I don't agree with your opinion on that.
Where's the difference between "the properties of the op amp change the sound" and "the interactions of the op amp with the circuit change the sound"? How the op amp interacts with a circuit is governed by it's properties.
Not sure you would say that...as if all circuits or part selection is infallible. If that were the case, we wouldn't need to test gear, it'd all be good.
To a degree, yes. But why would you assume that precisely the op amp is the problem? And why would changing that component alone improve the whole circuit and especially improve it to an audible degree? That's where it fails for me.
Could be many reasons:
1) To hit a price point... NE5532 might do the job a a basic level of performance, rather than design a higher performing circuit that requires more exotic parts
2) Scalability... allows them to get a product out the door that performs well, but builds in enough headroom to take performance further with opamps with better specs so that users have upgradability
Manufacturers can always solder the opamps right to the circuit board, they don't have to put in sockets to allow op amp rolling, but many do.
Why is that the case if they are not supporting/encouraging op amp rolling? Many even give you a choice of opamps to configure with their device too.
If you say they would never design a circuit with this in mind, then why do they?
Socketed op amps are 100% marketing. Look at yourself: It essentially becomes part of the audiophile hobby to change devices and "tune" them by switching components. Socketed op amps are an incentive for you to buy the device.
Soldering op amps would be more reliable and even give a miniscule performance advantage due to less contact resistance and less stray capacitance. From an engineering perspective, it does not make sense to socket op amps (outside of testing and development environments). It's a pure marketing/sales driven decision.
You're just repeating what was already in the video I am referencing. I don't think we are disagreeing on this point.
EDIT:
Which video?
If what you say is true, that for audio, an NE5532 is all we will ever need...than why do manufacturers even bother to design other opamps intended for audio use that have a broad array of specs that are better than the NE5532?
I do agree it all revolves around the design of the circuit but....how do you know in their design, they don't want to push the beyond the boundaries of what an NE5532 can do?
If it were such a commodity at such a lower price, there would be no market for better performing opamps in audio. Every manufacturer would just use the NE5532 and be done with it. Reality is many do not use the NE5332, so I'd love an explanation.
I didn't say you don't need other op amps. I said that:
A) You can design transparent devices using NE5532's.
B) If a device is audibly transparent, you cannot improve the sound any further, no matter how much better another op amp may look specs-wise.
That doesn't mean we don't need anything other than 5532's. For certain circuit designs, other op amps are more appropriate. Some may have a higher currently noise but lower voltage noise than an NE5532 and the PCB designer calculated that this would give him a performance advantage in his specific circuit. Good. Doesn't mean it changes the sound, but it will still improve the specs. Other example: A 5532 might cause instability in a circuit, while another specific op amp doesn't. Choose the stable one - good.
You can improve designs further concerning noise and distortion, even if they are already audibly transparent. Again: I didn't say there are no better op amps than NE5532 - there are. Even if they sound the same, an amp with lower THD on the spec sheet might simply sell better.
And there is even more marketing reasons. Stuff like
product differentiation is a thing: Selling a cheaper version of your amp with NE5532's for price conscious customers and a supposedly higher end version with "better" caps and op amps. Because in the audio market, superstition rules and many, many buyers think more expensive components
must sound better. It's silly, but it sells.
Yah, I don't disagree with you here on how bias works, but I am also not a fan for using it of to explain everything we can't yet measure or think we understand.
The way you talked about bias in your last post would definitely be something I wouldn't agree with. But maybe we're not that far apart on the ideas behind it.
My main problem is that a common subjective audiophile theme is that people claim to hear something and therefore it must exist / must be true. But they fail to provide any verifiably evidence. In science, the burden of proof is on the claimant - and for good reason. Since it's been proven that our senses are easily fooled, the most plausible explanation if someone claims something extraordinary is that it's just another case of fooled perception or bias. That assumption holds until any objective evidence of the contrary is provided.
Purist claim Class D was not harsh for many many years, we are not hearing what we are hearing - placebo, bias whatever.
Reality was the testing was incomplete. Didn't cover cases where load fluctuated and now we have amps with PFFB to address that and know how to measure it.
The designers knew how load dependency affected the output. This wasn't some unknown effect or couldn't be measured. Testing wasn't incomplete. To the best of my knowledge, load dependency was an accepted engineering flaw of those amps. Maybe it was a cost-driven decision, maybe PFFB wasn't understood well enough to implement it broadly, maybe there were other reasons.
I'm a believe in science, but not a believe we know how to quantify everything we do not yet understand, or that we even have a complete understanding.
There's always newer science that proves the old methods were flawed. Keep an open mind.
There is always new science. But once you reached a threshold in measuring capabilities, you don't unreach it. We can measure better than human hearing in all aspects, be it distortion, noise, phase or frequency response. And not by a little - orders of magnitude better than human hearing.
People can
hardly detect -40 dB THD in controlled testing with real music, only very few lucky people reach slightly below -50 dB. Modern measurement equipment easily detects -120 dB THD. That's over 3000 times more sensitive than human hearing. These are the scales we're talking about here when I say "measurements are better than human hearing".
There may be some complex aspects which we can measure, but not interpret or connect to human perception, yet. Mostly with speakers and headphones/IEMs. Like, what quality results in people perceiving that a speaker "vanishes" in a room? We can fully characterize the speaker, room and reflections but might not know yet which exact measured property causes that perception.
Amps and DACs are not as difficult to characterize and their interaction with other components is more limited and easier to measure and especially easier to model. Not having mechanical parts makes them way easier to predict.
