And a damn good job somebody did. Thank the stars for measurement. Now that it's been observed visually I guess we can all 'trust our ears' to hear it.
Given that the engineering position is that all opamps sound the same when operated within their specifications, then none of them should be running into loads outside of what they are specified for, because that's not a fair comparison. Doing it like this just loads the test in favour of discovering an invalid conclusion.
Come on guys, I shouldn't have to point that out at this late stage.
The silliness that goes on about "peeking" really amazes me - I've looked very closely at waveforms many times over the years, in Audacity, and I haven't a clue, so far, about what I should be looking for to tell me whether one version of a track sounds better than another. Normally I hear a difference and then I visually zoom into that area, and usually it's quite apparent what the "worse" version is doing, or not doing ... but to go the other way around, that's sticking a finger into the wind stuff ...Frank, that is peeking.
Dan.
Sorry to disappoint you, as I just said, the visual 'aid' does nothing for me - Audacity serves as a very convenient mechanism for rapidly switching between short segments of the tracks on repeat, it's merely a tool for AB'ing. Personally, I find Audacity to be a little lacking in playback quality, because it does extra processing while running ... something like the Nero player is what I would use to improve quality, when needed ...
Mooly, please don't publish the results on this thread but refer those who want the answer to another linked thread. I don't want to accidentally read the results in this thread.
I'll have to try & download the files but it will take me a long time with slow internet
I was a Blind Listening Test guru for a large part of my previous life and would like to make some comments.
Those of you who have done the test, please don't reveal anything on this thread. Do it in the results thread. I think its valid to post things that you notice, eg the violin sounded slightly edgy at 3:32 but not identify which presentation you noticed it on.
When you do Blind Listening Tests for nearly 2 decades, you find Peer Pressure (among other stuff) is a HUGE influence.
I'll have to try & download the files but it will take me a long time with slow internet
I was a Blind Listening Test guru for a large part of my previous life and would like to make some comments.
- for something like this, levels MUST be matched to 0.1dB or better
- 10 is far too large a number of choices. I always did ABC tests. For something like this, 2 out of the 3 presentations might be identical
- You need to weed out the deaf Golden Pinnae. An easy way to do this is to have 2 presentations of each item. Anyone who gives Chalk & then Cheese to the same item (disguised) gets his results binned. Definitely have this with the Original.
- You ONLY ask preference of those who give similar (very rarely identical) rankings to the same item in its two different disguises.
- Someone posted a link to an ABC test but I'm not sure its the same as what I used to do
- Some of my Blind Listening Test Panel were remarkably perceptive. But most (all?) audio reviewers do much worse than the Man in the street .. especially those who claim they are Golden Pinnae. There are very few exceptions that prove this rule.
- Even the best true golden pinnae are opinionated & biased. There are people who's ears I trust very much in Blind Listening Tests and much less in a sighted test. I'm one of them.
- The Woman in the street is usually better than the Man in the Street
- The true golden pinnae ALWAYS say something like ".. I think they all sound very similar but ... "
- Chalk & Cheese comments are a sure sign of deaf Golden Pinnae
- There is music that is better for tests of this sort. Also test signals .. but you NEVER listen to test signals until you have your full set of results for music. Also ways of recording music so its more useful for these tests.
Those of you who have done the test, please don't reveal anything on this thread. Do it in the results thread. I think its valid to post things that you notice, eg the violin sounded slightly edgy at 3:32 but not identify which presentation you noticed it on.
When you do Blind Listening Tests for nearly 2 decades, you find Peer Pressure (among other stuff) is a HUGE influence.
Last edited:
To kill 2 birds with one stone - many would regard 128bps MP3 as the pits, badly mangled sound, easy to hear the 'damage done'. Therefore, it should be easy to see that damage - in something like Audacity - there must be visual clues ...
Okay, here is a snippet of some intense drumming, solo work with cymbals going flat out, one track is the original WAV, the other is a 128bps MP3 of the same - size of the whole WAV file is 1,015KB, the MP3 is 185KB - so "severe" loss of information in the latter.
So, which is which?
Okay, here is a snippet of some intense drumming, solo work with cymbals going flat out, one track is the original WAV, the other is a 128bps MP3 of the same - size of the whole WAV file is 1,015KB, the MP3 is 185KB - so "severe" loss of information in the latter.
So, which is which?
Last edited:
Richard, some people here don't like the idea, because it's 'visual',
, but Audacity is ideal for this - you can always adjust the interface so that you don't see anything if it bothers you ... or even better, look away! When something is very close I usually go for a bit of a 'wander' - move around in the room and just let the sound filter in through sideways, so to speak ...But none of you GE guys heard it?
If that is not a "night & day" difference ....
It IS a "night & day" difference. Many "senior" members have noticed how the orange is stand out. The one before me (john??) mentioned that the image on orange is strange and I agreed. When you switch L/R you will notice change to sonic (worse) and image (better). I really thought the orange is the "wild card" Mooly talked about. There was a thought that the L/R were switched but I didn't think Mooly would do something like that (And 80% of my listening was done in mono).
Indeed, I need to listen again on the orange, I believe that the preference result regarding orange should be invalid.
Now look at the leading Pomegranate. It doesn't surprise me. First, it could be the loudest file in SPL. Second, if it is not louder, then the sonic of it is superior, probably due to slew rate. Or, the noise has made it louder, which mean it is a noisy file. And you can see that "senior" members including me do not like pomegranate.
AND I said it with 100% CERTAINTY when I said which file stands out with worse distortion (Orange) and which file stands out with least distortion (Peach? Don't remember, will make sure again soon).
I'm happy to know something is wrong with orange, because otherwise it makes things can not fit into places, when I have to guess what chip orange could be, thus which file is which opamp.
Last edited:
In one of my preamps (my best opamp based actually), I used vintage LM301 in the preamp and OPA620(?) in the power supply. I compared many opamps and I preferred this setup, even tho LM301 clearly stood out in having the highest noise.
When I remember this and compare with this test, pomegranate is equivalent to the LM301. And because LM301 and CA741 are both vintage opamps, my current bet is that pomegranate is the 741.
From your tone I take it that you have taken measures in ensuring the latest offering (TLE2072 with SR of 45V/us) is stable in the circuit.
So, when there is no design "mistake", I believe I could prefer the TLE (even tho I have never heard it). It may stand out in sonic, or it may stand out in distortion (like most non musical expensive opamps).
It is not easy to stand out in both sonic and distortion, so I believe TLE will stand out in distortion. WAIT! There is one file is very close to the "original". The location of the file is not far from Peach and as of now it gets zero or one vote
I will check this file at home tonight.
No, it was noticed - but most of us didn't realize what it was. Which is kinda lame, I admit. But we thank you for pointing it out to us.
These things take time to get rolling. It was the same way with my tests. And most people don't want to take a test, they just want to read and comment. That's the nature of internet forums.
Hehehe good question.
But it takes time. Easier is to pick what stand out. In the middle, they could be "equal", including the original should be in the middle.
When I did the blind test, some files just get comment like "Hmmm... no mistake but nothing interesting". Some got "whatever it is it should be one of the worst". "This should be one of the best", etc.
Based on those comments I rank them. Most of the files in the middle of the ranking just didn't get commented because nothing was special about them, nor sonic benefit, nor distortion issues.
And it really takes time because there are 4 musics in each file. That's why I only did one cycle for all music and the rest cycles are for the first music only because I can do it by pressing the "next" button.
There were serious cycles where I didn't move from my listening location. There were cycles where I just try to pick my preference based on sonic and musicality. This is done in non-serious background listening, and wait for that moment like "Hey this is very musical. Not perfect but I'm okay with this one" or "This is terrible. This belongs to the worst", then go check the name of the file.
I can check my note at home to see the ranking. I think it is something like:
1) Banana
2) Orange (?)
3) Kiwi
4) Peach
5) Apple
6)
7)
8)
9)
10) Pomegranate
But Orange will surely taken different position soon (with the problem fixed), can be better or worse.
No matter what you do for a test, someone will object. If Mooly had given them all an easy load, someone would complain that was invalid because - whatever.
If you can't easily tell them apart, and can't easily pick the original, that's valid. No complaining will change that fact.
