I ask because it sounds like prefer using a circuit favoring your desired outcome of the experiment. But when people say op-amps have a sound, they probably are not usually talking about a single stage line amplifier with a gain of 10. They are probably taking about set of varied circuits in which there is a characteristic to the sound that appears to be associated with the type of op-amp used, because the characteristic sound seems to follow the op-amp around from circuit to circuit when used in the various different types of circuits. If such conditions are those for which the claim is made that op-amps sound different, then the test you propose would not appear to be testing the claim being made.
In a world of audio IC op amps, one more is not necessarily going to be heard. This is why, when I evaluate audio circuits, including op amp based ones, I use sources that are FREE of op amps, both in front, and after, including the power amp.
For example, I did evaluate a JC-3 IC based phono stage with a direct disc record from the 70's, that used all tube electronics, and perhaps another direct disc record with class A discrete electronics in the recording chain. THEN, I have a chance of hearing differences in op amps. Playing around with just any IC based source and power amp, would be mostly a waste of time, at least for me. And of course, a simple X10 gain stage is the easiest on the IC op amp that I can possibly imagine, but that does not eliminate any sound differences, IF your hi fi reproduction system is otherwise IC free. My hi fi reproduction system is all discrete, Class A, except for the final output stage (that runs as hot as I can manage), and that's the way I will keep it.
When I have to make an IC vs discrete design decision, (and I do this on a regular basis) I worry a great deal about WHICH IC that I might be forced to use for practical reasons, instead of a Class A discrete, (preferably FET) design that I know would work OK. I would NEVER start with a 4558, it is just too limited, AND I rejected a Sony SACD player when I found that I could not make out any specific listening differences between DVD and SACD, and that I concluded was 'clouded' by the 4558's in the player.
For example, I did evaluate a JC-3 IC based phono stage with a direct disc record from the 70's, that used all tube electronics, and perhaps another direct disc record with class A discrete electronics in the recording chain. THEN, I have a chance of hearing differences in op amps. Playing around with just any IC based source and power amp, would be mostly a waste of time, at least for me. And of course, a simple X10 gain stage is the easiest on the IC op amp that I can possibly imagine, but that does not eliminate any sound differences, IF your hi fi reproduction system is otherwise IC free. My hi fi reproduction system is all discrete, Class A, except for the final output stage (that runs as hot as I can manage), and that's the way I will keep it.
When I have to make an IC vs discrete design decision, (and I do this on a regular basis) I worry a great deal about WHICH IC that I might be forced to use for practical reasons, instead of a Class A discrete, (preferably FET) design that I know would work OK. I would NEVER start with a 4558, it is just too limited, AND I rejected a Sony SACD player when I found that I could not make out any specific listening differences between DVD and SACD, and that I concluded was 'clouded' by the 4558's in the player.
May be it is the time for me to start following fashions and urban legends too? ;-)
Capacitors sound, resistors sound, opamps have signatures, 12AX7 Telefunken has more bass, etc... 😀
Capacitors sound, resistors sound, opamps have signatures, 12AX7 Telefunken has more bass, etc... 😀
Different use case to me though. I am looking at the 'performance adjustment' side (or you could call it remastering). Heresy I know. I have miniDSP but the UI is a little clanky needed a laptop, which is not as useful as a small number of knobs to twiddle.
I call it "changing the presentation". So you could use the same record for multiple uses - from soft background music to critical listening at higher volume levels or headphones. If you can't do that, then your "system" becomes quite limited as a "single purpose" one. I.e. it will be satisfying only with a limited set of "audiophile" records and only for some particular listening goals. Anything else, and it will sound "a little bit off". Or more.
This is where those "purist" systems fall short in a very obvious, audible way.
This is where those "purist" systems fall short in a very obvious, audible way.
And another thing wrong with opamps in the dip package, if you leave them label down on a seat, they really can be a pain in the a..
Scott,
Not quite certain what it is that you are suggesting here... "10 boxes" etc...
But in general, I'm amenable to anything that is sensible and reasonable.
As I said before, the details and the discussion for much of something like this would best be served by private discussions via email.
When there are things that would do well in the forum, so when appropriate, then put them up.
There are implementation details that I suspect are non-trivial.
BTW, who is "we"??
_-_-bear
Please keep in mind that in my current system, there is exactly one opamp which is at the output of a 24/96 DAC (it might be a 24/192, don't recall). It drives a fixed resistive load of 25kohms in the form of two metal film 1/2w resistors.
Disclaimer: I have not done proper FFT measurements on the output yet.
After that is an amp that is DC coupled consisting of one pair of complementary JFETs and 6 complementary Hitachi Mosfets, Class A, SE.
Speakers: Quad ESL 57.
It's is completely trivial to hear the differences between a 5534 and the AD797.
This mirrors my experience using opamps at all other times, although in this case there is only one in the signal path.
Now, I may find something improper in the DAC's layout, circuit implementation that is causing the audible differences. This IS possible. Improbable, but possible. I would not rule it out.
So, it is entirely plausible that after some effort that it may be found that "opamp sonic differences" are eliminated or significantly narrowed or reduced or conversely that these things are due to X, Y and/or Z.
In which case those things could be reported and clearly shown, described, and the means by which the "issues/problems" may be eliminated/significantly reduced laid out in specific. (obviously, these would be things that somehow are not being universally applied now, and/or are specific to specific chips and implementations, etc...)
But, I think that there well may a significant amount of investigation and time involved in this to yield such a result or outcome.
Ok?
Disclaimer: I have not done proper FFT measurements on the output yet.
After that is an amp that is DC coupled consisting of one pair of complementary JFETs and 6 complementary Hitachi Mosfets, Class A, SE.
Speakers: Quad ESL 57.
It's is completely trivial to hear the differences between a 5534 and the AD797.
This mirrors my experience using opamps at all other times, although in this case there is only one in the signal path.
Now, I may find something improper in the DAC's layout, circuit implementation that is causing the audible differences. This IS possible. Improbable, but possible. I would not rule it out.
So, it is entirely plausible that after some effort that it may be found that "opamp sonic differences" are eliminated or significantly narrowed or reduced or conversely that these things are due to X, Y and/or Z.
In which case those things could be reported and clearly shown, described, and the means by which the "issues/problems" may be eliminated/significantly reduced laid out in specific. (obviously, these would be things that somehow are not being universally applied now, and/or are specific to specific chips and implementations, etc...)
But, I think that there well may a significant amount of investigation and time involved in this to yield such a result or outcome.
Ok?
Ed, I think we've long since explored the health and safety factors with regards to DIP packages. Sheesh, keep up man! Has your shop taken up specific protocols for safe handling yet? You really should get on it.
Close to the DAC output is more likely to show up opamp differences due to HF noise etc.
Better test in this case is further downstream and further filtered.
In my experience in typical line stages, swapping functionally similar opamps yields sonic changes that are easily heard and distinguished....eg LM833 is quite different to LM5532.
Over on Gearslutz, there is discussion but nil argument that opamps can sound different, down to same type number/different manufacturer sounding different and subjectively ranked.
The 'challenge' is to build a line level board with 'perfect/ideal' layout and considered choice of passive components.
Once a set of input filtering and resistances, feedback network values, output buildout resistance and output networks are agreed upon to suit a particular range of opamps, it is then trivial to swap out opamps and do the subjective listening/objective measurements.
'Solder blob' switching between inverting/non inverting modes would be useful to flesh out further differences.
I suggest battery powered unity gain line RCA connected line stage unless there is definite reason to run 10x gain.
Dan.
Better test in this case is further downstream and further filtered.
In my experience in typical line stages, swapping functionally similar opamps yields sonic changes that are easily heard and distinguished....eg LM833 is quite different to LM5532.
Over on Gearslutz, there is discussion but nil argument that opamps can sound different, down to same type number/different manufacturer sounding different and subjectively ranked.
The 'challenge' is to build a line level board with 'perfect/ideal' layout and considered choice of passive components.
Once a set of input filtering and resistances, feedback network values, output buildout resistance and output networks are agreed upon to suit a particular range of opamps, it is then trivial to swap out opamps and do the subjective listening/objective measurements.
'Solder blob' switching between inverting/non inverting modes would be useful to flesh out further differences.
I suggest battery powered unity gain line RCA connected line stage unless there is definite reason to run 10x gain.
Dan.
I have been using an ESP phono stage for a while and have just used what I think is the best sounding op-amp I have tried so far. I can't find any mention of it in any thread. It is the Ultra Analog THE-2. Any opinions?
We anyone that wants to join the fun. Sorry I left that out 5 boxes each of a simple gain stage of the two chosen op-amps. Eliminating "cheats" won't be easy. I might have time for this after I retire.
After thinking about it sending blank boxes around to identify would be a waste of time. Someone could easily drive the output to the rails and separate the groups by the slew-rate and overload characteristics.
What exactly do you mean by "at the output of a DAC"? If you mean connected to the DAC chip with no intervening filter then the opamp is probably not handling just audio but audio plus spikes or edges. The behaviour of an opamp in this situation tells you nothing much about how it behaves with audio alone.bear said:
So is this thread about opamps and audio, or opamps and spikes?
Yes, I've seen manufacturers rated on their "company sound". Considering the numerous process and fab changes across any one company I find this silly and don't take any of it seriously. And as the OP has already said read all op-amp rolling threads and they all say something different about each op-amp eventually.
Easy enough for me to make some black boxes loaded with unity gain opamps. RCA in RCA out, sealed and numbered.
How is 5532 vs LME49710 ? Have lots of both. (Too cheap to use up that many AD797s!)
I can build them send them to ? and Bear can listen and ID without reaching clipping or the noise floor.
- Status
- Not open for further replies.