OPamps still inferior to discrete circuits?

Casouza - Jeff Rowland uses, in great sounding Capri linestage, fully differential high speed op-amps OPA1632.

OPA1632 has:
0.00002% THD
0.00005% IMD
1.3nV/SQRT(Hz) noise
180MHz bandwidth

discrete amp like that is not a trivial task (layout!). Can you imagine the cost?

Kijanki, I understand the technical difficulties of making discrete layouts and the costs that it will generate. But we are audiophiles and we are only interested in sound quality, not technical specifications. So do you think the best OPamps with the best specifications ever sound as good (= musical) as the best discrete layouts? Or even better?

Chris

I have not tried Burson discrete op amps, but IME the sonic performance of monolithic op amps is very much dependent upon the quality of the power source. Typically the same build-to-cost considerations that lead manufacturers to use inexpensive op amps, also leads them to cheap out in the power supply. Therefore few people hear op amps at their best. Executing op amps together with batteries and well engineered discrete DC regulation circuits is a different animal. And if this is done correctly, then monolithic op amps may have the advantage insofar as there less is less metal, solder joints, etc. to add degredation. Moreover, at Burson's price point it is unlikely that he uses top-quality piece parts.

Chris - This is difficult to answer. Example of Capri shows what is possible. The real question is what you can buy for the difference. It might be better to buy Capri for $3,500 instead of discrete preamp for $10k and spend $7k more on speakers. Benchmark DAC1 praised by many uses $0.50 opamps (NE5532).
I agree with you that it might be better to get discrete circuit if money is no object but even then you might not beat opamps in offset drift to eliminate caps from the signal path.

Problem with opamps is that they are not designed in general for the audio. First opamp really designed for audio was NE5532. It was about 25 years ago and Benchmark claims that it performs better than OPA627. Now National Semi. came with LM4562 but many claim that NE5532 has better midrange.

I was surprised with Rowlands selection of 180MHz amplifier and even more surprised when I checked datasheet - It is an audio amp. It has 180 MHz bandwidth but its full power bandwidth is only 800kHz (Capri swings 7V rms = 20Vp-p). Distortions start to climb-up starting from 10kHz so it was probably necessary to spike-up bandwidth to avoid distortions and any phase shift at 20kHz.

Dgarretson - Jeff Rowland used batteries in his preamps but now he "switched" to switching power supply (SMPS) in Capri. It should be easier to filter out high frequency than 120Hz and also to make stronger supply within same dimensions (and SMPS are load and line regulated) but unfortunately manufacturers use SMPS technology to make it cheap instead of better.

Somebody said that there are two types of fools in this world: One says "it is old and therefore good" while the other says "this is new and therefore better"