Inane question (I couldn't satisfy myself just looking at the "Go Advanced" tab)--would it be possible to put spoilers (or something where you have to click to show said text) in place? It'd at least let some of us who are trying to remain moderately blinded from being *as* influenced by others.
Not perfect, but it's a happy medium between closed submissions and free-for-all.
Anyhow, I should mention, by at least what I heard, I prefered the B-E family to the AF family (albeit, weakly). If I were looking through this from my scientific lens, I'd say my preferences are below any threshold of significance.
Not perfect, but it's a happy medium between closed submissions and free-for-all.
Anyhow, I should mention, by at least what I heard, I prefered the B-E family to the AF family (albeit, weakly). If I were looking through this from my scientific lens, I'd say my preferences are below any threshold of significance.
I just spent some time with Lacinato and these 6 files. I am using a crap setup of laptop output to a mediocre set of Altec Lansing desktop speakers.
The differences are incredibly subtle. In fact, for most of the files I cannot tell the difference. I isolated my listing to between 00:53 and 01:05, listening specifically to the subtle, low-level guitar sounds (fingers on strings, etc.) in the right channel.
I find "C" to have a clarity or "impact" in the fine detail. "E" and "A" (to a lesser extent) come close in this same area.
I am not 100% confident, however, that I could pick one from the other in ABX testing.
This is very interesting to me so far. Enough that I plan to listen on my proper home system this weekend (bypassing my EQ which is jammed full of 4558s, of course. 😀 )
The differences are incredibly subtle. In fact, for most of the files I cannot tell the difference. I isolated my listing to between 00:53 and 01:05, listening specifically to the subtle, low-level guitar sounds (fingers on strings, etc.) in the right channel.
I find "C" to have a clarity or "impact" in the fine detail. "E" and "A" (to a lesser extent) come close in this same area.
I am not 100% confident, however, that I could pick one from the other in ABX testing.
This is very interesting to me so far. Enough that I plan to listen on my proper home system this weekend (bypassing my EQ which is jammed full of 4558s, of course. 😀 )
Last edited:
I take some care to make sure that I don't view any in between posts at all, in such a thread before I send off a set of results; only then do I go back and read what others have said - this involves nothing more than being careful in the use of the browser.
I've failed the course, then. I don't see why you can't just pick one file and use that as the reference. I did. Then listened for any differences. Yes? No? Better, worse?
I also raised some objections to the test and lack of reference early in the thread. I don't want to assign a "sound" to some opamp when I don't know what else is in the circuit. Does the unknown part have a coloration so large that it masks differences in the DUT? Don't know.
But if we can accurately tell some of these opamps apart, that's a good start. If not, we have to look at why we can't. Because they aren't audibly different, or because the differences are masked by something else?
Well as hinted at yesterday, I think its time to put the reference file up. This is a direct rip that hasn't been through the playback process of Micromega CD player > opamps > AD conversion etc.
REFERENCE
I'll hold off identifying the various files for the moment as you may want to compare your choice (your best and worst) with this reference.
REFERENCE
I'll hold off identifying the various files for the moment as you may want to compare your choice (your best and worst) with this reference.
That would be great if we could do that. Can't recall ever having come across anything like that on any vbulletin based forum though.
It's hard to interpret your comment, as I can see both positive and negative comments at the same time 🙂
But it seems that you have treated A as a "reference", as though it was the original or direct file. Interestingly, I found A to have some characteristics of direct/original file, i.e. tonal balance and phase characteristics. But from it's noise, it cannot be an original/direct file.
I don't think so now. All that are going to listen to the files will have done so. It might be interesting at this stage to get comments on how the two vary audibly.
Arny, you're still as cantankerous as when I saw you on the rec.audio newsgroups years, and going on decades ago. I haven't read 'em much if at all in recent years.
I've just read through the thread and have only listened to the reference (is it cheating to listen to the reference first? 🙂)
One complaint I have about the first post is the schematic with the "Dirty ground" line - if you give the thing to a few layout peeps you'll get several layouts with grounds of varying dirtiness. Unliess I missed it, we don't have much of an idea what you have. Can you post a pic of the actual board you've put the opamps in? Maybe both sides? That should be interesting...
Late to this as usual.
I found them all pretty similar but would be happiest listening to C. A was a bit flat.
The reference track was plainly a fraction clearer than all the others but I still found it fatiguing compared with what I am used to. As soon as the voice came in, it seemed unnatural to me and I wasn't comfortable with it.
I was listening on a MacPro via external DAC into Sony TA-F300 and Sennheiser HD518. I'll try something better later. This is my 'snug' system and the Quad ESLs may reveal more.
I found them all pretty similar but would be happiest listening to C. A was a bit flat.
The reference track was plainly a fraction clearer than all the others but I still found it fatiguing compared with what I am used to. As soon as the voice came in, it seemed unnatural to me and I wasn't comfortable with it.
I was listening on a MacPro via external DAC into Sony TA-F300 and Sennheiser HD518. I'll try something better later. This is my 'snug' system and the Quad ESLs may reveal more.
Son of a gun! (Using cheap/old laptop to a headphone) Sonically it is actually in the following order:
1.Original/Direct/Reference
2.A
3.C
4.D
5.B
6.E
7.F
It's very hard to guess what A is from the specification of the opamps. In this difficult circuit, A displays some good quality (sonic drive) as well as some bad quality (noise). This sonic must relate with output current or driveability, so JRC4560 matches the condition.
Output current of 4560 is 50mA, outperforms LM4562 which is only 26mA. But the noise of RC4560 is terribly 1200nV compared to others which are no more than 18nV!
1.Original/Direct/Reference
2.A
3.C
4.D
5.B
6.E
7.F
It's very hard to guess what A is from the specification of the opamps. In this difficult circuit, A displays some good quality (sonic drive) as well as some bad quality (noise). This sonic must relate with output current or driveability, so JRC4560 matches the condition.
Output current of 4560 is 50mA, outperforms LM4562 which is only 26mA. But the noise of RC4560 is terribly 1200nV compared to others which are no more than 18nV!
Thanks for listening to these Gus 
Giving them a whirl on the ESL's should be extremely revealing I would imagine.
Giving them a whirl on the ESL's should be extremely revealing I would imagine.
Mooly,
Next time please include not only direct cd rip original.
Please include your dac adc loop without DUT.
Thx, sika
Next time please include not only direct cd rip original.
Please include your dac adc loop without DUT.
Thx, sika
Not at all. See what audible differences you can winkle from it vs any of the files.
Interesting ! You think 😉
Given the very low impedances involved (CD player output impedance <100 ohm and the opamps all configured for unity gain) I felt a simple breadoard layout was sufficient. The 'dirty ground' was a connection back to the PSU. Certainly nothing worse than a typical commercial layout, in some ways better because there is no other circuitry pulling current (signal or otherwise) through the ground trace.
Fair comment.
The reason I did not was simply that it would be instantly identifiable as 'different' without having to listen to the files.
Hi Mooly.
Your avatar wasn't made by Denise Wilton was it? It looks exactly like some of her work!!
Your avatar wasn't made by Denise Wilton was it? It looks exactly like some of her work!!
Rule 1: Start by defining the specific question an experiment is supposed to ask. The question Arny wants to ask is "Do any of these files sound different than the reference?" That's not the question Mooly is asking (which is "Do any of these files sound different from any of the others?"), so Mooly's experiment is, not surprisingly, different than one which would answer Arny's question.
It really does help to be explicit.
- Status
- Not open for further replies.