No by ear. But then the system used had resolution way in excess of virtually anything you can buy at any price. However I had easily confirmed this on more commonly available hi-end type of equipment.
Surprised you think that measurements will give meaningful information as to sound quality.
Sure, have used test equipment to confirm distortion figures of some open loop gain stages in an amplifier front end that yielded 0.03% max, but you could swap out a triode cascode for a bipolar or MOSFET and see no difference in measured distortion but it sounded like crap.
Once you realise the distortion mechanisms in audio you realise that distortion measurement equipment is mostly a waste of time as it doesn't capture what is important. Out of them all, the only one that seems a little useful is the spectrum analyser.
When you can hear the difference between mica and PP capacitors in an anti-aliasing filter and not measure a damn gnat's hair difference then you know distortion measuring equipment has grave limitations.
Nice advanced hobby circuits on your web page.
I know your bio scott. You have a lot of psyche tied up in your body of work and I can understand your belief in all that you have done.
BTW the 797 is a decent low noise op-amp and has some good application in that environment, however the 627 sonically is in a different league. It's more open, relaxed with natural timbres. The 797 is more clinically cold and not engaging musically. Still better than most other audio op-amps like the OPA604, OPA2604. The hi-biased 5534s (were they Japanese specials?) are not bad, a little grungy and warm but still maybe better than the too clean 604.
Ever tried balancing the impedances at the input of an op-amp with symmetrical i/p stage, i.e. not a current feedback type?
Improves settling in data acquisition systems but also improves sound quality in audio systems. But then they need to be high resolution. A set of NS10s and Onkyo amplifier is not going to cut it.
@CraigBuckingham
What do you think are the causes that many people do not differentiate between the files of this and others tests - mp3 320kbps vs FLAC, 16/44 vs 24/96....?
What do you think are the causes that many people do not differentiate between the files of this and others tests - mp3 320kbps vs FLAC, 16/44 vs 24/96....?
Last edited:
So what is your proposed mechanism for these effects? Why do they not show up in any other T&M, aerospace, medical, etc. applications? Just arguing from your own personal experience is not going to cut it; the internet is full of deaf people who think they can hear just about anything.
Do you think LIGO's gravitational waves are clinical and cold? Maybe if they used OPA627s they would get more musical gravitational waves?
OK. Do you have any proof supported by a controlled test that you have really heard the difference?
Frankly, the main reason why I prepared this test is that in the past I have performed many uncontrolled listening tests where the op amps were swapped and I was sure I heard significant differences, together with guys who participated in those tests as well. The most job was done with this
pre standard preamplifier
We had pretty good audio equipment for these tests, with speakers up to Wilson Audio Maxx (I assume most audiophiles would not protest against this one here). But regardless speakers, the “difference” had always been there, at least we were sure about it. The preamp was finally used during Prague Audio Show in 2004, with Wilson Audio Sophia.
http://pmacura.cz/exp1.jpg
Quite a long time passed since that event, digital sources have improved a lot (though recordings got much worse, except for classical music) and I have started to evaluate less by “heart” but rather be “evidence”. Audible differences started to disappear. Recently I designed an AB box that would allow to test two power amplifiers to same speakers by A-B switching. The AB box has had 2 independent link signal paths with 2 Alps potentiometers and op amp buffers, to match gain differences of the amplifiers under test. It also has had 2 relay switched power paths to link 2 stereo amplifiers with 1 pair of speakers. The AB box is built into quite well shielded metal Al box with top and bottom covers made from 3mm Al metal sheet. Clean stabilized power supply with TL431, op amp and 2 transistor regulator is used. Now, it was possible to perform unbiased AB and ABX tests of power amplifiers. It became quite difficult to find any difference, though it is still possible in case of very different designs.
As a secondary use, this AB box may be easily utilized for op amp testing, provided that IC1 and IC2 positions are occupied by different dual op amps and 2 Alps pots are set at the same volume. So, I decided to make an op amp blind test for myself, after having made many op amp tests by IC swapping in the past (like 15 years ago), when "clear differences" were audible in uncontrolled tests. Frankly, after those 15 years my current audio system is much better tested and thoroughly measured, compared to that one I had 15 years ago.
So I had put op amps into sockets IC1 and IC2, matched volumes of both paths within 0.1dB and started A-B testing. And - nothing was heard, both paths sounded the same. I was unable to tell the difference. So I dug out the oldest dual op amp I still have, the Tesla MA1458 (LM1458 equivalent) from about 1980, and put it in the test against XXYYZZ, one of the best contemporary audio op amps. Again - almost nothing. With some music material, I thought I heard some difference, but was unable to confirm it by a valid AB test protocol.
Based on this experience, I decided to prepare the op amp test with 4 music files recorded through 4 different dual op amps. This test was then published at one of our local forums and here at diyaudio.
The uncontrolled tests are nice and we may have a lot of fun with them, but a lot of cheating and bias as well, due to the complex process of human hearing and evaluation of sound.
Last edited:
It's a very common belief, so how come surprised? You don't sound like an easily surprised type.
The mechanisms for distortion in audio amplifiers are many and varied. I spent a lot of time researching them so I'm not going to give away that hard earned information for free. It's like Wenzel designed oscillators, he gives some ideas about low noise squaring circuits and multipliers, but do you think he reveals what he has in his state of the art DOD restricted designs or even main shelf items?
One mechanism does show up in data acquisition systems in regard to settling times. If you refer to my previous post I explain that and I give a freebee clue away to improving their performance. As I am not interested in wasting time with the limitations of op-amps, I have no qualms realising it. It was in a National Semiconductor application note first I noticed. From the 70s IIRC. I believe Analog Devices may have touched upon the subject in a low noise op-map signal conditioning design for data acquisition also.
I can assure you I am far from deaf. I very rarely have had anyone hear anything I cannot. Generally it's me doing the coaching.
I have no idea what a LIGO's gravitational wave is, sorry I can't answer that question.
Yes, it is a very common belief that a sound difference must reflect in the difference in the electric signal. Can you disprove it? 🙂
Okay, no evidence will be shown then, so why a post like this? Is not it a bit pointless? 😉
I assume that anyone here knows that "the mechanisms for distortion in audio amplifiers are many and varied", it is just an empty statement. Are those "mechanisms" unmeasurable? Is it a problem today to have a part with all distortion artifacts below >>-120dB through whole audio band?
Then we can speak about external influences like EMI (air coupled) and PSU coupled noise/interference. Again, are those phenomena unmeasurable? Or is it just necessary to create a hype in audiophile audience? 🙂
Well Pavel, you are conflating my reply to a question regarding the mechanisms to my general initial post. They are two different issues and I was only offering the courtesy of a reply to the question asked. My initial reply was in regard to wasting time with op-amps.
I did the same thing you did back in the early 90s and I had some guy saying the same thing to me. I thought I was brilliant and capable enough to be able to improve on the implementation limitations to make op-amp circuits beat his discrete designs. So I went with o/p class A biasing, and no measly 2mA, but close on 10mA. Then esoteric open loop low impedance shunts fed from cascode current sources. Balancing common mode impedances. Then I looked at one of his designs, biased it up a little more and that was the end of op-amps for me forever.
Sorry, I've really ended up hijacking your thread, so I should really butt out and let you cintinue on with the ABX comparison which has a lot of merit in itself.
Would be nice to know the D-A-D chain used.
I really wanted to try this again, and like yesterday evening, I just feel relaxed and able to do this. Can't explain it. The concentration required is intense but I'm listening at a moderate to high level via Sony MDR V7 headphones that must be 30yrs old now. The playback setup is a 10.5yr old Acer laptop running W10 with default Windows drivers for audio. No response tweaking or anything like that employed. You can see from the times that I take quite a while before moving on to the next 'test' with it taking nearly 5 minutes for just ten attempts.
Attachments
So, PMA you are lucky man, after 15 years, now you can make your self some "gainclone" chip power amp. It makes no difference after all... 🙂
Mooly, how did you do the listening? I guess you picked a very short segment and compared the impact or pressure? Aa and bb are both the extremes in term of (subjective) sound pressure.
Hi Karl!
great result 🙂. So I took your hint and used one of the worst headphones I have here, JVC I bought at the Manchester airport 2 years ago at some 15 GBP (remember when I called you from Manchester?), and immediately got better result
Code:
foo_abx 2.0.2 report
foobar2000 v1.3.7
2017-06-13 13:38:57
File A: aa.wav
SHA1: 81e0cc3b0146df3562d4421385098251239b3789
File B: bb.wav
SHA1: 4da5d9045285da07488f69c24cda9187be2882c9
Output:
WASAPI (push) : Speakers (Cambridge Audio USB Audio 1.0), 24-bit
Crossfading: NO
13:38:57 : Test started.
13:39:28 : 01/01
13:39:40 : 02/02
13:40:03 : 03/03
13:40:44 : 04/04
13:40:54 : 04/05
13:41:16 : 05/06
13:41:30 : 06/07
13:41:39 : 06/08
13:41:39 : Test finished.
----------
Total: 6/8
Probability that you were guessing: 14.5%
-- signature --
26994466d1ad7e78f19f720799029ed6752abf23Still not sure what it means and I doubt I could make it repeatedly, however I agree that the worse the system, the more we "can hear" component changes like swapped ICs etc.
Attachments
Yes, I seemed able to make the judgement based on the first couple of seconds of the tracks and by concentrating intensely on that first hi-hat and the shimmer and spread of it. I would for example keep playing 'aa' repeatedly, over and over for that first couple of seconds and then 'play X'. Was it different on the shimmer and decay or the same. The acoustical memory seemed good over that short time span.
Also, and I've never studied this with Foobar but the results seem to show I wandered off toward the end of each run of ten, as if the concentration was fading. I was getting 1 of 1, 2 of 2, 3 of 3, all correct up to the 8th test. Then the concentration perhaps went.
Also, and I've never studied this with Foobar but the results seem to show I wandered off toward the end of each run of ten, as if the concentration was fading. I was getting 1 of 1, 2 of 2, 3 of 3, all correct up to the 8th test. Then the concentration perhaps went.
I do remember, and your pictures of Winter Hill.
Hey, I don't have to come here to be insulted you know
those Sonys were a top performer back in the day although sad to say the earpads no longer seal as well as they did.
Your concentration wavers to I see (toward the end of the tests) 🙂
Hey, I don't have to come here to be insulted you know
those Sonys were a top performer back in the day although sad to say the earpads no longer seal as well as they did.Your concentration wavers to I see (toward the end of the tests) 🙂
- Status
- Not open for further replies.