I know of one user with a Benchmark, and I don't think I heard anything from him explicitly.
What I know is that each driver of which I can expect to report decently, all report 32 bits (at feeding with 24).
All 'n all, I did not find one application that would swallow the 24 bits (but i.e. a Juli@ "SAYS" it can do 24. It only actually works with 32 though).
Please keep in mind that everything gets more vague with a soundcard in between the PC and the DAC, that soundcard actually reporting *and* being capable of conversions to the DAC without my notice. The Fireface is one of them (needing 32 bits btw).
What I know is that each driver of which I can expect to report decently, all report 32 bits (at feeding with 24).
All 'n all, I did not find one application that would swallow the 24 bits (but i.e. a Juli@ "SAYS" it can do 24. It only actually works with 32 though).
Please keep in mind that everything gets more vague with a soundcard in between the PC and the DAC, that soundcard actually reporting *and* being capable of conversions to the DAC without my notice. The Fireface is one of them (needing 32 bits btw).
XXHE said:
abzug said:
Thanks guys.
I may try to do this if I can find the time and the inclination at some point.
For now, I'm quite happy to simply enjoy listening to music... and leaving the development of PC playback software to experts like yourselves.
Wingfeather said:
What the hell are you talking about? XXHighEnd sounds better to me than Foobar with ASIO. OK, I'm basing my evaluation on phenomenology and not empiricism (at this point at least). But my evaluation has nothing to do with belief...
Mani.
My point was that no valid reasons have been given for which XXHE can sound better than any other bit-perfect player. Features of XXHE have been hinted at, such as an improvement in output jitter, but hinting is all that has been done. The fact that there is simply no mechanism by which a piece of software can have an effect on output jitter in a PC (or at least, nobody has shown a convincing one) is being totally ignored.
The name of the software and the inclusion of GUI controls which supposedly affect the player's performance (the Q1 slider? What exactly does that do? Is it even connected to anything?) is highly likely to increase people's expectations of sound quality. As are the multitude of different settings you're supposed to try to squeeze the last bit of performance out.
Back when I used to believe in expensive cables I would have sworn they made things sound different. I mean, I genuinely heard differences. Now that I can't see any possible reason for them to be better than the cheap ones, I swear they sound the same. Doesn't prove me right one way or the other, but it does show that one's expectations are very important and it is possible to be swayed by them.
Yeah, I think the devil is in the details. Unless one has a really in depth understanding of the issues with coding, moving the information around, etc, the only way to judge is by examining the end product with ones ears. Theoretical stuff like 'bits is bits' tend not to hold up in the real world.
Many of us have an understanding of the issues, and I say you're wrong. I'm a software developer in the day and a hardware developer in the evening, and I'm backing the "bits is bits" comment. If it sounds different (in blind test of course), then something is different--and there are only two things if power supply etc. noise is properly decoupled--bits and jitter. If bits are the same and interface jitter is eliminated through the use of an asynchronous connection, then all is the same, and any claim to the contrary is extraordinary, tantamount to saying the tooth fairy is real. Extraordinary claims require extraordinary evidence.cuibono said:
Wingfeather said:
Yes, I see what you're getting at.
Well look, you believe that there is no mechanism by which a piece of software can have an effect on output jitter in a PC.
In which case, you have the expectation that XXHighEnd should not sound better than say Foobar with ASIO, right?
So why not test it? Just give the demo version a listen (try to use Engine #3 on a Vista machine).
According to your logic, your expectations will simply be met.
If they're not, then you'll be in the same boat as me... trying to understand why two pieces of supposedly bit-perfect software sound different.
But in any event, I'd be interested in hearing your findings.
For my own part, I didn't believe that a PC could out-perform my CD transport. But I couldn't ignore the evidence. I didn't believe that XXHighEnd could sound better than Foobar with ASIO, but I refuse to ignore the evidence that it does.
Mani.
Hehe I don't think that azbug said that software couldn't influence jitter, but based upon my own expression that *I* don't claim it is jitter which changes things, azbug now says that combined with all being bit perfect things can't be different.
Besides that only few people will believe that software can influence jitter. But in fact this is another subject.
Guys, this is too complex. If you'd rather not believe it, then just don't. Debating it without listening is not much useful I think, and I don't tell you to have a listen. I really really don't care, and the only thing I care about is creating a player for those who do care. Remember, I created the stupid thing just for my self at first, but then people starting to want it *including* further requirements (the only reason why it costs afew bucks btw. I think I have spent some 3000 hours on it by now).
I don't say I can't explain things, but I do say that differences for SQ can be applied easily. Mind you, all of the changes I apply are reasoned out in advance, many times anounced in advance (already for people hearing "wrong" things) and so far it always worked out for the objective in mind.
Now, in order to call me crazy if you like, with the last version (0,9u opposed to 0.9t) I wanted more metal in cymbals, and it just worked out like that.
Might you measure a different jitter signature or something else, one thing I guarantee : it is still bit perfect (and that can be measured too of course).
If you'd follow the development from the beginning, you'd have seen that I hunt for one stupid thing only : better 1:1 representation of what's in the file, with the whole chain in mind (!). That this by itself leads to a different jitter signature is kind of logic because we can't think of anything else. However, I think there is more. This is just honest.
Peter
PS: If someone has some affordable jitter measurement device (say < 10 ps) available that allows my own software to play when measuring, I will be happy to buy it.
Besides that only few people will believe that software can influence jitter. But in fact this is another subject.
Guys, this is too complex. If you'd rather not believe it, then just don't. Debating it without listening is not much useful I think, and I don't tell you to have a listen. I really really don't care, and the only thing I care about is creating a player for those who do care. Remember, I created the stupid thing just for my self at first, but then people starting to want it *including* further requirements (the only reason why it costs afew bucks btw. I think I have spent some 3000 hours on it by now).
I don't say I can't explain things, but I do say that differences for SQ can be applied easily. Mind you, all of the changes I apply are reasoned out in advance, many times anounced in advance (already for people hearing "wrong" things) and so far it always worked out for the objective in mind.
Now, in order to call me crazy if you like, with the last version (0,9u opposed to 0.9t) I wanted more metal in cymbals, and it just worked out like that.
Might you measure a different jitter signature or something else, one thing I guarantee : it is still bit perfect (and that can be measured too of course).
If you'd follow the development from the beginning, you'd have seen that I hunt for one stupid thing only : better 1:1 representation of what's in the file, with the whole chain in mind (!). That this by itself leads to a different jitter signature is kind of logic because we can't think of anything else. However, I think there is more. This is just honest.
Peter
PS: If someone has some affordable jitter measurement device (say < 10 ps) available that allows my own software to play when measuring, I will be happy to buy it.
abzug said:
Correct. Abzug never said that software couldn't influence jitter...
Wingfeather said:
... but Wingfeather suggested otherwise.
But one thing that we can all agree on, I think, is that it is possible to get a bit-perfect data stream from a PC running Vista and that if done well, the results can be stunning
Mani.
JimOfOakCreek said:
ackcheng said:
Part of the key is in the "improve on 100% perfect?", because bit perfect is only for a relative small amount related to perfect. And I say relatively small, because of the huge improvements on SQ realized over time, so far.
Very indirectly you could say that what is in "a" way in the file, should arrive exactly the same at the DAC's input (this is the bit perfect part) and should arrive exactly the same at the speaker (even better would be out of the speaker).
It is this part I work upon.
Things get funny when you see that some weeks ago I (found and) anounced that CD playback always had been wrong since its invention, that I applied something to make it "good" as should (a better 1:1 playback means), and that the change to SQ had never been that huge for the better. All is still bit perfect ...
Only yesterday I have found something else which looks very very wrong to me in CD data, but in order to attack *that* it needs DSP ... (which I actually refuse).
Like with the last example and the before one, all is such a huge pile of, say, thinking differently, that it takes years to come to the proper conclusions. This is one reason why I do not eleborate on things that much : I am just not sure what actually is happening.
If I create playback software, and suddenly it comes to me that I and my room are not subject to the dreaded standing waves anymore, at that moment I can only shout outloud that, well, this player doesn't show standing waves. Now, almost two years later I'm still working on the why (please don't jump in the standing waves subject here, I'm just trying to make clear how difficult things are).
From the beginning I say that a zillion things are wrong with audio playback as a whole, and I'm just counting down on solving them all.
Looking at the CD data, it is not only the "bits" needing to get into the DAC unmangled, it is also the, say, transients which must be followed by everything and all. Is it ? careful now, because were talking digital data in the base, and these are just squares.
Mind you, those (verry tiny) squares leave the DAC the same as they entered the DAC, if the DAC only can follow. For the amp the same, for the speaker the same. So when everyting is represented 100% 1:1 we'd hear those squares. And in fact, currently I do, and things get over the hill.
When those squares float in air, they produce harmonics already right in air. The harmonics of tiny squares, which actually were nice sines (or anyway a smooth analoge line).
I can't repeat it enough : the only thing I do is better 1:1 representation of the CD data, which actually comes down to better timing in the DAC, that by itself implying different jitter.
Please note I'm so much against/near/in the data the DAC actually uses itself, that it may need som imagination of possibilities, compared to ASIO, Kernel Streaming and let alone DSound. You could say that if I want a byte to arrive some later, I can do it. And again, what goes in is still a stream of bit perfect data.
I hope this helps a bit.
Peter
And what else is there that software can affect, if there's no modification of the bits? That is the question I'm asking--other than getting through the data unmodified, and potentially influencing jitter through things such as system load or whatever, what other thing is there left that the software can affect? If there really is something, then you must know what it is because of your claim that your player improves it.XXHE said:
XXHE said:
I don't understand any of this. Can you answer my question directly and to the point?
The computer streams packets of data through the USB . Each packet has a header with information that defines the file type, the bit rate, etc. The DAC reads the header and adjusts for the bit rate and adjusts for the file type.
Does your software improve on that? How does your software improve that?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.