"Boutique" opamps from chinese vendors. Anyone tested?

[Walter66]

Are you talking about power amps or small signal stuff ?

My Studer uses NE OP amps from the early 1980's. Small signal stuff. But you can use the newest Burr Brown audio OPA types, it wouldn't have changed the result. Discrete beats integrated, at least for my ears.
Today, there are chip amps that include small signal and power amps. Just put them on a place to let the heat out, put a box right beside and connect a digital source and they will play music. So all is integrated today, until a few watts. Nobody takes them seriously when it comes to music audio high end systems. They just have good data, thats all. Good sound still missing, but good enough for a radio player in the car or for the workbench anyway.

 

[user 37518]

Can we agree on the fact, that the IC chip amp is and was THE cheap, quick and dirty solution the audio industry always strived for? And does it sound the best? Maybe it has good measurements (0.000001% THD), but does this mean it should sound good?

That is the problem, people who make the same argument you do switch back and forth between a quasi-scientific and a totally subjective approach. You mention that semiconductor integrated technology is not good for audio, but the basis for your argument is subjectivity. Exactly how do you quantify if something sounds good if you discard the figures of merit that are supposed to do exactly that?

Secondly, it is very convenient to criticize something when absolutely no scientific or technical arguments are being said. You can, again, switch back and forth between something like

To build an IC needs a special technology and a silicon layer is needed. The process was invented after there was a long history of building it with discrete transistors.

And:

In theory, thats correct, sir.
But, as Goethe Said:
Gray, dear friend, is all theory,
And green life's golden tree.

Very convenient, indeed.

Just ask a question: why are the majority of good amps still build discrete instead of using this low cost IC solution?
I don't necessarily have to think of old IC, this ist still the fact untill today.

That proves nothing, that is the typical ad populum fallacy "just because most people are doing it, it has to be true". The main reason why discrete is still being done for studio use is plain and simple: hype & marketing. People are still stuck with the same idea you are proposing, this started in the 60's and 70's and has perpetuated till this day. It is true that you can create a DOA that can handle larger voltages than IC's in general, but the main reason why discrete survives is mainly hype & marketing. Just hire Chris Lord Alge or whomever you want to say "this product sounds great and I use it all the time" and there you have it.

Don't get me wrong: I have no problem with my audio IC in the Studer R2R, it sounds fabulous. But did you ever hear such a Studer equipped with discrete build amps? Much better sound. Thats all that counts to me.

Exactly what is good sound? And who gets to decide? If you say that ICs measure great but sound bad, well, what metric do you propose in which what you say can be demonstrated?

Technology doesn't mean anything and has no value looked at it without an intention. And if the intention is best sound, discrete often beats IC.

Why do discrete often beats IC? If no metric satisfies you (like THD) then how can you judge discrete beats IC? because it has "better sound"? that is equivalent to saying "because I say so", that is completely subjective and means northing, really.

 

[user 37518]

but sound-wise, they are more like fast food, and of course there are very good amplifiers using ICs

Another baseless argument that means nothing, only flashy analogies and no hard evidence that demonstrates that what you are saying is even remotely true.

 

[sahib]

The advantage of ICs is that stability and distortion figures are better than discretes, but sound-wise, they are more like fast food, and of course there are very good amplifiers using ICs, but they can be more expensive than discretes.

My Studer uses NE OP amps from the early 1980's. Small signal stuff. But you can use the newest Burr Brown audio OPA types, it wouldn't have changed the result. Discrete beats integrated, at least for my ears.
Today, there are chip amps that include small signal and power amps. Just put them on a place to let the heat out, put a box right beside and connect a digital source and they will play music. So all is integrated today, until a few watts. Nobody takes them seriously when it comes to music audio high end systems. They just have good data, thats all. Good sound still missing, but good enough for a radio player in the car or for the workbench anyway.

A load of garbage.

 

[gyraf]

Perhaps a bit too harshly worded, but in essence; yes

 

[Ricardus]

My Studer uses NE OP amps from the early 1980's. Small signal stuff. But you can use the newest Burr Brown audio OPA types, it wouldn't have changed the result. Discrete beats integrated, at least for my ears.

This is of course, NONSENSE.

I am sure you can tell the difference when you know which is which, but I would bet my car that you couldn't tell the difference in a decently conducted double blind test.

 

[Newmarket]

This is of course, NONSENSE.

I am sure you can tell the difference when you know which is which, but I would bet my car that you couldn't tell the difference in a decently conducted double blind test.

What's the car
Hopefully its ECU is fully discrete

 

[Ricardus]

What's the car
Hopefully its ECU is fully discrete

I don't like to brag, but I'm rockin a 2006 Saturn!

 

[Newmarket]

From the UK In don't know what that is tbh I'm cruising Sussex in a 2009 Kia Rio myself

 

[1audio]

To build an IC needs a special technology and a silicon layer is needed. The process was invented after there was a long history of building it with discrete transistors. Yes, IC has their design inherent benefits, but there are drawbacks, too. The two technologies sound different therefore.

The IC was invented a very long time ago, 1960, when transistors were still new. (Invention of the integrated circuit - Wikipedia) Transistors were built on Silicon then (except some retro Germainium parts that were notoriously unreliable) The transistors on a modern IC can be faster, lower noise, much better matched etc. Some functions cannot be done without the IC technology. DAC's and ADC's made from discrete parts can't get to the speed and accuracy of modern ADC or DAC. Ultrastable voltage references are IC based.

However maybe the "flawless" performance possible is not what you want. Perhaps some nonlinearity in the discrete circuit or response irregularity actually has a positive euphonic quality.

I will acknowledge that you can get lower noise from carefully selected discrete parts IF you know the characteristics of your signal source. However once boosted if managed properly opamps won't compromise SNR, Distortion, Frequency response, Transient response or any other electronic aspect if your signal.

 

[abbey road d enfer]

It is clear that some optimizations are better done with discrete circuitry than with IC's.
For example, VLN peamps for low impedance sources, such as moving-coil cartridges and vibration sensors.
These require high-current gain cells drawing several dozen mA. If only due to thermal dissipation issues, this is hardly compatible with IC technology.
At the other end of the spectrum, preamps for very high impedance meters (electrometers) tend to favour discrete FET's.
And if you want to drive the 1:3 transformer loaded with 600 ohms, not many IC's can do it, so a discrete, or at least hybrid solution is required.
Power amps are a different subject. The real monomlithic solutions seem to have reached a reliable limit at about 50W. And the sound quality limitations due to thermal drift are not in line with the aspirations of HiFi enthousiasts, so yes, for high-quality analog power amps, discrete is the way.
For many other applications, which represent the dominant part of signal processing, IC's offer a perfect compromise between performance, cost and practicity.
Remember that the path from source to ear is a chain; the lack of performance of one single link ruins the rest.
A perfect preamp with a lousy power amp and magnificent speakers makesa lousy system.

 

[Newmarket]

It is clear that some optimizations are better done with discrete circuitry than with IC's.
For example, VLN peamps for low impedance sources, such as moving-coil cartridges and vibration sensors.
These require high-current gain cells drawing several dozen mA. If only due to thermal dissipation issues, this is hardly compatible with IC technology.
At the other end of the spectrum, preamps for very high impedance meters (electrometers) tend to favour discrete FET's.
And if you want to drive the 1:3 transformer loaded with 600 ohms, not many IC's can do it, so a discrete, or at least hybrid solution is required.
Power amps are a different subject. The real monomlithic solutions seem to have reached a reliable limit at about 50W. And the sound quality limitations due to thermal drift are not in line with the aspirations of HiFi enthousiasts, so yes, for high-quality analog power amps, discrete is the way.
For many other applications, which represent the dominant part of signal processing, IC's offer a perfect compromise between performance, cost and practicity.
Remember that the path from source to ear is a chain; the lack of performance of one single link ruins the rest.
A perfect preamp with a lousy power amp and magnificent speakers makesa lousy system.

Quite. Horse for courses. And the status evolves (expanding into all areas here - not only audio). I could give you chapter & verse on inclinometers (essentially damped accelerometers) but I'll spare everyone the detail. Although tbh that's more a question of dedicated ADC ICs Vs built in converters.
But the real issue here seems to be the irrational fetishisation of discrete vs integrated. When it's basically just form factor, thermal engineering etc
eg the HV electronics I do everyday can't (quite) be met with wholly IC solutions. But we use ICs where appropriate - for DC precision etc And power semis "in the loop" (with opto isolation) for the Oomph
Outcome over Process.

 

[MicUlli]

Going back to the initial question boutique opamps rom chinese vendors. Anyone tested?
No, I havent and i am not willing to do so. Because it simply makes no sense at all. Why? Every OpAmp design has to be judged against the circuit requirements. As long you do not know the circuit purpose you cannot really say that a type of OpAmp is suited or not. The OpAmp is just one of several factors that influence performance. So called DOAs are only hyped because some companies make a lot of money with it. They address the not very well educated audio jockeys willing to swap DIP8 devices for eternity

 

[pucho812]

Using chips vs discrete is still a source of discussion even in 2023?
It gets even more absurd when people tell you that a 2520 opamp made with smt parts sounds different than thru-hole parts. At the end of the day it’s all in the details and not if you went discrete or chip based.

For example on a manley 16x2 the channel side is all chip. The only tubes are in the stereo bus. So Input to a Chanel and it’s direct out for recording is all chip. It’s also centered around an ina103 running +/-24vdc. It sounds quite good and no one would be the wiser about it unless they were told.

 

[Walter66]

Well, you can see it that way. But it's just your opinion, man.
What I learned in four decades of audio: everything is audible, some to a lesser degree, some more. Of course, when changing a resistor out of hundred in the amp, its not audible. But if all are being changed or maybe 50 %, it could be audible. And so it is with discrete and IC, different technology, different sound.
When I change one tube in my amps from manufacturer a to a different one, it's audible. Same type of electron tube. So why we shouldn't hear a different tochnology being used in construction of parts? Makes no sense to me.

 

[sahib]

You hear what you hear, and I hear what I hear, so do the rest of the forum members. This is a subjective issue.

But when you say things like

............... Nobody takes them seriously when it comes to music audio high end systems. They just have good data, thats all. Good sound still missing, but good enough for a radio player in the car or for the workbench anyway.

even if you have practiced audio for 400 years I can not take you seriously.

 

[mobyd]

If DOAs are a nonsense, why the proliferation of 990 and 2520 style modules in "high end" studio equipment?. Is this merely driven by fashion?.

 

[amplexus]

If DOAs are a nonsense, why the proliferation of 990 and 2520 style modules in "high end" studio equipment?. Is this merely driven by fashion?.

One big consideration in these implementations is the ability to drive transformers. Most IC’s struggle with this and need a discrete transistor output stage to do so properly.

Obviously these days we enjoy transformers for tonal options and non linearities etc… but there are now and always have been significant advantages to having the galvanic isolation that transformers provide, and both of those reasons mean DOA’s enjoy widespread popularity.

Of course even in transformerless circuits the non linearities and “deficiencies” of a DOA design (2520’s being the classic example) are desirable for their colouration.

There is an argument to be made- and has been made- that using discrete components allows for choosing optimal devices that are purpose-built… ie: instead of capacitors, resistors, transistors, etc… all being made from the same substrate, you can use the best materials and processes to build the discrete components and so can get the “best” components from the ground up.

Imo this argument probably has the most weight when using the best of modern SMD’s and chasing a design like the 990 AND when needing to drive a transformer. Whistle Rock has done this well with their RED25 which is all high grade SMDs except the output transistors and two paired transistors.

 

[pucho812]

If DOAs are a nonsense, why the proliferation of 990 and 2520 style modules in "high end" studio equipment?. Is this merely driven by fashion?.

I would be the first to say they are not nonsense. But I would also say there is use for both. These are all design choices. The details as I like to say.

 

[pucho812]

Well, you can see it that way. But it's just your opinion, man.
What I learned in four decades of audio: everything is audible, some to a lesser degree, some more. Of course, when changing a resistor out of hundred in the amp, its not audible. But if all are being changed or maybe 50 %, it could be audible. And so it is with discrete and IC, different technology, different sound.
When I change one tube in my amps from manufacturer a to a different one, it's audible. Same type of electron tube. So why we shouldn't hear a different tochnology being used in construction of parts? Makes no sense to me.

In a blind ear test I would bet you that you would not be able to tell the difference in topologies. A deemed good mic pre is a good mic pre regardless of how it’s constructed. However, how it’s constructed can be a good selling points for the marketing department.

 

[user 37518]

Well, you can see it that way. But it's just your opinion, man.
What I learned in four decades of audio: everything is audible, some to a lesser degree, some more. Of course, when changing a resistor out of hundred in the amp, its not audible. But if all are being changed or maybe 50 %, it could be audible. And so it is with discrete and IC, different technology, different sound.
When I change one tube in my amps from manufacturer a to a different one, it's audible. Same type of electron tube. So why we shouldn't hear a different tochnology being used in construction of parts? Makes no sense to me.

Is there anything you can say that is not just anecdotal? Can you provide at least a tiny bit of demonstrable evidence?

 

[abbey road d enfer]

If DOAs are a nonsense, why the proliferation of 990 and 2520 style modules in "high end" studio equipment?. Is this merely driven by fashion?.

That may be your perception, in regard of a business that has been under many changes in the last 20 years (DAW, plug-ins,, almost infinite track count...) but when big analog behemoths mixers were ruling the studio world, not many were equipped with DOA's. The only two significant names are, IIRC, API and Quad-Eight. Most others used IC's. Actually API and QE used DOA's sparsely, only in stages where they had a proper justification. IIRC API had only two DOA's per channel, QE I don't remember but their parametric EQ couldn't have been DOA-based.
Today, the recording methods concentrate analog processing on a small number of high-quality (cost?) channels, so can dispense with the stringent quality vs. cost concern of earlier designers and users.
The cost of a few high-quality preamps, EQ's and compressors is far less than that of an SSL, Neve, Focusrite, whatever.

 

[sahib]

If DOAs are a nonsense, why the proliferation of 990 and 2520 style modules in "high end" studio equipment?. Is this merely driven by fashion?.

I do not see anybody using the word nonsense for DOAs. But what is nonsense is making unqualified and unsubstantiated comments on IC op-amps.

 

[user 37518]

I do not see anybody using the word nonsense for DOAs. But what is nonsense is making unqualified and unsubstantiated comments on IC op-amps.

I did mention that the main reason why DOA is so widespread over ICs is due to hype and marketing, but he is just making a straw man. The real issue is that we are not attacking or saying DOAs are nonsense, the real issue is that they are saying ICs sound bad or at least inferior to DOAs without providing any hard evidence whatsoever.