The audio IC Opamp is dead and these guys claim they killed it

[rsjsouza]

Quite suspicious claims of "high open loop gain", "improved class A performance", "better pair matching than ICs"... Dodgy marketing perhaps?

https://sparkoslabs.com/discrete-op-amps/

What are the better folks experienced in audio around here can tell about this? An interesting set of comments on Facebook is attached.

[Kleinstein]

The give typical specs for +-12 V - high grade audio more like needs more swing and thus more like +-15 or +-18 V supply. While possible too, with the high current consumption the discrete OPs would run way to hot to be comfortable, especially the diode version. Who cares about low noise when you need a powerful fan. 

It looks like the input current and thus also current noise is high, so they kind of need a low impedance FB network and thus quite some extra power to drive the FB network. This makes the thermal problem even worse.

With a relatively high +-3.5 V lost to both sides the normal +-15 V supply may be to low and a directly drop in replacement may not work.  Another point here is the 2 pole compensation - this can work in some circuit, but it can fail in others (e.g. high closed loop gain - those cases where the high open loop gain may be needed). The limited bandwidth does not compensate for complications due to the 2 pole compensation.

Due to different source impedance there is no OP that fits all needs.

[WattsThat]

...All of this translates into a more detailed and engaging listening experience with better imaging and soundstaging than IC op amps can deliver. In short, discrete designs are the best op amp for audio.

Whenever I see the term “soundstage”, i’m done with it, it’s complete wankery.

The crowning jewel in their line-up has got to be the magic woo drop-in discreet voltage regulator. It’s amazing the different ways people can be enticed to dispose of their (hopefully) excess monies.

[magic]

Long ago in a galaxy far away, Samuel Groner published distortion measurements (and noise specs) of a few dozen opamps, including discrete. Some of them did very well against ICs (except for price), others were gimmicks.

If by "two pole compensation" they mean the same thing as described in Self's power amp book, their cricuit will oscillate like mad at closed loop gains above some threshold. That's why general purpose opamps don't use such compensation.

By "high open loop gain" they probably mean higher OLG at audio frequencies thanks to TPC. This is legit.

[Haenk]

Burson has discrete "drop-in-replacement" OP-Amps as well, tzhis just FYI:

https://www.bursonaudio.com/products/supreme-sound-opamp-v6/

Not sure if there is a point or not in installing these...

[Cyberdragon]

Burson has discrete "drop-in-replacement" OP-Amps as well, tzhis just FYI:

https://www.bursonaudio.com/products/supreme-sound-opamp-v6/

Not sure if there is a point or not in installing these...

Nah, you gotta go all the way and get these...

[Circlotron]

And what about the audio horrors caused by all those opamps upstream in the recording studio? I bet it's only a matter of time before someone sells a cure for that too.

[DimitriP]

And what about the audio horrors caused by all those opamps upstream in the recording studio? I bet it's only a matter of time before someone sells a cure for that too.

Shhhh     

[Haenk]

audio horrors caused by all those opamps upstream in the recording studio

Master production is rarely done analog, for many years now (became affordable 20 years ago). So there is a good chance the overall production (mastering) sound quality is better than your home equipment can reproduce, which means there is room for improvement. (usually going with 24bit resolution, then downmixing - however that does not help with "noisy" instruments, like old synthesizers and only works on analog stuff with good ADCs...)

[dzseki]

And what about the audio horrors caused by all those opamps upstream in the recording studio? I bet it's only a matter of time before someone sells a cure for that too.

Also consider that recording is happening in parallell fashion, eg. each instrument is recorded in its own channel. Instruments usualy produce rather simple and predictible spectra. Things get difficult (audiowise) when you start to mix together different insturments. Even thinking analog at that stage there are no tons of opamps in the sound chain anymore.

[magic]

They actually have a datasheet.
https://sparkoslabs.com/wp-content/uploads/2018/03/SS3601_SS3602.pdf

Input current is 4.5µA, so 1.2pA/rtHz shot noise.

130dB at 1kHz is unmatched by any unity gain stable IC opamp and few can touch 90dB at 10kHz. But phase margin at 70~90dB closed loop gains is rather thin.