RandomEar,
First of all, I want to say I appreciate your thoughtful responses and the dialog.
My pleasure. Thanks for a friendly discussion.
In my head, I'm not necessarily convinced on all the arguments, but I'm certainly no engineer, just seeking to understand better how things work and questioning things when they don't make sense to me.
I know that this is AudioScienceReview and I do see plausible scientific explanations on what I am hearing.
For me, ultimately, this is a hobby and what my ears prefer is what gives me enjoyment (not sure why some users on here think anything put 100% accurate is stupid unless they also believe in not seasoning their food) despite it not measuring as accurate - so I want to learn about what might cause sound to change, even if it is adding second order harmonics, distortions and color.
I get it. Ultimately we want to appreciate well rendered music. For me, the goal of ASR is to offer people good information to approach this in an objective, bullshit-free way.
Marketing/sales...yes, could very well be. That's why I'm so interested in this topic.
Agreed.
Ok, so your condition is that is it already audibly transparent. So we are talking about the diminishing returns (or in fact no more returns after you reach audible transparency).
Yes. If I remember correctly, Amir qualifies "transparent" as SINAD better than 115 dB and +-0.5 dB from 20 Hz to 20 kHz. From my point of view, CD quality should be sufficiently transparent for almost any application - that would be 96 dB SINAD. Almost all DACs and most amps today surpass that threshold for "almost certainly transparent".
But let's say the circuit is not audibly transparent...
I do get your point about circuit stability - being somewhat binary (not your words, but how I think about it) or not, but would this depend on the overall environment and operating conditions? So it might be stable in 90% of the perfect operating conditions but maybe perform worse if one or more factors is outside of range?
Stability is mostly an example of what you could make worse by swapping op amps. There are
three types of stability for linear systems: Asymptotically stable, marginally stable (rare in practice and just as bad as unstable for audio applications) and unstable.
Stability is not really a concern for music reproduction because you won't use an unstable or marginally stable amp for that - it would immediately screech your speakers to death.
And of course this is a layman's view - but wouldn't heat, noise, voltage, current over time impact stability? If the circuit is not perfectly designed to always supply the opamp with ideal conditions, I would imagine it could impact the sound. We have PSSR, output impedence, input offset voltage, slew rate, gain bandwidth etc...
It's hard for me to imagine a system that close to the border of stability in an audio device. As noted above, stability is not something you usually need to worry about. And definitely not with prebuilt amplifiers.
The properties of semiconductor components are also guaranteed within their operating temperature windows, which are usually uncomfortably high for normal amps - 150 °C are not uncommon for mosfets and such.
An amp either has bad distortion in its default state, or it doesn't. The measurements can change slightly when heating up, but from room temp to 50 or 70 C inside the hottest chips, the change isn't order of magnitude.
Higher currents are similar: As long as you stay within the rated limits of the main chips, it's fine. Distortion over power plots tend to have that extremely steep increase close to the rated power. Before that, not much is happening.
Is it not possible to design a circuit to take advantage of an opamp better specs in these categories for better sound (assuming its not yet audibly transparent)? ...Even if that same circuit would still be stable under most conditions with an NE5532?
Maybe? Ignoring stability, I would expect that in a bad measuring amplifier the op amps are typically not the performance limiting factors. But if distortion were already audible due to bad circuit design, a different op amp might change the distortion profile.
But we are taking about "bad, bad" in this case. As in: tube-like -45 dB distortion. That's not something you will encounter in any modern transistor or chip amp.
That's also why I'm that adamant about op amp rolling being useless. This thread is about the Ampapa D1 and the 3e A5/A7. We have data for the 3e amps and it shows a SINAD of better than 100 dB for both. That's better than what audio CDs can reach (!). From my point of view, that's transparent. There would be no point in rolling for these amps. The same is true for most modern amps.
I take your point on improving things that are beyond our sensory capabilities, but I do keep hope out there that we don't know it all yet and that some cool new science/innovation will improve our experiences even further that we didn't know about. Otherwise there would not be much to look forward to.
I don't think "audio reproduction won't get any better if you select these three devices because they are essentially perfect" is a bad outlook. A bit boring maybe, for people who like to constantly change their setup. But you could still look forward to listening to all that new music that's still to come
Again - really appreciate you sharing your thoughts with me.
Cheers.
Cheers.
