So, different DAC could be the reason for the Claire#2 sounding very different than #1 and #3?
I like Claire#2 for being the most musical, but a bit disappointed by the thin soundstage and instruments that are pushed to background. But this still sound like 744 that I know. My best active currently using 744 for LF (input buffer etc.) as it is good there and it doesn't have to deal with HF which is bad (just like the Claire#2).
So I preferred preamp #1 on both, but yet Claire 1 sounded much better than WS1 ; I’m wondering how much difference between dac configuration there was?
If I’m understanding correctly WS1 was inferior recording + inferior dac setup into preamp 1 but C1 was better recording + better dac configuration into same preamp 1?
If that’s the case it’s hard to tell how much the recording quality actually mattered.
In reference to “can’t hear evident difference between 1 & 2”
I found C2 quite veiled compared to C1.......what was the ‘mod’ ?
Seems like a lot of variables not just a change of recordings.......I only change one thing at a time when I do subjective testing, otherwise it’s hard to tell what’s what!
Last edited:
+1
For now, the few answers given confirm many of my beliefs:
1: We (all of us) are not looking at all for the same things, when we concentrate on sound reproduction*.
2: There still are audible differences low, low under what we believe as thresholds of audibility. (Why and how ?)
3: Measurements do not illuminate these differences and do not explain them.
4: Blind ABX is not the best way to figure out subtle differences.
5: What is the importance of these micro differences, compared to those of the source (Both technical and artistic quality) ?
* As a sound engineer, i was trying to have, at home, the most "revealing" system I can to can watch each and every mistakes and beauties in my own (and others) mixing work.
Some kind of a magnifying tool. In a way, hyper realistic.
Good to easy listen to music ? Not so sure.
It's sometimes aggressive, it invites you to listen at realistic level, which is quickly tiring, it highlights all the flaws (there are records that I love and that I can only listen on poorer systems). By example, impossible to use it to reproduce poor quality TV sounds. I realize that, on a daily basis, I'm using-it ... less and less.
At the end, where is the interest to have 10 records that sound fantastic, and the other 1000 where the play of the musicians is spoiled by small production defects that underlines your system ?
PM: PMA, on all your (nice) preamp designs, may-be you should try to buffer (a lot) the outputs ?
May depend on the amp being used. At least Tournesol put C2 at the bottom list, meaning that he could hear the 'problem' with it. He also mentioned 'bass bump' with C2 (Indra1 also mentioned more bass with C2). I won't use 'more bass' to describe the sound, rather 'more drive' or 'more control'. AD744/746 is well known as bass lovers' opamp. But like i have mentioned, the biggest difference is not with bass. This might not easy to spot if not using my transparent amps using 60MHz NEC output transistors.
BTW, C1 is very transparent from bottom to top, but I can still perceive the hard fatiguing bass, even very slightly, the reason it is not my taste.
Tournesol - Re buffering, please visit my website
Pavel Macura audiopage
and you will see that most of the preamps have buffered outputs.
The preamp#1 here (which I normally use for measurement purposes) uses ADA4898 opamps at the output that have 40mA output current capability. Buffering has IMO many advantages but some disadvantages as well.
I agree with you that many of us concentrate on different sound attributes and also have different sound systems which counts as well. Similarly as you (probably) I ask for transparency and NOT for smoothing of poor recordings.
P.S.: those horrible tablet checkers - mistake corrected - NOT asking for smoothing of poor recordings!!
Pavel Macura audiopage
and you will see that most of the preamps have buffered outputs.
The preamp#1 here (which I normally use for measurement purposes) uses ADA4898 opamps at the output that have 40mA output current capability. Buffering has IMO many advantages but some disadvantages as well.
I agree with you that many of us concentrate on different sound attributes and also have different sound systems which counts as well. Similarly as you (probably) I ask for transparency and NOT for smoothing of poor recordings.
P.S.: those horrible tablet checkers - mistake corrected - NOT asking for smoothing of poor recordings!!
Last edited:
I agree that C2 has worse impact and attack than C1, but not worse than C3, at least not to me on my headphones, amps and speakers. It may be system dependent and I do not think it is worth arguing. I can see we all prefer C1 and this is the setup as designed. C2 mod was used just to have third option for the test. Two options only are more bias prone, IMO.
There are people who will argue all day that you must be imagining things that aren't real. Research to answer your questions authoritatively is probably not going to happen in the foreseeable future since no one is willing to pay for the cost of doing it. On the other hand, if you want educated guesses rather than research, how would you know who's opinion to trust? No easy answers, IMHO.
We have known for a long time that we don't hear exactly the same way we measure, although some people will argue all day that we already measure everything we could possibly hear. If you want research to provide explanations, same problems as for your last question above.
Changing the subject for a moment, I have some trouble seeing the point of these listening comparison threads. Guess they might sway a few skeptics to accept that we may be able to hear some things that didn't seem possible before. Other than that, is anything useful coming out of it or is it just something to do solely for entertainment?
Measurements are not to be thrown away. Good or rather perfect measurements are a necessary condition. I do not know the situation that I would subjectively prefer a poorly measuring component to a well measuring one. I think this test is a confirmation. With a technically poor recording, worse component that masks troubles may be preferred. With good recordings it is not the case. I do not like the situation when people who do not have enough experience and skills in measurements say that measurements are useless. That's why I appreciate listening comments from anyone, but comments on measurements and their evaluation only from experienced and educated people.
You are definitely not forced to participate, so please feel free to ignore the thread. There are many many other threads that may satisfy you. To me, as a thread starter, results and comments are very interesting.
Some interesting words from ifi on measurement vs sound quality: iFi iDSD Micro DSD512 / PCM768 DAC and Headphone Amp. Impressions, Reviews and Comments. | Page 475 | Headphone Reviews and Discussion - Head-Fi.org
So like mark says, maybe these kind of tests will show there is indeed audible differences between level matched gear. (EH anyone!)
And I sure would like to know what my preferences equate to in measurements and Pavels tests do reveal something about that.
But I’m still thinking there’s more to it than what we currently know.
And I sure would like to know what my preferences equate to in measurements and Pavels tests do reveal something about that.
But I’m still thinking there’s more to it than what we currently know.
One falls back on one's own experience in these circumstances. We are all struggling to make sense of what/why we hear what we do
There's always another measurement or a more sensitive measurement - see ScottW's post of ESS plot - only when input signal was notched out was the pattern of amplitude variations visible.
It's always interesting to see fundamental research like this - just as it's interesting to hear anecdotal reports of listening. I don't believe these things have to be framed in terms of swaying one way or another - just like listening, we all seem to extract from it different aspects. I for instance, was surprised that PMA admitted ABX listening was not giving him the above random results that his long term sighted listening "easily audible" differences suggested should show as near 100% discrimination in Foobar ABX listening test.
Ah but when has something reached 'perfect measurement' condition? As in ScottW's ESS plot there can easily be some uncovered measurement of imperfection?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.