Does THD accurately predict good sound quality? And is subjective SQ useful to assess amps?

Haha audio will never be reproduction because of the room it will never sound like a copy,

Of all technical aspects I think thd is the least important, noise is more important, then psnrr, then assymetry, feedback stability driving various impedances, damping, and tone of parts used

That is why a simple dac sound so bad in some applications then can be made to sing with surrounded parts, but thd is all very low there
 
Only THD numbers say absolutely nothing about the sound of an amplifier. The easiest way to lower THD is applying lots of negative feedback, just like they did in the 80's and some still do today but on a much lower level than they used to back then but even then it has it's own issues like:

  • Transient intermodulation distrotion due to insufficient slew rate which lead to typical harshness or graininess even when THD measueres very low
  • Phase shifts and stability issues specially at high frequencies which leads to an edgy sound, even when THD measures almost zero
  • Emphasis on higher order distortion becuse feedback typically lower low oreder distortion more than higher order distorion which leads to the typical "clincal" or "sterile" sound
  • Overcorrection if the open loop gain isn't linear which leads to a typical "digital" sound like if there is a lot of jitter
  • Latency in correction, the time between detecting an error and corecting it is not zero. This leads to typical "dead" sound with very little resolving power in small transients
  • Load dependancy. High open loop feedback is sensitive to back emf from the loudspeaker that can cause ringing and instabilty in the amplifier

 
  • Like
Reactions: adason and gabdx
  • Transient intermodulation distrotion due to insufficient slew rate which lead to typical harshness or graininess even when THD measueres very low
  • Phase shifts and stability issues specially at high frequencies which leads to an edgy sound, even when THD measures almost zero
  • Emphasis on higher order distortion becuse feedback typically lower low oreder distortion more than higher order distorion which leads to the typical "clincal" or "sterile" sound
  • Overcorrection if the open loop gain isn't linear which leads to a typical "digital" sound like if there is a lot of jitter
  • Latency in correction, the time between detecting an error and corecting it is not zero. This leads to typical "dead" sound with very little resolving power in small transients
  • Load dependancy. High open loop feedback is sensitive to back emf from the loudspeaker that can cause ringing and instabilty in the amplifier

If amplifier designer cannot achieve sufficient slew rate, then he/she is not capable to designing amplifier. In this forum, there are several design amplifier that have very low distortion and high slew rate. Even, the distortion is very low at 20 Hz to 20 kHz and all level below clipping.

Insufficient slew rate cause is not because of negative feedback. It is very easy to design amplifier with high slew rate, the power bandwidth higher than 200 kHz with low distortion.
 
Only THD numbers say absolutely nothing about the sound of an amplifier. The easiest way to lower THD is applying lots of negative feedback, just like they did in the 80's and some still do today but on a much lower level than they used to back then but even then it has it's own issues like:

  • Transient intermodulation distrotion due to insufficient slew rate which lead to typical harshness or graininess even when THD measueres very low
  • Phase shifts and stability issues specially at high frequencies which leads to an edgy sound, even when THD measures almost zero
  • Emphasis on higher order distortion becuse feedback typically lower low oreder distortion more than higher order distorion which leads to the typical "clincal" or "sterile" sound
  • Overcorrection if the open loop gain isn't linear which leads to a typical "digital" sound like if there is a lot of jitter
  • Latency in correction, the time between detecting an error and corecting it is not zero. This leads to typical "dead" sound with very little resolving power in small transients
  • Load dependancy. High open loop feedback is sensitive to back emf from the loudspeaker that can cause ringing and instabilty in the amplifier

You just describe a very bad amplifier design!

NFB doesn’t pose a problem with any decent amp design. What objections do you have on Bruno Putzey’s article “The F word” or Jan Didden’s article “Feedback Cannot Fix What Has Already Happened”?.

The F word
Audio Myths: Feedback Cannot Fix What Has Already Happened
 
  • Like
Reactions: twister1264
Example: common "blameless" amplifier topology, the slew rate depend on collector current of the LTP and the compensation. To reduce distortion on high frequency, you can use 2 pole or more pole compensation. Use must use VAS transistor with very low Cob so you can use "light" compensation. It explained at Bob Cordell book and Douglas Self book very detail. But if you want more advance topic with conceptual explanation, I suggest you to read book from Dr. Arto Kolinummi.
 
  • Like
Reactions: abstract
If amplifier designer cannot achieve sufficient slew rate, then he/she is not capable to designing amplifier. In this forum, there are several design amplifier that have very low distortion and high slew rate. Even, the distortion is very low at 20 Hz to 20 kHz and all level below clipping.

Insufficient slew rate cause is not because of negative feedback. It is very easy to design amplifier with high slew rate, the power bandwidth higher than 200 kHz with low distortion.
Yes, that's absolutely right. All I meant to say is that THD figures alone don't say much about how an amplifier actually sounds.
 
  • Like
Reactions: gabdx
You just describe a very bad amplifier design!

NFB doesn’t pose a problem with any decent amp design. What objections do you have on Bruno Putzey’s article “The F word” or Jan Didden’s article “Feedback Cannot Fix What Has Already Happened”?.

The F word
Audio Myths: Feedback Cannot Fix What Has Already Happened
And this is very corect as well but like my previous answer, all I meant to say is that THD figures alone don't say anything about the sound of an ampligier. And yes, there still are many very bad designs out there that measure very low THD but sound like crap.