Replace OP amps. Completely pointless, or not?

[antcollinet]

There are engineers (one in this thread) that would apparently disagree to some extent.

Then either they are not engineers who understand the application of op amps - or they need to come forward, and explain in engineering terms what they think is happening. Perhaps they should back that up with engineering measurements, to show measured improvements at audible levels.

 

[cgallery]

Then either they are not engineers who understand the application of op amps - or they need to come forward, and explain in engineering terms what they think is happening. Perhaps they should back that up with engineering measurements, to show measured improvements at audible levels.

Wouldn't one of them simply point to a study like this:

AES Paper Digest: Do Audio Op-amps Sound Different?

This is a digest of the Audio Engineering Society (AES) conference paper titled, “Distortions in Audio Op-Amps and Their Effect on Listener Perception of Character and Quality.” (http://www.aes.org/e-lib/browse.cfm?elib=16029) As the name indicates, it is a controlled listening test to see if...

www.audiosciencereview.com

While the differences were slight, the op-amps in this test were capable of producing levels of distortion within their normal, unclipped operational range that were detectable by listeners. It is possible for listeners to differentiate between op-amps based on distortion characteristics alone. The ability to distinguish between the distortion characteristics of op-amps appears to be dependent on a complex interaction of the THD vs. Frequency, THD vs. Amplitude, and the harmonic spectrum of the distortion. This low-level distortion is enough to alter the character of the sound but does not necessarily reduce listener preference.

 

[Old_School_Brad]

Wouldn't one of them simply point to a study like this:

AES Paper Digest: Do Audio Op-amps Sound Different?

This is a digest of the Audio Engineering Society (AES) conference paper titled, “Distortions in Audio Op-Amps and Their Effect on Listener Perception of Character and Quality.” (http://www.aes.org/e-lib/browse.cfm?elib=16029) As the name indicates, it is a controlled listening test to see if...

www.audiosciencereview.com

While the differences were slight, the op-amps in this test were capable of producing levels of distortion within their normal, unclipped operational range that were detectable by listeners. It is possible for listeners to differentiate between op-amps based on distortion characteristics alone. The ability to distinguish between the distortion characteristics of op-amps appears to be dependent on a complex interaction of the THD vs. Frequency, THD vs. Amplitude, and the harmonic spectrum of the distortion. This low-level distortion is enough to alter the character of the sound but does not necessarily reduce listener preference.

I believe this test doesn't reflect a real-world scenario, making the conclusion unreliable. They routed the signal through four op-amps before listening. And didn't disclose the used op-amps.

Notice that they ran the signal through two of the fixtures shown before. So the total picture is the op-amp being in the signal path four times. It is not explained why they had to do this.

There is a lot of fine print here of course with quadrupling the use of op-amps

 

[solderdude]

swapable opamps usually are -1 or +1x gain buffer stages so have huge bandwidths and very low distortion.
As long as you don't load them with unexpected low impedances ( < 600 ohm for most op-amps) they will all perform just fine.
Sure there will be measurable differences in noise floor, spectrum and HD and IM distortion but none of them will reach the levels they created in the described test.

 

[Cbdb2]

This test is not really applicable to home audio. The paper says its the distortion of the peaks that is audible. The peak levels in the tests were +22dbu rms or 14V peak (approaching many OA limits), not sure what the load was. In studio gear (where the nominal level is 10db hotter) this could definitely be a consideration. But when I think about all the OAs (before the 90s almost all NE5532s) the signal goes thru in music production and it can still sound great does it really matter? Besides, once the distortion from the peak limiting to the max is added the OAs distortion is completely inaudible.

 

[antcollinet]

swapable opamps usually are -1 or +1x gain buffer stages so have huge bandwidths and very low distortion.
As long as you don't load them with unexpected low impedances ( < 600 ohm for most op-amps) they will all perform just fine.
Sure there will be measurable differences in noise floor, spectrum and HD and IM distortion but none of them will reach the levels they created in the described test.

Yes, and as can be seen from the circuit - the audio was put through 2 40dB (x100) (distortion) gain stages plus a 9dB buffer and put through that circuit twice.

Other problems with the study was that the circuit rolled off the FR in the audible range (starting at 10kHz) and differently for each op amp, Also the fact that the op-amps tested were not disclosed in the paper. (are they - for example - the types typically used in and appropriate for audio applications) All of this is discussed at the link provided above.

It obviously is not applicable to the line level circuits at or close to unity gain, and with much wider bandwidth, as used in DACS and Amps where op amp rolling is typically carried out. It might be applicable to high gain circuits such as phono pre-amps. However, these rarely have more than one gain stage, let alone 4, and in good designs they will always use op amps already optimised for the application.

Wouldn't one of them simply point to a study like this:

They might - but that would be another indication they are not understanding the application.

 

[DanielT]

We have a pretty decent protocol for phase one. Saturday is the day. I just hope the connectors (RCA and especially banana) are up to this, the connectors are a bit on the flimsy side.

If you are going to plug in and out RCA connectors quite often in connection with changing your two A07 in your blind test, it may make sense to have lockable RCA connectors. The kind you screw to get a suitable hard pressure around the RCA female, not too hard not too loose. If you have the time and desire to fix such, that is. Plus you need to solder the cables to them, but then on the other hand you now get exactly the length (as short as possible) that suits you.


What kind of HF duty is it that your A07 takes care of? Just tweeters? I was thinking about this with load dependence, as the A07 is.

Blind test A07 vs non load dependent amp. There is actually a possibility that you can detect differences in such a test (depending on what your A07 is power up). A more complicated test to carry out but I mention it nonetheless.

Speaking of blind tests.Basic (cheap) system (A) vs High End system (B). Do you think they hear a difference?:

Matrix HiFi --> Blind testing high end full equipments

 

[solderdude]

A downside to those lockable ones is that one might accidentally break the sockets when removing them and forgetting to unlock which makes them not really suited when needing to do some fast cable swapping.

 

[DanielT]

A downside to those lockable ones is that one might accidentally break the sockets when removing them and forgetting to unlock which makes them not really suited when needing to do some fast cable swapping.

You have a point. I didn't think of that.Since there will be many and quick swapping, RCA connectors/cables that are well suited for that purpose together with cgallery's HiFi equipment.

 

[cgallery]

The objections to the test cited are ringing a little hollow. Where I come from, we'd call this a game of "you can't win." That is, if someone demonstrates audible differences in op amps (each working within specification and none of them driven to clipping), the goal posts are immediately adjusted.

While I acknowledge the criticisms, not a single card-carrying member of the "there are no audible differences in op amps" association acknowledged that the subjects had minimal time to become acclimated with the gear, nor the sheer number of challenges involved. This was not a test conducted using the personal gear of the subjects, so there was no familiarity (which may have lowered the threshold of detection).

For both sides ("I can hear a difference" and "no you can't, nobody ever can"), I'm afraid this has become a bit of religion.

 

[DanielT]

The objections to the test cited are ringing a little hollow. Where I come from, we'd call this a game of "you can't win." That is, if someone demonstrates audible differences in op amps (each working within specification and none of them driven to clipping), the goal posts are immediately adjusted.

While I acknowledge the criticisms, not a single card-carrying member of the "there are no audible differences in op amps" association acknowledged that the subjects had minimal time to become acclimated with the gear, nor the sheer number of challenges involved. This was not a test conducted using the personal gear of the subjects, so there was no familiarity (which may have lowered the threshold of detection).

For both sides ("I can hear a difference" and "no you can't, nobody ever can"), I'm afraid this has become a bit of religion.

One thing that you have to be aware of is the bias that you can have if you carry out a blind test. If I were to do a blind test with the notion that I will NOT hear any differences, I most likely won't. How the case would have been if I had done such a test with a different mental attitude, what do I know.

Or wait now I remember. A few years ago I tested, compared blindly a Blu-ray, Sony BDP-S570 vs and a CD player, Marantz CD5001. I tried to implement it without preconceived notions, bias. As best as I could with my rudimentary knowledge regarding such a blind test.
Results: I heard no difference. Incidentally, in a way implemented like you will do in terms of level matching. I never got to round two regarding measuring levels:

What is this? No golden ears!

Got an idea last week. Connected a Blu-ray, Sony BDP-S570, and a CD- player Marantz CD5001 to the pre amp. It has line selectors. Went to the local flea market and managed to find duplicates, CDs. Exactly the same CDs, the same recording on the discs. Same master. Enter with the discs in the...

audiosciencereview.com

 

[antcollinet]

The objections to the test cited are ringing a little hollow. Where I come from, we'd call this a game of "you can't win." That is, if someone demonstrates audible differences in op amps (each working within specification and none of them driven to clipping), the goal posts are immediately adjusted.

While I acknowledge the criticisms, not a single card-carrying member of the "there are no audible differences in op amps" association acknowledged that the subjects had minimal time to become acclimated with the gear, nor the sheer number of challenges involved. This was not a test conducted using the personal gear of the subjects, so there was no familiarity (which may have lowered the threshold of detection).

Well they've demonstrated a (very slight) audible difference in a circuit specifically created to amplify distortion by 80dB and put it through that circuit twice. So 160dB distortion amplification. If anything this proves the opposite of what you are expecing. A very slight audible difference with 160dB gain for distortion. How audible is the distortion if it is 50 million times smaller?

If you want to think that is representative of line level buffers with typical gains in the region of 0dB to 6dB - with no specific extra distortion gain - then you go right ahead.

On the other hand, ff you want to provide engineering data, or a study that shows the sort of changes you are proposing can be audible, then you need to do that with representative data generated in a circuit with similar characteristics to the circuit you are modifying. Better still, using the actual circuit.

Not with something that bears no relationship at all to what you are using.

For both sides ("I can hear a difference" and "no you can't, nobody ever can"), I'm afraid this has become a bit of religion.

Except one side is presenting engineering knowhow and test data, and the other side is presenting listening impressions without proper controls. Which side sounds more like it's operating on the basis of faith?

 

[antcollinet]

If I were to do a blind test with the notion that I will NOT hear any differences, I most likely won't. How the case would have been if I had done such a test with a different mental attitude, what do I know.

This comes under the catagory of "you can't prove a negative"

It is not possible with listening tests to prove there is no difference to be heard. Only that "I can't hear it"

Only people who believe they are able to hear something - are in a position to prove it with a blind test.

 

[NTK]

One thing that you have to be aware of is the bias that you can have if you carry out a blind test. If I were to do a blind test with the notion that I will NOT hear any differences, I most likely won't. How the case would have been if I had done such a test with a different mental attitude, what do I know.

Proper experiment design should include positive and negative controls.

Positive Control vs Negative Control: Differences & Examples

A positive control is designed to confirm a known response in an experimental design, while a negative control ensures there's no effect, serving as a baseline for comparison. The two terms are defined as below:

helpfulprofessor.com

 

[cgallery]

Well they've demonstrated a (very slight) audible difference in a circuit specifically created to amplify distortion by 80dB and put it through that circuit twice. So 160dB distortion amplification. If anything this proves the opposite of what you are expecing. A very slight audible difference with 160dB gain for distortion. How audible is the distortion if it is 50 million times smaller?

If you want to think that is representative of line level buffers with typical gains in the region of 0dB to 6dB - with no specific extra distortion gain - then you go right ahead.

On the other hand, ff you want to provide engineering data, or a study that shows the sort of changes you are proposing can be audible, then you need to do that with representative data generated in a circuit with similar characteristics to the circuit you are modifying. Better still, using the actual circuit.

Not with something that bears no relationship at all to what you are using.

Except one side is presenting engineering knowhow and test data, and the other side is presenting listening impressions without proper controls. Which side sounds more like it's operating on the basis of faith?

Right, you've said all this. The problem is, you didn't mention any of it until confronted with a study demonstrating that properly selected/functioning op amps that aren't driven to clipping (and still presenting some impressive overall distortion characteristics) can still be differentiated rather quickly by experienced listeners.

 

[antcollinet]

Right, you've said all this. The problem is, you didn't mention any of it until confronted with a study demonstrating that properly selected/functioning op amps that aren't driven to clipping (and still presenting some impressive overall distortion characteristics) can still be differentiated rather quickly by experienced listeners.

No, you're right. I didn't mention a study I was not aware of. Also bear in mind : we are specifically talking about line level low gain circuits. We always are when talking about op amp rolling. So even if I had been aware of the contents of that study, I probably wouldn't have brought it up, as not being relevant to the circuits at issue.

And again they could only determine "rather quickly" slight audible differences with massive amplification of distortion - I don't know why you keep referring back to that study. I'll make you a deal, you stop doing that, and I'll stop repeating myself regarding why it is not relevant..

I have in the past stated that it is possible (though I've also not seen any specific tests or measurements) that audible improvements might be made in high gain circuits such as phono preamps. Probably then only by using a lower noise device - but that is just opinion and guesswork.

 

[Cbdb2]

The objections to the test cited are ringing a little hollow. Where I come from, we'd call this a game of "you can't win." That is, if someone demonstrates audible differences in op amps (each working within specification and none of them driven to clipping), the goal posts are immediately adjusted.

While I acknowledge the criticisms, not a single card-carrying member of the "there are no audible differences in op amps" association acknowledged that the subjects had minimal time to become acclimated with the gear, nor the sheer number of challenges involved. This was not a test conducted using the personal gear of the subjects, so there was no familiarity (which may have lowered the threshold of detection).

For both sides ("I can hear a difference" and "no you can't, nobody ever can"), I'm afraid this has become a bit of religion.

What we are saying is that there may be audible differences if you make the test hard enough; lots of gain, large signals, multiple OAs, large loads, but not if you use OAs they way they are used in home audio systems. This is only applicable to music production, as in mic prees and other analog devices which are getting rare.

 

[solderdude]

replacing op-amps is beneficial for certain:
A: it keeps the amp owner busy.engaged.
B: op-amp sellers and manufacturers have extra income.
C: People love to read about the sound differences they hear.
D: People love to tell others their findings (especially the ones that don't believe)
E: The owner is happy with the improved sound (regardless what it is caused by)
F: People like to engage on the web about this.
G: threads about it keep on going and giving
H: People may feel validated or are annoyed
I: manufacturers of amps with sockets are selling more stuff as swappers love this stuff especially when the manufacturers use it as a selling point with audible benefits.
J: It is fun an easy to do and gives the owner the idea they are being 'technical'.
K: troll heaven
L: people can get a certain fame for their 'research'

etc..

furthermore ... blind tests are mainly to prove something to yourself but should be done correctly. The last part is the difficult part.
To prove something to others (in a scientific sound way) is a totally different ballgame.

 

[Killingbeans]

The objections to the test cited are ringing a little hollow. Where I come from, we'd call this a game of "you can't win." That is, if someone demonstrates audible differences in op amps (each working within specification and none of them driven to clipping), the goal posts are immediately adjusted.

There's no such thing as 100%, but everybody would like the certainty to get as close a possible.

All you really need to do is present evidence strong enough to make the people rejecting it seem just as nutty, as the ones who blindly accepted the baseless claim.

I does require a rather large amount of rigour, but once you reach that point, whatever goal posts people demand to be adjusted won't really matter much.

 

[Cbdb2]

An interesting result was that preference was random. Higher measured distortion did not mean lower preference. So rolling OAs even if (a big if) there's an audible difference doesn't mean you'll pick the better performing, more linear one.