John Kenny SABER DAC

[Guest Muon]

Umm...Bill you said "It is not 100% understood how different bit perfect players sound different"

Two different developers will code a media player differently, the code won't be identical, and there approach will differ, that's enough to make them sound different.

I'm definitely not having a go at the producers of the software here.

[jkeny]

Umm...Bill you said "It is not 100% understood how different bit perfect players sound different"

Two different developers will code a media player differently, the code won't be identical, and there approach will differ, that's enough to make them sound different.

.......

Wow, that's a loose statement - can you explain how these differences in code makes the output sound different (presuming both outputs are bit-perfect)?

[sir sanders zingmore]

I was wondering the same thing myself. If the output of the 'black boxes' are identical why should it matter what happens inside those boxes?

[hochopeper]

Wow, that's a loose statement - can you explain how these differences in code makes the output sound different (presuming both outputs are bit-perfect)?

One thing I have wondered, having only very rudimentary experience with writing software is: how do you write a test that checks the output for bit-perfectness?

[Dr X]

It is not 100% understood how different bit perfect players sound different but blind tests show they do.

What blind tests?

[bhobba]

What blind tests?

The blind tests that I have done and others like Kdoot have posted about. I will not get into a silly discussion on methodology etc because believe me a blind test is not required - it is 100% obvious and easy to spot - no blind test required. For example when checking out the JK-DAC a guy that took it away was shocked how easy it was to spot using my Mac-mini. He could not hear as big a difference with his windows machine. My guess is that's because many of the true audiophile players such as my favourite, Audirvana, are Mac specific. If you can't hear it, to be blunt your system is simply not revealing enough.

But we are getting into double blind testing territory which the moderators have deemed, correctly IMHO, to be the province of the great debate section. I however will not be contributing to it because anyone that hears my system will spot it immediately - no blind testing or such required. If you choose to disbelieve it that's fine - its your prerogative .

Thanks

Bill

Edited September 14, 2011 by bhobba

[Guest Muon]

Wow, that's a loose statement - can you explain how these differences in code makes the output sound different (presuming both outputs are bit-perfect)?

I'm talking about software media players!

And that's the content I have been responding to all along *sighs*

[bhobba]

Wow, that's a loose statement - can you explain how these differences in code makes the output sound different (presuming both outputs are bit-perfect)?

Hi John

When you think about it a bit its not hard to see how it is possible. For example one person may read from the disk and output it while reading from the disk. The disk hash can then travel down the USB cable and interfere with the working of the DAC just like Jitter does. A second person may decide, like Audirvana does, to do everything in memory. It reads it into memory, decodes it into memory, then plays it from memory. Less disk activity while playing so less hash to go down the USB cable. They are both bit perfect in the sense of if you attach a recording device on the end of it they will both register exactly the same ones and zeroes. But a DAC is not a recording device - it depends a lot on timing information that can easily be mucked up. While the async you use makes a big difference even with that it is very easy to hear differences between bit perfect players.

One of the problems here is guys at my level, and the majority of peoples here level, are not DAC engineers and we have a very limited understanding of the detail. At our level we would expect two different bit perfect players to sound the same. But myself and tons of people I have demonstrated it to can easily, and I do mean easily, it is far from subtle, hear night and day differences. This is left us scratching our heads and have come up with some tentative explanations such as the hash hypotheses I am putting forward. That is what I mean by it is not known 100% for sure what's going on. Trouble is the engineers who design this stuff, and the software developers of stuff like Amarra, have a marketplace advantage in not releasing the exact detail so we will probably never really know. It may even be they are telling porkies when they say their product is bit perfect - I don't think anyone really knows.

Thanks

Bill

Edited September 14, 2011 by bhobba

[Guest Muon]

You are on the right track, Bill.

But it goes further than that, for example I compared playback between cPlay and jplay and while both load into, and read from RAM they sound different. Not black and white different but still different.

Edited September 14, 2011 by datafone
grammar

[Cyber_Murphy]

I compared playback between cPlay and jplay and while both load into, and read from RAM they sound different. Not black and white different but still different.

In your opinion, which one was better... I only ask because I am currently deciding to replace Media Monkey with either cPlay or jplay??

Ta

CM

Edited September 14, 2011 by Cyber_Murphy
typo

[bhobba]

But it goes further than that, for example I compared playback between cPlay and jplay and while both load into, and read from RAM they sound different. Not black and white different but still different.

Exactly. And as a person that has heard my system you know what I am saying is true - night and day differences exist but as you point out the differences between some plays is not so obvious - but still exists. For those players ie those that both do it from memory, the reason is far from obvious - to me at least anyway - in fact I do not even have a working hypothesis for that.

Thanks

Bill

[nofixedaddress]

One thing I have wondered, having only very rudimentary experience with writing software is: how do you write a test that checks the output for bit-perfectness?

Easy, Feed the software with an audio file that has already been analysed and the bits known. Then compare that to the bit values being spat out by the computer and software. I'm fairly sure that is how the Weiss DACs do it (Proft can enlighten us more on the that).

NFA

Edited September 14, 2011 by nofixedaddress

[Dr X]

The blind tests that I have done and others like Kdoot have posted about. I will not get into a silly discussion on methodology etc because believe me a blind test is not required - it is 100% obvious and easy to spot - no blind test required. For example when checking out the JK-DAC a guy that took it away was shocked how easy it was to spot using my Mac-mini. He could not hear as big a difference with his windows machine. My guess is that's because many of the true audiophile players such as my favourite, Audirvana, are Mac specific. If you can't hear it, to be blunt your system is simply not revealing enough.

But we are getting into double blind testing territory which the moderators have deemed, correctly IMHO, to be the province of the great debate section. I however will not be contributing to it because anyone that hears my system will spot it immediately - no blind testing or such required. If you choose to disbelieve it that's fine - its your prerogative .

Hang on a minute, you brought bind testing up not me.

You're post lacks clarity...first you say blind tests that you've done and Kdoot...and then you say they are not required.

So did you actually do one or some or none at all?

If you're not prepared to provide all of the details, then I think you should refrain from statements like "It is not 100% understood how different bit perfect players sound different but blind tests show they do." because it's simply false and misleading.

[bhobba]

Hang on a minute, you brought bind testing up not me. You're post lacks clarity...first you say blind tests that you've done and Kdoot...and then you say they are not required. So did you actually do one or some or none at all? If you're not prepared to provide all of the details, then I think you should refrain from statements like "It is not 100% understood how different bit perfect players sound different but blind tests show they do." because it's simply false and misleading.

Yes I did - and I explained it was blind tests me and others have done. Kdoot has even posted about them. That really should have been the end of it. However I said, and I stand by it, they were not required - it was no blindingly obvious. I will not be entering into a discussion of methodology, nor do I think it is appropriate for this thread and should be taken up in the great debate section. But I will not be contributing to it.

I am basing statements like "It is not 100% understood how different bit perfect players sound different but blind tests show they do." on the blindingly obvious differences people have heard but to be sure some, including me, have subjected them to blind tests. IMHO it is false and misleading to suggest in a thread discussing subjective differences that it should be held to a standard different than subjective impressions.

I am really scratching my head why the details of those blind tests is of any real importance. In my case it was with a group of people and a guy switched randomly between playing with different players. If I recall correctly Kdoot did it with his wife doing the switching. In both cases it was reliably picked 100% of the time. I suspect after giving such detail the moderators may now consign the post to the great debate section where it properly belongs. I hope not and because of that regret mentioning it - its just I though it important enough for people to understand we were so surprised we did go to that length - but with the benefit of hindsight I am sorry I mentioned it.

Thanks

Bill

Edited September 14, 2011 by bhobba

[Dr X]

Yes I did - and I explained it was blind tests me and others have done. Kdoot has even posted about them. However I said, and I stand by it, they were not required - it was no blindingly obvious. Your posts are of the type that doubts what people hear unless it is done with the methodology you will accept. That is your precognitive just as it it my precognitive to say I do not agree the methodology that you want is required - and I don't. I will not be entering into a discussion of methodology, nor do I think it is appropriate for this thread and should be taken up in the great debate section. But I will not be contributing to it.

I am basing statements like "It is not 100% understood how different bit perfect players sound different but blind tests show they do." on the blindingly obvious differences people have heard but to be sure some, including me, have subjected them to blind tests. IMHO it is false and misleading to suggest in a thread discussing subjective differences that it should be held to a standard different than subjective impressions.

Thanks

Bill

I don't think it's logical to mention blind testing and then not provide any details on what you did. I think you should stop it because it's misleading and only seems to cause arguments.

I've got no problem with you sharing your subjective sighted impressions about every DAC and digital transport in the universe...go for your life buddy!

[naggots]

Bill, have you tried the Biperfect app with your mac mini? How does it compare to Audirvana in your system?

[bhobba]

I don't think it's logical to mention blind testing and then not provide any details on what you did. I think you should stop it because it's misleading and only seems to cause arguments. I've got no problem with you sharing your subjective sighted impressions about every DAC and digital transport in the universe...go for your life buddy!

Fair enough and as I said in hindsight I should not have mentioned them at all. Thanks very much for your understanding in not pursuing the matter in this thread as I believe the moderators will shut it down pretty quickly.

I just noticed you got an early version of my post where I said 'your posts are of the type that doubts what people hear unless it is done with the methodology you will accept. That is your precognitive just as it it my precognitive to say I do not agree the methodology that you want is required - and I don't.'. I removed it because I realised it was unfair. But because you got the version with it I apologise for saying it because it is unfair. I did mention blind tests and it is fine to ask about detail from someone doing so. I should not have done so in the first place.

Thanks

Bill

Edited September 14, 2011 by bhobba

[bhobba]

Bill, have you tried the Biperfect app with your mac mini? How does it compare to Audirvana in your system?

Hi Naggots

Yes I have and it is pretty good and in fact sounds a lot like Audirvana. But I find Audirvana with SRC upsampling, even though it really thrashes my CPU, to sound just so liquid, fluid and entrancing it is the one I seem to always come back to. It also seems to be DAC dependant. On the Metrum and other DAC's its a toss up between Pure Music and Audirvana - Audirvana sounds more crystalline, pure, exact and neutral - especially in the treble which is really clean - Pure Music seems to make detail more obvious and on some DAC's sounds darn good - as good as Audirvana - but different. Yet on the PDX Pure Music sounds not a lot better than Itunes to me, and others think it sounds worse. Weird hey?

Thanks

Bill

Edited September 14, 2011 by bhobba

[lebowski]

Bill, interesting you say BitPerfect sounds a lot like Audivarna because that's not what I found. To me, and in my system Audivarna clearly sounds better, more relaxed and balanced sounding top to bottom. BitPerfect has more pronounced bass and with less detail. I can't hear the same level of decay with BitPerfect as I can with Audivarna. Using Audivarna and JK hiFace into my DAC is still the closest I can get to analogue sound without a TT.

Edited September 14, 2011 by Lebowski

[hochopeper]

Easy, Feed the software with an audio file that has already been analysed and the bits known. Then compare that to the bit values being spat out by the computer and software. I'm fairly sure that is how the Weiss DACs do it (Proft can enlighten us more on the that).

NFA

Exactly! So what I suppose my question was leading to is that, we have a variety of levels of bit-perfectness. Yet we are only testing this with our ears with many different stages to consider.

If the software is the creating a problem then we need to develop a method for testing that! Their should be some method to intercept the output of the software before it goes to the operating system sound driver. I've not used it before but something like the software package JACK should be able to achieve this with some plugins I think.

This way we can have:

1. known input

2. a tested output prior to OS sound driver

3. output from DAC

When a software developer writes a test it will be for their own specification of what is bit-perfect. After this the signal then runs through a variety of computer/transport hardware until it interfaces with the DAC, this is for me an area that is more difficult to test. We should be able to test for either side of each black box in the chain and see at what stage the differences are introduced.

I'm now intrigued enough that when I next get some free time I'll try to work something out to objectively compare software players input/output.

I may have oversimplified some of the problems and have now gone way off topic of jk's DAC.

[bhobba]

Bill, interesting you say BitPerfect sounds a lot like Audivarna because that's not what I found. To me, and in my system Audivarna clearly sounds better, more relaxed and balanced sounding top to bottom. BitPerfect has more pronounced bass and with less detail. I can't hear the same level of decay with BitPerfect as I can with Audivarna. Using Audivarna and JK hiFace into my DAC is still the closest I can get to analogue sound without a TT.

Hmmm. Interesting. Have you done the comparison lately because bit perfect is being updated all the time.

However I fully agree with you in that Audivarna is the best. I have tried all sorts of stuff and I always come back to it, especially upsampled using SRC - it is just so liquid, fluid and analogue like - the detail is just so natural and not overly obvious - yet when you listen you realise more of it is actually there. It makes stuff I never really liked before such as Barry White because its sibilance sound good - before it blead my ears and was one of the tracks I used to demonstrate the sibilance issues with the WFS. I was just listening to it a minute ago and really enjoyed it - not so with other players.

Thanks

Bill

Edited September 14, 2011 by bhobba

[nofixedaddress]

The difficult section (after the software) is what the Weiss DAC includes in the way it tests the bits. The DAC itself tests the incoming stream being created from the Weiss test file as played by the software and compares it to the file stored in the DAC. I least that is how I understand it to work.

I guess your problem will be, how to capture and and analyse the outgoing stream from the PC in a software environment without developing hardware to do that.

From memory Proft still reported hearing differences between software even when the Weiss test passed both as bit perfect - then again I could have KRAFT memory.

[Guest Muon]

In your opinion, which one was better... I only ask because I am currently deciding to replace Media Monkey with either cPlay or jplay??

Ta

CM

To tell the truth I can't really say.

But you need to remember that cPlay is ideal when used with cMP, a XP based built for media PC http://www.cicsmemoryplayer.com/index.php and I have only used it under a general purpose win7 PC.

[Guest Muon]

Exactly. And as a person that has heard my system you know what I am saying is true - night and day differences exist but as you point out the differences between some plays is not so obvious - but still exists. For those players ie those that both do it from memory, the reason is far from obvious - to me at least anyway - in fact I do not even have a working hypothesis for that.

Thanks

Bill

Umm..hang on Bill, wasn't the night/day difference at your place due to the OffRamp (hardware) in the chain then not?

But I agree that the difference between something like jplay/cPlay and other media players like Winamp is huge!

Edited September 14, 2011 by datafone

[lebowski]

Hmmm. Interesting. Have you done the comparison lately because bit perfect is being updated all the time.

However I fully agree with you in that Audivarna is the best. I have tried all sorts of stuff and I always come back to it, especially upsampled using SRC - it is just so liquid, fluid and analogue like - the detail is just so natural and not overly obvious - yet when you listen you realise more of it is actually there. It makes stuff I never really liked before such as Barry White because its sibilance sound good - before it blead my ears and was one of the tracks I used to demonstrate the sibilance issues with the WFS. I was just listening to it a minute ago and really enjoyed it - not so with other players.

Thanks

Bill

I haven't updated it since I first downloaded it about 2 weeks ago. Yes, it's good but ultimately loses out to Audivarna in my system.

You've hit the nail on the head with - the detail is just so natural and not overly obvious - yet when you listen you realise more of it is actually there.

This is also what the JK hiFace does, and when I use this combination the music sounds just so right.

Cheers

Mark