Calling all EE's, another subjective amp quality thread!

[ski2xblack]

Bring this dead horse back to life for more thrashing!

Regarding the whole subjective amp quality question, I was wondering if any of you had read this:

http://www.its.caltech.edu/~musiclab/feedback-paper-acrobat.pdf

Just to warn you, this is in the flavor of "prevailing measures are inadequate at describing an amplifiers subjective sound quality".

Here is another about coming up with specific metrics to guage this:

http://milbert.com/Files/articles/DanCheever.pdf

Who else out there in audioholic land agrees that it's time to combine engineering and human perception for a new measurement method to describe amp quality? Or are the authors of these papers just a bunch of biased tube jockeys trying to justify their addictions?

Let the discussion begin...er, continue. (I sincerely hope that this is enough of a tangent to be thread-worthy. It seems there is a bit of interest, given the rapid growth of the most recent 'do amps all sound the same' thread. The papers at the links are long, but fascinating reading. Lemme know what you think.)

 

[mtrycrafts]

Bring this dead horse back to life for more thrashing!

Regarding the whole subjective amp quality question, I was wondering if any of you had read this:

http://www.its.caltech.edu/~musiclab/feedback-paper-acrobat.pdf

Just to warn you, this is in the flavor of "prevailing measures are inadequate at describing an amplifiers subjective sound quality".

Here is another about coming up with specific metrics to guage this:

http://milbert.com/Files/articles/DanCheever.pdf

Who else out there in audioholic land agrees that it's time to combine engineering and human perception for a new measurement method to describe amp quality? Or are the authors of these papers just a bunch of biased tube jockeys trying to justify their addictions?

Let the discussion begin...er, continue. (I sincerely hope that this is enough of a tangent to be thread-worthy. It seems there is a bit of interest, given the rapid growth of the most recent 'do amps all sound the same' thread. The papers at the links are long, but fascinating reading. Lemme know what you think.)

Of course there isn't measurements to correlate with subjective evaluations. By its very nature it is subjective.
Boyk may be a fringe type of professor especially with his attempt to convince about ultrasonics which he has not done in a credible manner.

It would be an interesting exercise to have Doug Self comment on this amp paper of his.
http://www.aqpl43.dsl.pipex.com/ampins/dipa/dipa.htm
The second paper is in the same boat. Has no real meaning. Telling that it was a masters paper but, it doesn't have to be accurate to pass

 

[ski2xblack]

The second paper is in the same boat. Has no real meaning. Telling that it was a masters paper but, it doesn't have to be accurate to pass

It also contains some spelling and syntax errors, which, if I were among the reviewing profs, would never fly.

Even if more research needs to be done, the hypothesis that the non-linearities inherent in human hearing could mask amplifier distortion is at least plausible, at least to those who have actually heard no-NFB designs and find them subjectively superior. I also find compelling the idea that higher order distortion products in high negative-feedback circuits negatively influence the listener's perception of low level detail.

It would be nice to see more perceptual research, kind of like what Earl Geddes has done regarding speaker distortion. I didn't notice anything more recent regarding Total Aural Disconsonance or Cheevers in my brief visit to Google U. I'll check out the link you posted...

There is one group of specifications I would like to see from amp/recv manufacturers which may better correlate to subjective sound quality, and that would be low level performance. You typically only see specs at rated power. It seems that almost every modern amp (class AB, some feedback in the circuit, and fantastic bench test results) have more distortion at 1 watt, more at .01 watt, and even more at .001 watt than they do at rated power. The difference can be 20-30db or more than rated distortion at these low power levels. Some single ended no-NFB amps have comparably lower distortion at these lower power levels, quickly ramping up as power increases. Could the perceived differences be simply a reflection of this? (Who knows at this point, but I'm glad I have both types of toys to play with.) I typically listen at 65-95db levels, where my amps are just loafing along, not at 125db pushing rated power, and I suspect most of us who cherish our hearing do the same. Being able to compare low level performance could possibly be quite relevant for judging an amplifier's sound quality, at least for fairly typical home use.

Recordings that don't suck would help. Stop the brutal compression already!

 

[mtrycrafts]

...Even if more research needs to be done, the hypothesis that the non-linearities inherent in human hearing could mask amplifier distortion is at least plausible, at least to those who have actually heard no-NFB designs and find them subjectively superior.

Subjectively? With no bias controls in place? If so, then, that result is suspect. I have not seen any credible DBT results comparing such amps, so far. But then, I don't think I have seen all of them.
Also, if there are those nonlinearities, that should mask equally others too, no?

... I also find compelling the idea that higher order distortion products in high negative-feedback circuits negatively influence the listener's perception of low level detail.

Maybe, maybe not, depending on the level of those high orders.
http://www.mastersonaudio.com/features/20040401.htm
Axiom has done some research into distortion too:
http://www.axiomaudio.com/distortion.html#
A good read and interesting.

... There is one group of specifications I would like to see from amp/recv manufacturers which may better correlate to subjective sound quality, and that would be low level performance. You typically only see specs at rated power. It seems that almost every modern amp (class AB, some feedback in the circuit, and fantastic bench test results) have more distortion at 1 watt, more at .01 watt, and even more at .001 watt than they do at rated power. The difference can be 20-30db or more than rated distortion at these low power levels. Some single ended no-NFB amps have comparably lower distortion at these lower power levels, quickly ramping up as power increases. Could the perceived differences be simply a reflection of this? (Who knows at this point, but I'm glad I have both types of toys to play with.) I typically listen at 65-95db levels, where my amps are just loafing along, not at 125db pushing rated power, and I suspect most of us who cherish our hearing do the same. Being able to compare low level performance could possibly be quite relevant for judging an amplifier's sound quality, at least for fairly typical home use.
...

Yes, but most if not all DBTs that have been published are at normal listening levels that would be in your ballpark. Some specs do list 1 watt distortions and the above links on distortion audibility would apply equally at low levels, I am sure.

 

[M Code]

Why debate sonic quality..
Buy a Bose wave-radio and enjoy.

Just my $0.00...

 

[jliedeka]

I think the negative feedback issue is what I would call audiophile lore. Early designs weren't that well executed in modern terms, especially in the early 70s. That's no longer true but persists because people have better things to do that keep up with developments in engineering.

A similar phenomenon is the belief that tube amps sound better than solid state. Early transistor based amps were more or less tube amps with transistors replacing tubes. After a few years, designers learned how to design with/for transistors and produced better amps.

I suspect that tube amp fans also like something they can fiddle with as opposed to a black box they just hook up and turn on. That probably also explains the enduring love for vinyl.

Jim