Gedlee on distortion

[ksneote]

I was reading Gedlee and his speaker design/room theories, to learn more. I ran across an interesting paragraph in this paper, down near the bottom: http://www.gedlee.com/downloads/Philosophy.pdf

"What is the distortion level?

There is a problem inherent with this question. GedLee has shown that THD
and IMD are meaningless numbers in relating the perception of distortion. So
what would be the point of measuring or showing them? Another point is that it
would be extremely loud (read hearing damage) before these drivers were even
close to an excursion domain where nonlinear distortion was even an issue."

More information can be found here, but it's quite beyond me.

http://gedlee.com/distortion_perception.htm

Any thoughts? Should we ignore distortion graphs etc?

[dcibel]

Re: Gedlee on distortion

You have to take Gedde's words very literally. THD itself doesn't tell you much, that does not mean that distortion is not important. Relative levels of each harmonic must be analyzed, forget about the totals.

[johnnyrichards]

Re: Gedlee on distortion

Ignore Geddes, actually.

[Pete Schumacher]

Re: Gedlee on distortion

Ignore Geddes, actually.

+1

It's a really stupid paragraph actually.

You read distortion graphs to gauge where a driver should be limited in range.

While the drivers he's talking about are probably large, pro-style woofers and compression tweeters, it probably is true that in order to hear the distortion, you'd have to be running the drivers at very high SPL.

But for smaller drivers, it's definitely a valuable data set.

Such conclusions in that paragraph should not be applied across the board to drivers in general, that's for sure.

As an aside, Geddes argued with me and others in a thread that room gain did not exist . . . :rolleyes:

[isaeagle4031]

Re: Gedlee on distortion

Ignore Geddes, actually.

+2.

[civit]

Re: Gedlee on distortion

Geddes' meaning is very precise. He is saying that THD is not a perceptual measure. He is not saying that distortion doesn't matter, he is saying that it is possible to distort a signal in such a way that THD increases but it doesn't actually sound worse. This is not really a controversial statement - there is no reason to believe that it would.

I suspect his statement about room gain is similar - he could even simply be making the pedantic point that a room doesn't add energy to the system and therefore can't have 'gain'.

Geddes' work is pretty great, I don't know why he should be ignored.

[Face]

Re: Gedlee on distortion

I suspect his statement about room gain is similar - he could even simply be making the pedantic point that a room doesn't add energy to the system and therefore can't have 'gain'.

Geddes loves to be cryptic at times and play the semantics game.

[Jeff B.]

Re: Gedlee on distortion

Geddes is a smart guy with a lot of knowledge in this field, and a lot of good information on acoustics. He is also quite prone to playing semantics games and spinning information to precisely fit his "highly acclaimed Gedlee loudspeakers". When he is wrong, don't expect an admission, which is just silly. (I was in a discussion with him on another forum on modeling diffraction and he was just wrong about the importance of modeling the first baffle peak in a simulation, which his model leaves out. The rest of us moved on.)

Regarding his comments on distortion; I don't know if he is just being very precise and narrow in his statement (but he doesn't really infer this), if he's just mistaken in his views, or if he is purposely spinning information to suit his needs at the expense of a larger truth. Whatever it may be, he knows exactly how he wants his comments to be interpreted.

As for me, after taking lots of distortion measurements on a range of different drivers, and then playing and listening to these drivers under different conditions, I am sufficiently satisfied that higher distortion components produce easily audible effects and I can hear very quickly what it is and what it sounds like. Consequently, very low distortion drivers actually do sound better in a number of ways to me, but it does depend on drive level and where that audibility enters the picture for that particular driver and crossover combination. Still, for some drivers, this threshold is quite low. I am satisfied enough based on my own testing that it wouldn't really matter to me if someone else stated something otherwise.

Jeff B.

[dlr]

Re: Gedlee on distortion

Geddes is a smart guy with a lot of knowledge in this field, and a lot of good information on acoustics. He is also quite prone to playing semantics games and spinning information to precisely fit his "highly acclaimed Gedlee loudspeakers". When he is wrong, don't expect an admission, which is just silly. (I was in a discussion with him on another forum on modeling diffraction and he was just wrong about the importance of modeling the first baffle peak in a simulation, which his model leaves out. The rest of us moved on.).

Jeff, can you point me to this? I don't think that I've seen that thread. I'm surprised that he would dismiss the first peak entirely since that is usually the single biggest impact of baffle diffraction, at least on-axis. He's primarily concerned with diffraction internal to horns if I read him correctly, but the peak should affect the mouth of a horn as well, or so I would assume.

dlr

[Jeff B.]

Re: Gedlee on distortion

Jeff, can you point me to this? I don't think that I've seen that thread. I'm surprised that he would dismiss the first peak entirely since that is usually the single biggest impact of baffle diffraction, at least on-axis. He's primarily concerned with diffraction internal to horns if I read him correctly, but the peak should affect the mouth of a horn as well, or so I would assume.

dlr

In this thread here: http://www.diyaudio.com/forums/multi-way/249310-nearfield-farfield-curve-splicing.html

It's a discussion with Dave Thomas, Andrew Jones, Geddes, and several names you will recognize from here. We are discussing my paper on using the "Blender" spreadsheet that Charlie and I made. The discussion a few pages in moves to diffraction where Geddes states that he only uses a spherical model in simulations. I make the point that his slope will be off if it doesn't include the first peak compared to the spherical diffraction. This is probably around page 9 or so. I haven't dug into it for a while.

[Pallas]

Re: Gedlee on distortion

Any thoughts? Should we ignore distortion graphs etc?

One of the most important parts of reading comprehension generally is to understand the scope of a text. Here, the scope of Geddes' advertising material is clearly stated in its first paragraph: "Earl Geddes line of loudspeaker systems"

On their face, it's hard to see how his statements cannot be true, as applied to those speakers. But as to a 7" 2-way, it's simply not within the scope of Geddes' comments.

As to THD generally, there's no reason not to (mostly) ignore harmonic distortion graphs. They are useful to see how low one can take a tweeter or small midrange. And...that's about it. Otherwise, there is no serious study of which I'm aware that shows a meaningful correlation between THD and listener preferences in loudspeakers. If somebody has one, I'd love to see a cite.

Dr. Toole is, perhaps, more instructive here than Dr. Geddes, because his scope is the much wider universe of loudspeakers generally.

"The end result of this is that traditional measures of harmonic or intermodulation distortion are almost meaningless. They do not quantify distortion in a way that can, with any reliability, predict a human response to it while listening to music or movies. They do not correlate because they ignore any characteristics of the human receptor, an outrageously nonlinear device in its own right. The excessive simplicity of the signals also remains a problem. Music and movies offer an infinite variety of input signals and therefore an infinite variety of distorted outputs.
***
In loudspeakers it is fortunate that distortion is something that normally does not become obvious until devices are driven close to or into some limiting condition. *** In the general population of consumer loudspeakers, it has been very rare for distortion to be identified as a factor in the overall subjective ratings. This is not because distortion is not there or is not measurable, but it is low enough that it is not an obvious factor in judgments of sound quality at normal foreground listening levels." -Sound Reproduction, at 452-3.

But back to Geddes, it's worth noting that unlike the participants in this thread so far, he and his wife actually developed a distortion metric that is claimed to be more useful than the crude, useless THD. Lesser minds often go out of their way to show disdain for their obvious betters, tritely dismissing important contributors rather than engaging them and probing them for truth. That unfortunate part of human nature is unfortunately visible in this thread.

[Pete Schumacher]

Re: Gedlee on distortion

One of the most important parts of reading comprehension generally is to understand the scope of a text. Here, the scope of Geddes' advertising material is clearly stated in its first paragraph: "Earl Geddes line of loudspeaker systems"

On their face, it's hard to see how his statements cannot be true, as applied to those speakers. But as to a 7" 2-way, it's simply not within the scope of Geddes' comments.

As to THD generally, there's no reason not to (mostly) ignore harmonic distortion graphs. They are useful to see how low one can take a tweeter or small midrange. And...that's about it. Otherwise, there is no serious study of which I'm aware that shows a meaningful correlation between THD and listener preferences in loudspeakers. If somebody has one, I'd love to see a cite.

Dr. Toole is, perhaps, more instructive here than Dr. Geddes, because his scope is the much wider universe of loudspeakers generally.

"The end result of this is that traditional measures of harmonic or intermodulation distortion are almost meaningless. They do not quantify distortion in a way that can, with any reliability, predict a human response to it while listening to music or movies. They do not correlate because they ignore any characteristics of the human receptor, an outrageously nonlinear device in its own right. The excessive simplicity of the signals also remains a problem. Music and movies offer an infinite variety of input signals and therefore an infinite variety of distorted outputs.
***
In loudspeakers it is fortunate that distortion is something that normally does not become obvious until devices are driven close to or into some limiting condition. *** In the general population of consumer loudspeakers, it has been very rare for distortion to be identified as a factor in the overall subjective ratings. This is not because distortion is not there or is not measurable, but it is low enough that it is not an obvious factor in judgments of sound quality at normal foreground listening levels." -Sound Reproduction, at 452-3.

But back to Geddes, it's worth noting that unlike the participants in this thread so far, he and his wife actually developed a distortion metric that is claimed to be more useful than the crude, useless THD. Lesser minds often go out of their way to show disdain for their obvious betters, tritely dismissing important contributors rather than engaging them and probing them for truth. That unfortunate part of human nature is unfortunately visible in this thread.

I'd like to know what makes Geddes "our obvious better?"

I've actually listened to GedLee speakers and I'd put most any Jeff Bagby design ahead of anything Geddes has.

For all the write ups and "claimed to be better metrics" it sure hasn't resulted in anything all that special from a sonically superior product perspective. Foam inserts into a waveguide to quell obvious colorations is hardly a reason to cite him as a "better" than many of the actual designers/builders on this forum.

Jeff pretty much spelled out the quirky nature of Geddes. Much of his bluster seems to be to make himself to be more of an expert than he really is. Dismissing baffle diffraction and room "gain" are just a couple of the obvious blunders in the guy's audio character. He is basically saying Allison didn't know what he was talking about.

His statement about THD measurements being meaningless is something that cannot be carried into a general discussion of driver evaluation, something about which even you agree. Simple THD measurements certainly don't tell you everything about how a driver actually sounds under music conditions, but for someone well versed in speaker design and implementation it is also not something that is meaningless.

I'll take Jeff B's advice over Geddes 100% of the time based on the sound of the actual delivered product.

[isaeagle4031]

Re: Gedlee on distortion

I'd like to know what makes Geddes "our obvious better?"

I've actually listened to GedLee speakers and I'd put most any Jeff Bagby design ahead of anything Geddes has.

For all the write ups and "claimed to be better metrics" it sure hasn't resulted in anything all that special from a sonically superior product perspective. Foam inserts into a waveguide to quell obvious colorations is hardly a reason to cite him as a "better" than many of the actual designers/builders on this forum.

Jeff pretty much spelled out the quirky nature of Geddes. Much of his bluster seems to be to make himself to be more of an expert than he really is. Dismissing baffle diffraction and room "gain" are just a couple of the obvious blunders in the guy's audio character. He is basically saying Allison didn't know what he was talking about.

His statement about THD measurements being meaningless is something that cannot be carried into a general discussion of driver evaluation, something about which even you agree. Simple THD measurements certainly don't tell you everything about how a driver actually sounds under music conditions, but for someone well versed in speaker design and implementation it is also not something that is meaningless.

I'll take Jeff B's advice over Geddes 100% of the time based on the sound of the actual delivered product.

+1. I've listened to and built (for someone else) a few of Geddes designs. Simply put. They were horrible in every aspect. From the quality of the supplied parts, to driver integration, to overall response. Just bad. With that in mind, I question his theories and his ear. On the other hand, folks such as Jeff have proven themselves time and again. Pallas, sorry that you are on the Geddes bandwagon, which brings to question your ears, again.

[weinstro]

Re: Gedlee on distortion

Middle road. The meat of Geddes position on distortion is that measured THD does not correlate well to listener perceptions. I suspect that we use THD as a figure of merit because it is easy to measure. I'm open to the possibility that a different measure, perhaps as simple as applying a weighting mechanism to measured distortion products, would improve the correlation between measured and perceived differences.

On Geddes website, he has some materials that go into measured vs. perceived differences in some detail. But, he falls short of proposing a measurement that offers a better correlation, at least not in the writings I've seen that were free.

Regards,

Rob

[craigk]

Re: Gedlee on distortion

Pallas, sorry that you are on the Geddes bandwagon, which brings to question your ears, again.

well said. the Nathan and Harper are horrible sounding speakers. I can understand how lesser minds who are closet audiophools at heart could be mislead to believe his gibberish.