Propeller Head Plaza Technical and scientific discussion of amps, cables and other topics.

---

[ Follow Ups ] [ Post Followup ] Thread: [ Display  All  Email ] [ Propeller Head Plaza ]

---

[ Alert Moderator ]

---

Distortion and Audibility...replay.

Posted by morricab (S) on June 17, 2008 at 09:53:13

A long standing issue I have in the audio world is the lack of correlation between what people observe when listening vs. what the measurements suggest they should hear.
With loudspeakers, one can get a reasonable picture from the measurements because they vary so wildly from speaker to speaker. Things like on and off-axis frequency response, cabinet (or frame) resonance, panel or driver resonance and/or breakup, harmonic and non-harmonic distortions (like port noise or panel buzz) all add up to give a discernable character. The measurements won't necessarily tell you if you will LIKE the outcome but they will give you at least some idea where the troubles for a given speaker could be. There is at least some correlation between good sound and good measurements for speakers.

Electronics; however, are another matter. The standard measurements: THD+noise, IMD, Frequency response, S/N ratio, etc. don't tell you much at all what to expect sonically. Now during the 1960s and 70s there were the THD wars where lower and lower was considered the ideal (to some extent this never went away...look at Halcro for the latest extremely low THD example). The end result; however, was bad sound...really bad sound. Then some clever guys like Matti Otala thought there was a better correlation with transient intermodulation distortion or TIM and to some extent this was true but the outcome of some amps designed around the low TIM philosophy??? Again not appreciably better sound. Now the hardcore objectivists were screaming "Hey the distortion is so low that it is impossible to hear it...it must be your imagination if you think one amp sounds different than another". But hear it we do, so the quest goes on.

Now Otala also came up with another paper that showed that distortion can occur in an amp with a high global negative feedback from the back EMF of the loudspeaker. Essentially, the return of signal from the speaker back into the output stage and thus the feedback loop of the amplifier would give a distortion signal back to the input of the amp that woudl then be reamplified. Now this signal was small and would not contribute significantly to the THD level but it would be completely non-harmonic and therefore potentially quite audible. Still, pieces of the puzzle are missing.

Now jump ahead 20 years and what do we see as the trend, less negative feedback in many cases or none at all, simpler topologies with emphasis on linearity at each stage...things like this. Why? Well it turns out that THD and IMD are HORRIBLE predictors of sound quality. In fact, Earl Geddes has found that there is even a slight NEGATIVE correlation with THD, meaning a higher THD amp might sound better in fact. This then begs the question, is it THD that is really important then? The answer is no! BTW, if you look further below a link posted by AndyC also found another paper, which found the same thing as Geddes regarding THD and IMD. Their metric is perhaps more generally applicable but still non-trivial to implement with real amplifiers.

I have now read three interesting papers where a metric is attempted to correlate listening results with an empirically derived equation. The first was Cheever and he basically found that if one takes into account the masking effect of the human hearing then it lends itself to a model in which the HARMONICS of the distortion pattern are weighted to account for the lack of masking at high order harmonic distortion. He is dealing with the measured output of the amplification device but not the steady state harmonics the dynamically produced harmonics. It turns out that HOW an amplifier is designed has a profound effect on its behavior in this regard.

More recently Dr. Earl Geddes has come up with his Gedlee metric. This metric uses the transfer function of an amplifier to generate this Gedlee factor. Sharp discontinuities in the transfer function result in higher G values and these correlate with listening tests. The higher the value of G, the worse the sound. They were able to achieve a correlation coefficient of > 0.9. The difficulty is of course using this in practice and getting the transfer function of a piece of electronics. The links to the papers are here: http://www.gedlee.com/distortion_perception.htm

Read Distortion perception I and II

Now I have found another interesting site with sound files based on amplifier models:
http://www.silcom.com/~aludwig/Amplifier_distortion.htm

This guys test and modeling is VERY interesting as well and he has found the following:
"For these initial tests a pure 440 Hz sine wave is input to the amplifier model, and the THD of the output signal is computed.

The distortion of a pure tone is quite audible at a THD level of 0.5% for the SS model. There are nine harmonics in the range of -50 to -60 dB, and it is the higher harmonics that are the most audible For the SE model distortion is only audible at a THD of 10%, where the 1st harmonic is 20 dB below the fundamental! "

So 10% vs 0.5% THD the 0.5% should be inaudible and the 10% clearly audible, right? Wrong! The non-linearity of hearing sees to it that that simple numbers like this don't tell all.

"I have posted pure tone distortion wav files with 0.5% distortion for the SS model, and 5% for the SE model. I can't hear the distortion in the SE file even though it is 10 times higher than the SS model.
"

"The listening test results quoted here were obtained using Army's double-blind tester"

"For a jazz selection, with a value of alpha equal to 0.5, I was 100% accurate in selecting between the reference and distorted files for the SS and PP models. I could not reliably distinguish the distorted SE file, and I could not reliably distinguish between the SS and PP models"

"I was 100% accurate in selecting the SE file as less distorted than the SS file."

"Referring to the previous figure, this value of alpha corresponds to a sine wave THD of about 5%, 0.2%, and 0%, for the SE, PP, and SS models respectively. So if the output powers were really equal, a SE design that would be rated at 5% THD sounds better than a SS rated at 0% THD!
"

Apparently the less stressed an amplifier is the better as well:

Suppose the SS amp puts out 4 times the power as the SE tube amp. If you make the output volumes equal, the alpha value for the SE amp is twice as large as for the SS amp. Compared to a SS with alpha = 0.4, the SE amp with alpha = 0.8 sounds much worse. A SS with alpha = 0.2 compared to a SE with alpha = 0.4 is also better, but the difference is less dramatic. So in this situation the advantage of higher SS power outweighs the advantage of less discordant SE distortion products"

Cheever's thesis:
http://www.next-tube.com/articles.php?article=articles/Cheever/abstract_en.inc&sub_menu_item=99

Just some food for thought regarding what people think they know about electronic distortion and what is really known.

This post is made possible by the generous support of people like you and our sponsors:

Kimber Kable

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---