL level goes up we begin to hear the horn start to sound rough, honky, whatever adjective you want to use here. It is natural to assume this is nonlinear distortion that increases with increasing level, but it is not, it a linear distortion whose perception is nonlinear. It is basically impossible for our ears to tell the two apart - but measurements can.
I believe we used the studio version and it has always sounded clean to me. And even though we did not do a comparison of modified versus non-modified in our protocol, I did such a comparison and had the difference been obvious I would have changed the protocol.
Art, this test was done in the most valid way that we could think of. Sure you can criticize it, but you can't just write it off. For the limited range of validity that it did have, it put a stake in the ground as a valid data point. Until there is as competent a test showing otherwise, I have to accept the results.
I should point out that your test is not blind and as such would fail even the most basic rules for psychological testing.
Earl,
1) I just reported what you wrote in the report under "Stimulus", I couldn't know if you changed it from "Live" to the studio version, but both have plenty of distortion, though probably less than the level of distortion in the drivers the music was played through.
2) My tests were made so others could hear what a variety of drivers sound like at different drive levels, not for psychological testing.
Attachments
Art
1) My mistake on the recording.
2) I understand that - we have different objectives and we have to assume that each others tests are valid within their limitations and figure out if they overlap somewhere and in that area do they disagree. I don't think that they do, but then again, I am not sure that there is much overlap.
1) My mistake on the recording.
2) I understand that - we have different objectives and we have to assume that each others tests are valid within their limitations and figure out if they overlap somewhere and in that area do they disagree. I don't think that they do, but then again, I am not sure that there is much overlap.
There's a whole world outside your listening room where budget and size constraints are real factors. Your utopic ideas does not apply there.
If indeed eliminating audible non-linearity isn't the hardest or even most important design goal, and I think I agree that it isn't, then what is?
I suspect that former requires little more than choosing drivers appropriately and then using them within their limits. That is especially easy if budget isn't an issue but even budget drivers can sound good within their limits. Synergy horns in particular can get amazing results from inexpensive drivers.
Going back to my question....Gedlee seems to have put a very high degree of effort into eliminating all sources of diffraction and HOMs. From older threads, I see Dr. Geddes and Tom Danley have agreed to disagree on whether eliminating diffraction or providing point source behavior is more important for a home system. And then there is getting the power response correct.
To get back to the topic of the thread: the latest Gedlee speakers are an example of a very good cabinet systems. If I recall correctly, Dr. Geddes is molding the foam boards himself in what is no doubt a very professional but still DIY environment. I find that quite admirable but also too hard for me. I'm most comfortable with standard baltic birch plywood construction and have begun augmenting it with a 2nd layer, using green glue for viscoelasticity and isolation mounting for bass drivers. I don't claim its the best but I'm pretty confident it can be made good enough, and its a technology I can master.
I suspect that former requires little more than choosing drivers appropriately and then using them within their limits. That is especially easy if budget isn't an issue but even budget drivers can sound good within their limits. Synergy horns in particular can get amazing results from inexpensive drivers.
Going back to my question....Gedlee seems to have put a very high degree of effort into eliminating all sources of diffraction and HOMs. From older threads, I see Dr. Geddes and Tom Danley have agreed to disagree on whether eliminating diffraction or providing point source behavior is more important for a home system. And then there is getting the power response correct.
To get back to the topic of the thread: the latest Gedlee speakers are an example of a very good cabinet systems. If I recall correctly, Dr. Geddes is molding the foam boards himself in what is no doubt a very professional but still DIY environment. I find that quite admirable but also too hard for me. I'm most comfortable with standard baltic birch plywood construction and have begun augmenting it with a 2nd layer, using green glue for viscoelasticity and isolation mounting for bass drivers. I don't claim its the best but I'm pretty confident it can be made good enough, and its a technology I can master.
nc535 - I am not sure that Tom and I disagree on point source behavior, I think that we both agree that's important. Tom's design do not minimize diffraction as much as mine do, but then his target market is something that mine can't achieve. So its a matter of "goals". For his market, his approach may be optimal, I don't know, I haven't done any work in his market. So we may agree completely for each of us in our own markets.
The most important thing in a loudspeaker are the linear distortions. This is what Toole says and I agree completely. What Toole does not get into are the linear distortions that have nonlinear perception. His work is always done at one single fixed SPL level and as such his work would never encounter the nonlinear perception issue. There is only scant work done on this perception issue. Our paper is the only one that I know of and that was just a dip in the pond showing that such things do occur, but not calibrating it to anything.
The most important thing in a loudspeaker are the linear distortions. This is what Toole says and I agree completely. What Toole does not get into are the linear distortions that have nonlinear perception. His work is always done at one single fixed SPL level and as such his work would never encounter the nonlinear perception issue. There is only scant work done on this perception issue. Our paper is the only one that I know of and that was just a dip in the pond showing that such things do occur, but not calibrating it to anything.
No wonder I just said FI & used a truck for bonding psi to laminate Mdf.
OK ..read all the lack of concensus, back and forth nitpicking, mostly OT, nanny-nanny-poopoo
techo-yap....That the only certainty I can conclude (even with BS Applied Math) is that I now know of several product lines to avoid (rep of own product): Same reason roosters aren't eaten.....their behavior.
OK, not funny. But all kidding aside.....seriously !
2nd & 3rd harmonics....
....Put either palm of hand flat on a table,
Observe difference as table struck with my ball-pean hammer comparing round anvil to flat.
WACK ! . Sorry, ... just couldn't resist
But Now... Ya got somerhing to really argue loudly about !
Sorry Mod-man, was just there..... soo wide open, NMF
OK ..read all the lack of concensus, back and forth nitpicking, mostly OT, nanny-nanny-poopoo
techo-yap....That the only certainty I can conclude (even with BS Applied Math) is that I now know of several product lines to avoid (rep of own product): Same reason roosters aren't eaten.....their behavior.
OK, not funny. But all kidding aside.....seriously !
2nd & 3rd harmonics....
....Put either palm of hand flat on a table,
Observe difference as table struck with my ball-pean hammer comparing round anvil to flat.
WACK ! . Sorry, ... just couldn't resist
But Now... Ya got somerhing to really argue loudly about !
Sorry Mod-man, was just there..... soo wide open, NMF
