Taking into account cheapness of these opamps I would like to test them before going for discrete, could you reply to the following?
1. Did you personally compare the sound of V6s/SS3602 to OPA1612/1656 in any audio gear?
2. Is OPA1612/1656 better (as per specs or on listening tests) than OPA627 I have now?
3. Are 1612/1656 compatible with dual OPA627s I have in my DAC (like these
https://www.mav-audio.com/base/archives/643 )?
4. What are your thoughts on OPA1622?
Sorry I can't compare the sound of these opamps. For several reasons: Firstly I don't think that would be difference in sound quality. Even when it would it is really small and you can't say which is better.
The main reason I can't perform the test is because the current setups of mine are all soic8 package. I may compare opa1612 and opa1656 but can't do burson v6(which I have) and sparkos(which I don't have).
I did sever tests during past months. The results aren't consistent. So don't take it too seriously.
The first time I did test with opa1642 opa2156(precision opamp version of opa1656) opa1612. Both fet based opamps sounds a little bit more solid in the bass(not quantity but the imaging and the texture). Opa1642 sounds a bit smoother and silky at top end but a little bit fat in the mids. Opa2156 sounds a little more "detailed" and a little "grainy" in comparison. Opa1612 is generally slightly light weighted but very balanced otherwise. Smooth and natural. But just the bass sounds a bit far away in comparison. And opa1642 has a bit better soundstage I don't know why.
The second time I did test with opa1656 and opa1612. I can't really hear difference this time. Probably the circuit is better than the first one or something went wrong or I was just imaging all of those. I can't say for sure. The opa1656 this time is smoother than the opa2156 in previous test and opa1612 has better bass than previous.
The third time I added in njm4585 very cheap but pretty good opamp in the comparison. There is basically no difference between njm4585 opa1612 and opa1656. If I'd imagine, the njm4585 is a tiny little sticky sounding in the highs and less clean. And opa1656 and opa1612 doesn't sound any different.
Well, if I was imaging all those above, then i was. If If I wasn't imaging all of those, the possible change is the circuit layout and the final performance, including crosstalk, distortion, PS noise etc. A good peripheral design is crucial. The opamp itself is only part of the equation. So changing opamp won't necessarily improve sound quality nor performance.
Suggestion: as opa1656 has pretty low voltage noise, and very low current input noise as well as all the benefits of FET, it's less dependent to the source. And the high output current capacity makes it suitable for much more applications. So I suggest just getting opa1656 and enjoy music. Otherwise I think a overall better design complete equipment would be a better choice.
Side note: opa1622, my experiences with opa1622 wasn't all happy. The stability is hard to achieve on my end because the pole zero compensation technique utilized. Itself is more stable than without the compensation but it is much more limited for further improvement with supporting circuit. It's an ok drop in opamp for most applications: low noise low distortion, high output current, low output impedance. But I think I prefer opa1656 for the most part.