Mooly - go for all the same opamps in a test. Go buy a few that you need to fill all the slots.
Assuming whatever you are playing this through will actually permit you to hear whatever is being changed, I'd say you're aiming in an interesting direction.
You might be better off not trying to invert tone controls, and just set the thing for dead flat. Perhaps fixed 1% resistors??
An idea that occurred to me that is more along the lines of what Tom Danley suggested for finding speaker distortions. Not quite clear on the specifics of his method, but generally speaking he plays the speaker, records it, plays it, and repeats this a certain number of times. The principle being that the same distortions "stack" upon each other, increasing in amplitude with each pass.
I actually was thinking of doing a "Danley" test rig where multiple stages of opamps are set for unity gain, or some suitably low gain (but unity seems most practical) perhaps like 10 in series and then the output viewed.
Pretty easy then to compare with a single stage, even while the whole thing is running.
In addition one could muck about with power supplies.
See if there were any changes in the results.
Now as far as regulators go, although I have no data or measurements to support my conjecture, this is what I think may be happening: while the PSRR appears to be high, and the resulting "nasties" from even a lowly 78XX regulator ought to be so low in level that they are lost in the noise, apparently this does not happen in practice (yes, low level, but not inaudible).
Perhaps the reason is that PSRR is not a passive process. It's an active process that requires the circuit to null and reject common mode signals appearing on the rails. That same active nulling is taking place while the desired audio signal is present. That means that there must be some IM and other distortion produced. Perhaps this is why seemingly low level power supply noise and ripple is objectionable, and why there is a push for super regulators??
_-_-
So we're questioning whether differences exist... But how can we prove that it may exist???
Correct. That's why I don't like the ABX idea to prove audibility. Because I can't assume that we are all adults here. But remember this: There are many kinds of distortions and its audible symptoms. It is true that level difference is the easiest to perceive, but there are others unrelated to AM distortions...
That's why I often ask/offer to guess "which is which". For example, I can try to guess which one is 4558 and which one is 4562. Level difference will not affect judgement, because in the blind test I will try to listen to the speed and musicality of complex musical passage. The assumption is that high slew rate opamps like LM4562 will outperform 4558, especially at mixing position (U2).
imho it ought to be instantly obvious which is a 4558 vs an LME4562.
unless something else is (for lack of a better term) "wrong".
Again, I would stay away from "test circuits" where one mixes more than one type of opamp. Although I guess if you had say 6 positions and ONE was a 4558 and the rest were 4562 and compared that to where ALL were 4558 I'd expect to hear a difference - but it would be most clear if all the devices were the same. So why mix them up, if one is looking for clear differences between opamps??
_-_-
unless something else is (for lack of a better term) "wrong".
Again, I would stay away from "test circuits" where one mixes more than one type of opamp. Although I guess if you had say 6 positions and ONE was a 4558 and the rest were 4562 and compared that to where ALL were 4558 I'd expect to hear a difference - but it would be most clear if all the devices were the same. So why mix them up, if one is looking for clear differences between opamps??
_-_-
I like the idea of stringing say 10 or 12 opamps of the same type in series, each stage with a gain of say 20 dB followed by an attenuator of 20 dB.
You could then have 4 or 5 of these in parallel and could switch between them to get a feel for any sonic signature.
You could then have 4 or 5 of these in parallel and could switch between them to get a feel for any sonic signature.
I have not necessarily picked up on the 'possible' Foobar caused level differences between tracks 1 and 2 .... typically I am doing appraisals by jumping between tracks further down the playlist.
However, I do find that Foobar ABX applet and also VST wrapper add a layer/film of broadband noise over the music.
This 'white-ish' noise conspires to subtly mask/obscure fine differences between files under test, and erroneously encourage/forces null resultant.
Once aware of the Foobar applet limitations, one needs to consciously 'listen through' the Foobar created noise in order to discern actual differences between files.
This conscious 'listening through' takes high level of concentration, and is prone to wrong results due to mental fatigue and losing track/confusion of which file sound is the reference due to the Foobar masking.
I have repeatedly done the experiment of loading two only, very similar but slightly different files to a Flash drive or Android player, plugged into a car stereo.
The learning process includes, pressing Track Forward once switches to 'B' track, one more press returns to 'A' track etc...with practice one is able to recognize and diffrentiate the two tracks easily and reliably.
The next step is to press the TF button a random number of times, and then press known number of times whilst comparing tracks.
By this method I find that I am able to blind consistently correctly discern and identify very fine differences with perfect degree of statistical reliability.
In my experimentations, I am dealing with fine differences that on the surface are not necessarily obvious to some (1), but on further listening do actually become glaringly apparent.
This is the stuff that JC etal are on about.....a system can measure 'right' by standard measurements, but still sound 'wrong'.
IMO, the fact that Jay is achieving high certainty despite the confounders of Foobar caused noise masking and Foobar 'possible' caused level differences, actually reinforces confidence in his level of sensitivity.
Dan.
(1) - I find that any 'idiot off the street' immediately hears the fine but musical differences that I am dealing with...and it can be the technically trained/predisposed and/or self proclaimed audio 'experts' who have initial difficulty,
Last edited:
I am acutely aware of my limits. I am also acutely aware of some of the performance artists lurking on here!
Dunno about always but he's taller than me, stronger than me and knows where I live so I won't upset him!
I like to keep it simple, makes it easier for me. FWIW you might be the outlier with the hearing of a dog. I have insuffient evidence.
trust me you would not want to look in the mirror and see me. I have the T-shirt 'if you can't be an example, at least serve as a warning'.
@Mooly. Cool idea, do it. Would also be interested how you find the precision pre-amp tone stack as been tempted to build one of those.
Already done in the late '80s, but with a unitary gain string, followed by one with gain and a complementary output stage (Phili 139/140) in class A, into a set of headphones. (2 switchable sets)
Prior to that, the Elektor lab had done an opamp comparison for their upcoming design (The Preamp, line + phono), tested 5534 and OP27 of several manufacturers by measurement and ear.
First of all I think IQ testing is mostly utter nonsense as far as really measuring intelligence of a set of people. I was tested with an IQ in the range of 160 when I was rather young, this was based on puzzles and other things that I no longer remember but really all that shows is that I was good at deducing particular things and was good at that type of thinking. I don't believe for a minute that I am some genius, just that I have developed certain skills and that I am very good at visualization. I do design mechanical things but I think of others as having much more education than I do in particular aspects of design. I always listen to people saying what their IQ's are and I have meet few people who actually seemed that intelligent who claimed very high IQ's. I have very little electronics training so I stay out of discussion of particular parts and circuit design, I have no skills there.
Mooly,
Is that an actual tested circuit for the preamp your showing, does it work and would it make a good build for a pre with tone controls? I would be very interested in a nice pre with tone controls, I definitely am one who believes in the use of tone control for different music and spl levels. I'd love to see that as a board and kit if it is an actual design. I have read your replies in the headphone 02 thread for some time and know your skills at circuit trouble shooting.
Jay,
while I am not a computer programmer or expert I do know enough to understand that you can do a comparison of a computer file and compare if two files are identical, it is done all the time to make sure a download is exactly correct, a bit sum comparison for length and bit for bit accuracy is a common thing to find an error. For someone who just said they are a computer expert I find it hard to understand that you can't see that you were using two bit identical files and then saying you could detect a difference when that should be impossible. If Foobar is in error there should be a way to take the output from that program and find what the error propagating would be for a computer expert.
When someone says they are getting 8/8 results in testing and then admit to peeking then I have some difficulty not thinking that you have subconsciously, I will not say intentionally cheating, that you have made some connection to a particular phenomena that is giving you these exceptional results. Something is flawed in your test protocols, you may just be blind to fooling yourself.
Do I believe that all opamps are interchangeable, no, I believe there are enough basic differences in topology between designs and designers methods that seems rather unlikely. Do I believe that many similar opamps are out there and can be interchanges for audio and not be detected, yes. But it is up to the circuit designer to understand what to use where and what opamps are able to be interchanged. I don't think you would take a typical audio opamp and then insert an RF opamp and expect there not to be some difference and need for circuit modification to keep circuit stability and watch for oscillation problems.
There are many very smart people making well thought out statements here and others who are just holding a position, it is for each of us to determine who is doing what and who is just arguing for arguments sake. I know certain people I have read enough of their comments to understand they are very well versed in the electronic arts and others who are very well versed in communication skills and worldly knowledge. I find Jacco a very intelligent communicator at the same time I don't know what he is referring to sometimes and don't take the time to convert the saying to find out what he just said. Drunk or sober I don't think I would argue without knowledge with Jacco!
I respect many in this conversation, I think though that some are working at cross purposes in this argument, or what I really want to say that many are suspect of the claims of Jay as they seem more than just outlier in nature, they just seem to be in error of method.
Mooly,
Is that an actual tested circuit for the preamp your showing, does it work and would it make a good build for a pre with tone controls? I would be very interested in a nice pre with tone controls, I definitely am one who believes in the use of tone control for different music and spl levels. I'd love to see that as a board and kit if it is an actual design. I have read your replies in the headphone 02 thread for some time and know your skills at circuit trouble shooting.
Jay,
while I am not a computer programmer or expert I do know enough to understand that you can do a comparison of a computer file and compare if two files are identical, it is done all the time to make sure a download is exactly correct, a bit sum comparison for length and bit for bit accuracy is a common thing to find an error. For someone who just said they are a computer expert I find it hard to understand that you can't see that you were using two bit identical files and then saying you could detect a difference when that should be impossible. If Foobar is in error there should be a way to take the output from that program and find what the error propagating would be for a computer expert.
When someone says they are getting 8/8 results in testing and then admit to peeking then I have some difficulty not thinking that you have subconsciously, I will not say intentionally cheating, that you have made some connection to a particular phenomena that is giving you these exceptional results. Something is flawed in your test protocols, you may just be blind to fooling yourself.
Do I believe that all opamps are interchangeable, no, I believe there are enough basic differences in topology between designs and designers methods that seems rather unlikely. Do I believe that many similar opamps are out there and can be interchanges for audio and not be detected, yes. But it is up to the circuit designer to understand what to use where and what opamps are able to be interchanged. I don't think you would take a typical audio opamp and then insert an RF opamp and expect there not to be some difference and need for circuit modification to keep circuit stability and watch for oscillation problems.
There are many very smart people making well thought out statements here and others who are just holding a position, it is for each of us to determine who is doing what and who is just arguing for arguments sake. I know certain people I have read enough of their comments to understand they are very well versed in the electronic arts and others who are very well versed in communication skills and worldly knowledge. I find Jacco a very intelligent communicator at the same time I don't know what he is referring to sometimes and don't take the time to convert the saying to find out what he just said. Drunk or sober I don't think I would argue without knowledge with Jacco!
I respect many in this conversation, I think though that some are working at cross purposes in this argument, or what I really want to say that many are suspect of the claims of Jay as they seem more than just outlier in nature, they just seem to be in error of method.
Afair, that's exactly what the graduation prof at tech-u said in a few more lines. (he enjoyed listening to himself, but who am I to censure)
Of course I know how to do it, but it doesn't mean that I would bother to. One of the reason is because I was pretty sure that the software worked as expected. Put your feet in my shoes, what will you think if you duplicate a file with SPL difference, listen to it blind and you can hear which one is louder?? (And I did it many times with different SPL differences).
When Mark and others warned be that the files were identical, I renamed back the files and put them on FoobarABX and I still could ABX them. Put your feet in my shoes, what will you think? What are the possibilities?
(1) Foobar might have used the metadata information of the file to store the SPL playback information.
(2) Foobar might have bug (which was hard to believe until somebody confessed he had observed the bug).
(3) It might be a statistical coincidence. But no, how can it be, I know my records.
I have explained the reasons why it is possible. And says who they are (bit-by-) bit identical? No they are not!
Sure. But I'm not gonna do it. Never.
Yes, something is flawed there. I have mentioned the possibilities. But I'm not fooling myself. I have no burden, because I know who I am.
Hehehe... You know, people are talking about the strategy how to win debates. People are learning things like this, so it wouldn't surprise me if people also learn how to appear to be smart... No, I'm not like that. I'm an outlier in also psychology...
The files were identical. No gain difference, no metadata difference - I guess we don't know why..
But SHA1 being the same is a big clue, hash collisions are *extremely* unlikely, probability of the order of 2^80...
So that should have led you to do a quick file compare...
But SHA1 being the same is a big clue, hash collisions are *extremely* unlikely, probability of the order of 2^80...
So that should have led you to do a quick file compare...
Well I know you reckon to be able to pick the 4562 out at 50 paces, so you just might get your chance 😉
If the controls are dead flat then most of the opamps contribute little to nothing. We'll see how it goes. I haven't listened to any completed files yet so its all a bit of an unknown at the moment.
Here's hoping.
Perhaps Jay is not an outlier, and is observing real effects.
In my experimentation I am getting different sounding PB according to the USB cable used to transfer copies of the same file.
Binary file comparator shows zero file differences, however the two files consistently sound different, and 'any idiot off the street' is hearing these differences and indicating strong preferences despite PB on bog standard car stereo...ie SOTA hifi is not required to reliably discern these differences.
'Digital is digital'.....I am finding this not entirely true.
Dan.
Thanks 🙂
Yes, Dougs preamp was an actual design and imo the 'return to zero' tone controls were the best bit.
http://www.diyaudio.com/forums/soli...-built-listened-d-selfs-precision-preamp.html
Well, different playback systems may be... different or whatever. But identical files if played correctly on the same system will sound the same.
That's the very definition of it. Data is data.
.. We're also still waiting for the files you promised... 😀
Jay, your statements show you are very normal in terms of your beliefs about yourself. It doesn't help your credibility that you have the same illusions most people do. However, whatever you think of yourself compared to other people, it is independent of the extent to which you actually are. That's true for everybody.
In other words, that you are normal in such respects has no bearing on what you can or can't hear, as measured by some objective test independent of your own self-evaluation. Of course, I don't expect you to believe me, since you will trust your own perceptions first before somebody else's words. That's okay with me, since even if you have very good hearing, it would still be unlikely for you to also be exceptionally immune to naive realism. That you deny such immunity, however, doesn't make your other claims appear any more convincing. If anything, probably the opposite.
That's a very diplomatic way to describe it. 😀
All the clues are there to show exactly what's going on.
Data is data....yes of course. I am finding that digital data is not the only information that is contained/embedded in data files....something weird/spooky/interesting is going on.
Promised files will come.
Dan.
Last edited:
- Status
- Not open for further replies.