How does % work ? Like 4 people voted 2 times ? Or aa doesn't count ?
I used same short treble part and sound of decay when comparing.
I used same short treble part and sound of decay when comparing.
I'm not following you 🙂 You should only be able to vote once, and aa has collected four of the total votes cast.
Hi Mooly,
I installed Foobar and the ABX component late yesterday, and played with it a little. Now its easy to see why you describe the testing the way to do. Also, agree that cymbal shimmer seems to be the easiest thing to discriminate. It's much harder than I thought it would be, but I can see how in the right, relaxed state of mind and with a lot of concentration it should be possible to distinguish aa from any of the others.
However, when not trying to A/B/X test and only listening by itself,
(1) its much easier, and
(2) sensitivity to small differences seems higher.
More than one reason for that, but my sense is that
(1) operating the program is a distraction, and that it doesn't seem to be able to loop automatically is not helpful, but the undocumented hotkeys are at least are easier than mousing,
(2) I have been trying to think of some way to describe the next factor, but I'm still working on it. At the moment, maybe I could say that listening without testing it like kind of like listening to spoken language you speak and understand fluently from someone with a very poor accent. If you relax and listen carefully, you can just make out what it being said.
But, if it was a language you were not fluent at, and you weren't sure of the word definitions, so you had to think about dictionary meanings and try to figure out what the best translation would be in realtime, and simultaneously recognize the words with the heavy accents, that would be more like the concentration of testing.
Or, to put it another way, for me when I latch onto hearing a particular distortion, then it's easy to stick with it as a file plays, but with the 4 four buttons in Foobar ABX a,b,x,y each time I push a different button I have re-find the distortion and latch onto it again. Only with great effort, concentration and relaxation, all together at once, is it possible to briefly memorize the small differences in distortion, latch onto the distortion playing, and put the two together at once to recognize the distortion with near-certainty (maybe a separate process) enough to choose correctly while being distracted all the time by looking at and interacting with the program.
In other words, I think I might be able pretty soon to do about what you have done with the aa file and one other. However, while I can barely settle on the differences between non-aa files when not testing, I doubt that with the extra concentration of ABX testing whether it might either be impossible, or take quite a lot of practice to develop better skill at it.
I installed Foobar and the ABX component late yesterday, and played with it a little. Now its easy to see why you describe the testing the way to do. Also, agree that cymbal shimmer seems to be the easiest thing to discriminate. It's much harder than I thought it would be, but I can see how in the right, relaxed state of mind and with a lot of concentration it should be possible to distinguish aa from any of the others.
However, when not trying to A/B/X test and only listening by itself,
(1) its much easier, and
(2) sensitivity to small differences seems higher.
More than one reason for that, but my sense is that
(1) operating the program is a distraction, and that it doesn't seem to be able to loop automatically is not helpful, but the undocumented hotkeys are at least are easier than mousing,
(2) I have been trying to think of some way to describe the next factor, but I'm still working on it. At the moment, maybe I could say that listening without testing it like kind of like listening to spoken language you speak and understand fluently from someone with a very poor accent. If you relax and listen carefully, you can just make out what it being said.
But, if it was a language you were not fluent at, and you weren't sure of the word definitions, so you had to think about dictionary meanings and try to figure out what the best translation would be in realtime, and simultaneously recognize the words with the heavy accents, that would be more like the concentration of testing.
Or, to put it another way, for me when I latch onto hearing a particular distortion, then it's easy to stick with it as a file plays, but with the 4 four buttons in Foobar ABX a,b,x,y each time I push a different button I have re-find the distortion and latch onto it again. Only with great effort, concentration and relaxation, all together at once, is it possible to briefly memorize the small differences in distortion, latch onto the distortion playing, and put the two together at once to recognize the distortion with near-certainty (maybe a separate process) enough to choose correctly while being distracted all the time by looking at and interacting with the program.
In other words, I think I might be able pretty soon to do about what you have done with the aa file and one other. However, while I can barely settle on the differences between non-aa files when not testing, I doubt that with the extra concentration of ABX testing whether it might either be impossible, or take quite a lot of practice to develop better skill at it.
Maybe you missed that, it was posted again just last week from, let's see, 25yr. ago. You say you're familiar with what we do and actually ask if we've ever considered matching the input impedances on an op-amp or that we might not realize it doesn't work on a CFA? Are you serious? I heard NS10's at the Yamaha factory and couldn't wait till they stopped.
As mentioned LIGO is maybe 10 or more orders of magnitude more sensitive than the stylus vinyl interface, all IC's with their "micro-sized" components.
I did see some WIMA capacitors.
The % calculation seems to be off. By purposely selecting the wrong answer, I was able to score 2/10. Now, 5/10 would mean 100% that I was guessing, but for 2/10 it said 90% guessing.
So, it seems to assume that all wrong answers represent guesses, but that can't be correct because guessing should give 50% right.
A score of 2/10 should indicate that one is somewhat biased toward inverting the correct selection. In other words, someone is so confused as to get recognition completely backwards.
So, it seems to assume that all wrong answers represent guesses, but that can't be correct because guessing should give 50% right.
A score of 2/10 should indicate that one is somewhat biased toward inverting the correct selection. In other words, someone is so confused as to get recognition completely backwards.
Last edited:
I agree. The concentration is fading with increasing number of trials. Especially in case like here where the differences are tiny. In case of huge differences, like high noise/low noise in the same part of music, you can easily get 100/100 and repeatedly. Though it is boring then 😀
Mentally and it (Foobar) is very demanding I find, although you will get more comfortable using it with practice.
I tend to hover the cursor over say 'Play A' and with my eyes closed just keep hitting the button. Over and over and over. For a reality check I almost subconsciously move over to 'Play B' and hit that asking was it different.
If I was evaluating an opamp for use in a particular application I would normally leave it in place for days or weeks and keep asking if I was happy with the result. I did that with the OPA2134 vs the LM4562 in the preamp section of my amp. After several weeks I decided I just didn't like the OPA2134 in that application (inverting gain stage of around -2).
With Foobar and I usually have the start and stop times set over the part of interest and have the 'retain position' box unticked. Sometimes the sample I'm listening to might be just a couple of seconds. Its not a natural way to listen, that's for sure.
I still haven't figured all the words out in this one 😉
I tend to hover the cursor over say 'Play A' and with my eyes closed just keep hitting the button. Over and over and over. For a reality check I almost subconsciously move over to 'Play B' and hit that asking was it different.
If I was evaluating an opamp for use in a particular application I would normally leave it in place for days or weeks and keep asking if I was happy with the result. I did that with the OPA2134 vs the LM4562 in the preamp section of my amp. After several weeks I decided I just didn't like the OPA2134 in that application (inverting gain stage of around -2).
With Foobar and I usually have the start and stop times set over the part of interest and have the 'retain position' box unticked. Sometimes the sample I'm listening to might be just a couple of seconds. Its not a natural way to listen, that's for sure.
I still haven't figured all the words out in this one 😉
It turns out the a,b,x,y keys are all hot, which seems preferable to hovering the mouse. Easy to do with eyes closed.
EDIT: Agree about a preference for working with only a short segment, repeating play over and over, and 'retain position' unchecked. The repeat process is why I wish it could loop automatically.
EDIT: Agree about a preference for working with only a short segment, repeating play over and over, and 'retain position' unchecked. The repeat process is why I wish it could loop automatically.
Last edited:
Not sure if it's the worse the system or that headphones do a better job of isolation and depending on the response curve of the phones, will tend to emphasize the midband, where we're most sensitive. More balanced phones may not be as effective at highlighting the differences, though.
I'll give it a go after work today. Not that I have the best performing "system" but it'll be quiet here and hifiman RE-400's are pretty good for detail. Then again, I'm pretty deaf.
I haven't been here for a couple of years. So yes I missed last weeks post.
I'm chatting to a diverse audience on this thread. I read you data sheet notes on the AD797, they were your's correct?
But PMA and others in this discussion have different levels of knowledge. So I would assume you know, but possibly not the others.
Then to wonder why one would waste time with op-amps for serious audio then one would have to wonder about the level of knowledge.
It seems strange that I have to explain logical thinking to someone who should be strong in that area.
What's logical about it? PMA, jcx, Bob Cordell among others here are extremely experienced and knowledgeable. You have a new take on being insulting, almost so much so I could believe it's a put on.
I guess you do. Add one to the list.
Sure you can prove any sort of "logical" thing you want if your premises are completely bonkers. I'd avoid insulting people by the way -- hubris is pretty painful and all.
- Status
- Not open for further replies.