A first test and some food for audio thought

[BobH]

I, for one, will follow the thread with interest. I currently use JRiver in preference to iTunes for no reason other than it will auto switch sample rates and has half-decent library management. I've used all of the major players, and most of the minor ones, and never had a problem with the sound quality of any of them.

 

[audiozorro]

My criticism is that you may be attacking the problem on the back end instead of attacking the problem on the front end. Which is why I think the developers of Amarra, Pure Music or XXHighEnd are uniquely able to address their programming algorithms and codes that make their software sound different. Unfortunately for us, because of the proprietary and competitive nature of commercial software, I do not expect much disclosure here. So I assembled the below after a quick search on the Internet.

 

Amarra - Essentially, what Amarra does is play back audio files using iTunes as its GUI. It skips the playback portion of iTunes and uses its own mathematical equations to compute the audio data. Sonic Studios claim bit perfect output! Is it? Well, while I will admit I do not have the correct measuring equipment to prove this, I will say that it does sound quite different. [Positive Feedback Online – Issue 44]

 

Pure Music - Pure Music goes even further than previous player software, offering advanced features such as Memory Play, a 64-bit internal signal path, and all audio is rendered and played via Pure Music's low CPU overhead, audiophile quality 64-bit resolution playback engine. (Note: 64-bit resolution audio is an entirely different thing from the 64-bit memory addressing now available in Apple's OS X operating system.) [info from Pure Music website]

 

XXHighEnd - XXHighEnd is a memory player, with an attention to low software latency and incorporates several parameter settings that influence going jitter, dictates the interval for influencing the jitter, determines the strongness of the influence on the jitter and sets the resonance of the influence. [info from XXHighEnd website]

 

The above is subject to corrections from the developers, but could any of the above factors influence the sonics of these players from iTunes? Are there other factors? As Mani said, it looks like we’ve all got a lot to learn yet.

 

Personally, and especially since I’m not in the software development business, I don’t really care to know all the details. The important thing to me is that I hear the differences and my focus tends to be finding and using the best sounding software players.

 

If you don’t hear any sonic differences, then perhaps your first step should be to seek and hear the differences that others hear. You may have to hear these differences in someone else’s system or at a audio show. If you hear these sonic differences, you may develop a passion for investigating the seen or in this case heard, instead of the invisible or unheard.

 

Forgive the poor example, but why would you search for gold in your backyard if you didn’t believe there was any? However if you believed there was a fortune in gold you probably would be tireless in your search. I think the proper scientific methods are first observe, hypothesize, then test. Perhaps I am misunderstanding your intentions but it seems that you have not observed any sonic differences in the software players, you don’t believe there are any sonic differences, your hypothesis is that there is no sonic differences. So now you are conducting tests to prove that the differences that you haven’t heard and don’t believe in, don’t exist.

 

This kind of reminds me when one of my audio buddies tried to tell me that power cords can make a sonic difference in your system. For years I believed this to be nonsense, until I heard it myself in my system. A painstaking exploration ensued where 7 out of 10 cords sounded identical, 2 sounded better and only 1 sounded best for a particular audio component. Fortunately the cords were loaners, but the trying and comparing 40 different cords for each audio component is not something I call fun. In some ways I was happier when I thought there weren’t any differences, especially as is often said, there are bigger fish to fry.

 

 

[cfmsp]

AZ,

 

 

"... So now you are conducting tests to prove that the differences that you haven’t heard and don’t believe in, don’t exist."

 

That's not how I interpreted Barry's comments at all.

 

 

 

clay

 

 

 

[bdiament]

Thank you Clay.

 

Only someone who doesn't know me could suspect such.

I don't know exactly why but I'm reminded of Yoda's response to Luke on one of their journeys. When Luke has to enter some strange area on the planet they're on, he asks Yoda what he'll find there. Yoda says something like "whatever you bring in with you". ;-}

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[bdiament]

Hi audiozorro,

 

"...Perhaps I am misunderstanding your intentions but it seems that you have not observed any sonic differences in the software players, you don’t believe there are any sonic differences, your hypothesis is that there is no sonic differences. So now you are conducting tests to prove that the differences that you haven’t heard and don’t believe in, don’t exist...."

 

Thank you for your feedback. For what its worth, you are at the very best, completely misunderstanding my intentions. And you are attempting to tell me what I believe, which is just plain silly, since you don't really have any idea what I believe and it is clear you are not familiar with my work or anything I've been posting on the web for the past several years. If you knew, you wouldn't have made the statement you made. It might have occurred to you that we are on the same side. I don't know about you but I don't have a problem challenging my own beliefs. My feeling is that If they're correct, they'll stand. If they're not correct, I'll have learned something, so I win either way.

 

As someone who spends his days listening for small differences in my work, I don't agree that the approach I'm taking is "attacking the problem on the back end".

 

Regarding the software examples you mention, the Mac software will be included in my tests. Incidentally, I use soundBlade, the "engine" upon which Amarra is based, in my work. To my ears, it is the best sounding of the editing/mastering apps in my toolbox, all of which have sounded different to me and has already been verified as being bit perfect. It will be interesting to apply the same methodology to all those apps I use in my work. (I use several apps because each does things the others don't.)

 

At this point, I don't have a problem with the approach and contrary to any impression I might have given (or any that is inferred, despite my words), I'm not "attacking" anything and I don't see a "problem". This is a personal investigation and I'm drawing observations along the way.

 

Twice now, you have said "if you don't hear a difference" yet all I have spoken of is one experience with iTunes, the original master and one other (as yet unidentified) app. I don't know if you have compared iTunes playback with the master files used to create the CDs extracted into iTunes, so I don't know if you would hear a difference in this particular case either. In fact, not hearing a difference is actually a compliment to iTunes and to the other app; they sound indistinguishable (to me) from the original master used to create the CD. That's exactly what I seek in music server software and I can think of no greater compliment.

 

Despite your accusations, I am approaching these tests without preconceptions. (I wonder if you are under the impression I'm one of those engineers who is out to prove there are no differences. If so, that would be mistaken.) Not everything we hear can be measured but it can be recorded (otherwise we wouldn't hear it via our systems). If things sound different, I would expect (perhaps mistakenly but time will tell) there would be a difference in the recordings. To me, this is only logical. If I hear two different sounds, then a good recording of those sounds should preserve those differences. It works when I hear rooms that sound different or a player uses two different sounding instruments, etc. I see no reason why it wouldn't work for any sound I can hear.

 

Further, to my thinking, the best sounding app (whatever that is) will get the most out of what is in the master, without "enhancing" what is in the master. Since soundBlade and the MIO Record Panel are the best sounding apps in my experience for playing back master files, I would expect a music server to match these (or perhaps show what might be "wrong with them" if it can). Any other sonic discrepancy is to my mind, distortion, since the master is the reference. Note I'm drawing a distinction between what one likes (which is fine, whatever that is) and the sound of the master, which can either be reproduced or it can't.

 

I'm out to learn and to have some fun. I'm sorry if you don't approve of the methodology. If I'm wrong, I'll learn something anyway so I don't see any disadvantage.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

[audiozorro]

Perhaps I have doubts on the motivation of this post. I believe that we are usually very open and honest in our criticism and praise of products. When I read something like

 

“I wanted to see how playback via iTunes and from another “server” app (which I will not name until all my testing is complete) compared”

 

I feel the attempt to drag readers along until the end. Although I am not a fan of ABX testing were that applied in this case, you should be the one kept in the dark until the testing is complete, not we as readers Since that is not possible, I believe since you have the knowledge of what is being tested, then we should too, unless you’re trying to have fun with the CA readers by stringing them along.

 

For clarification perhaps you can answer whether you have ever heard any sonic differences in software players (other than your stated volume control error) and whether you believe sonic differences can exist in software player that are bit-perfect. One would assume someone in the recording and mastering business has some knowledge or opinion on these questions.

 

You stated that you are approaching these tests without preconceptions and stated earlier:

 

"First, I do not have any "conclusions", only observations based on what I found."

 

To which Mani quoted one of your earlier posts where you said:

 

"... I would have to conclude that any talk of iTunes "messing up the sound" is unfounded. And that anything that sounds different could well be altering the sound from the original, i.e. "enhancing" it but in doing so, taking it away from, not closer to, the sound of the original."

 

To which I would add that your original post reads like a carefully worded conclusion:

 

"If this is the case (and I’m not yet positive about this – just putting it out there for others to consider and perhaps share their thoughts, in which I’m interested), I would conclude iTunes is sonically the same as listening directly to the master and so is the other app. If this stands, then choice of app is a matter of preference for the user interface (or price) and not for sonics, where there would be in fact, no difference."

 

So at this point I feel that the bias is in and I would be shocked if your final conclusions are different at the end of your testing. I don’t know if you live close to any other CA reader like Mani or Chris, but I would prefer that you go someplace first to hear the sonic differences that many others have reported hearing, before conducting these tests.

 

I assume that if Chris has another CA Symposium or perhaps if someone at the RMAF conducts a software player comparison for a group of believers and non-believers to hear first hand, we will all be enlightened.

 

 

[TimDH]

This is getting pretty silly. Does anyone seriously think that Barry's preconceptions somehow made the playback bit perfect? Who's preconceptions are stronger here? Look, measurements can be incomplete (not enough to tell us everything) but to somehow deny that state-of-the art (if we accept that the MH ULN-8 is, which I think it is ) are actually WRONG is obtuse. A lot of measurements don't tell us enough to judge ultimate sound satisfaction (frequency response of amps/speakers). But this is actually a pretty simple measurement- whether the bits delivered to the DAC are the same through different software players. I have been very suspicious of Amarra and PureMusic precisely because they are so vague as to why their "sound engines" should sound better than iTunes. I think there could really be a benefit to playing files in their native resolution on the fly, which these players do. However, I would be very curious if Barry tried the SRC of iTunes in comparison to iZotrope and found a big difference. iTunes uses the Core Audio SRC, which looks pretty good in the various SRC comparisons posted on line. One of the primary principals of science is that we believe in reproducible measurements. Sometimes we have to learn to make new measurements. But we don't throw data out just because it differs with our preconceptions.

 

[bdiament]

Hi audiozorro,

 

"Perhaps I have doubts on the motivation of this post. I believe that we are usually very open and honest in our criticism and praise of products. When I read something like

 

“I wanted to see how playback via iTunes and from another “server” app (which I will not name until all my testing is complete) compared”

 

I feel the attempt to drag readers along until the end. Although I am not a fan of ABX testing were that applied in this case, you should be the one kept in the dark until the testing is complete, not we as readers Since that is not possible, I believe since you have the knowledge of what is being tested, then we should too, unless you’re trying to have fun with the CA readers by stringing them along.

 

At this point, I must say I find your posts continually missing my point and I don't know why. You draw conclusions (ABX is one - don't know where that one came from) with no basis, attempt to tell me what I believe and what I'm trying to do and in general have the whole idea wrong (I can say this since the idea is mine). And the continual accusations are tiresome too. You "have doubts about the motivation of this post"?

I don't know why but it seems you are in some sort of argument or "fight". If you don't like the thread, I don't really know who might be forcing you to read my posts. If you tell me who that might be, I'll ask them to stop.

 

Perhaps I don't want to do harm to a given manufacturer by saying I don't feel their product is in any way "special". (I much prefer to mention those products I have a positive experience with.)

Perhaps I am not at liberty to discuss the product in question even if I wanted to. Neither of these appears to have occurred to you.

 

 

For clarification perhaps you can answer whether you have ever heard any sonic differences in software players (other than your stated volume control error) and whether you believe sonic differences can exist in software player that are bit-perfect. One would assume someone in the recording and mastering business has some knowledge or opinion on these questions.

 

If you read my earlier posts, you will have found the answer to that question.

 

 

You stated that you are approaching these tests without preconceptions and stated earlier:

"First, I do not have any "conclusions", only observations based on what I found."

To which Mani quoted one of your earlier posts where you said:

"... I would have to conclude that any talk of iTunes "messing up the sound" is unfounded. And that anything that sounds different could well be altering the sound from the original, i.e. "enhancing" it but in doing so, taking it away from, not closer to, the sound of the original."

 

To which I would add that your original post reads like a carefully worded conclusion:

"If this is the case (and I’m not yet positive about this – just putting it out there for others to consider and perhaps share their thoughts, in which I’m interested), I would conclude iTunes is sonically the same as listening directly to the master and so is the other app. If this stands, then choice of app is a matter of preference for the user interface (or price) and not for sonics, where there would be in fact, no difference."

 

"Carefully worded conclusion"!?! I said if the nulls prove to be the case, I would reach a certain conclusion. I'm sorry if it is different from a conclusion that you would reach or if it is too "carefully worded" for you. Geez Louise are you serious?!

And I'm sorry Mani decided to take only part of my sentence instead of the whole thing, in full context. (Perhaps that would have made the argument less pointed?)

 

 

So at this point I feel that the bias is in and I would be shocked if your final conclusions are different at the end of your testing. I don’t know if you live close to any other CA reader like Mani or Chris, but I would prefer that you go someplace first to hear the sonic differences that many others have reported hearing, before conducting these tests...

 

And your point is?

Really audiozorro, you want to pin me to something you want to believe? Go ahead if it lets you sleep any better.

 

You have made it clear, repeatedly I might add that you feel "that the bias is in". Okay, you feel I'm biased. Noted for the record. Now that it has been noted, I would ask you to allow me to voice my observations as I experience them. If you find this distasteful in any way, I would suggest you don't waste any more of your time on me and what I think.

 

Can we get back to the thread now? Apparently, you have not read my previous posts carefully or simply ignored what did not fit your preconceptions. Clearly you are not familiar with the positions I've taken on various audio issues over the years. And you insist on seeing me as some sort of audio "enemy". For me, audio is a playground. I've been around too long to get into the nonsense of folks who insist on making it their battleground. If you want to have an open discussion, I'm there. As to your fight, from here on out, you're on your own. I wish you happiness.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

 

 

[cfmsp]

"...but I would prefer that you go someplace first to hear the sonic differences that many others have reported hearing, before conducting these tests."

 

so long we're just sharing preferences, I would prefer that you AZ quit your griping about Barry's approach, lest he do his tests in silence whereby we all lose the benefit of what he does. Frankly, I'm not sure why he's been so polite to your criticisms thus far. If it were me, I'd go on my merry way and spare myself unwarranted criticism.

 

You've been told multiple times that your assumptions are wrong, and yet you continue to object. What right do you think you have to suggest that his tests need to meet your conditions?

 

You're the one who seems to have preconceived assumptions here - attempting to damn Barry's approach before even understanding what he is doing and why.

 

 

 

clay

 

 

 

 

 

[bdiament]

Hi TimDH,

 

Thanks for your post.

If my "preconceptions" could alter results, I'd be pretty rich right now. ;-}

And Night Train would taste like Dom Perignon. ;-}

And if they did so in the test this past weekend, the other app would have very easily outperformed iTunes, for that is what I was fully expecting.

 

With regard to iZotope's SRC, I have compared it and many others in a long series of SRC tests conducted a few years back. To date, I've heard nothing that is close to the transparency of iZotope's conversion algorithm. (My criterion for this test is comparison against the unconverted original. Most SRC algorithms I've heard tend to brighten and harden the sound.)

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[salwani]

First, thanks Barry for the effort. This is one interesting topic.

 

I have 2 points to raise

a) Is the recorded output truely from the respective players ? A control here could be in the form of a comparism of an altered file (from within the player via EQ/gain, whatever) compared to an unaltered file.

b) Since most report subtle / slight changes in using various players, could it be the (in)sensitivity of the playback equipment in playing just the difference between the 2 players ?

 

But if the numbers are not lying, it would be interesting to see what the conclusion is. They do sound inexplicably different / placebo....whatever : i need to know so i can get the imagery of me in a nut house repeatedly mumbling 'but i can hear differences', out of my head.

 

Sunil

 

[Panelhead]

The results you reported do not match my experiences with Amarra and PM/PV. I hoped to see measured differences between iTunes and the other software.

This does not mean that I doubt your; motivations, methodology, equipment, or honesty. Hope you can spend more time testing and less explaining why and how.

Got to think the differences have to be in the digital side. If you feed the dac with identical data the analog output has to be identical. Or very close to it. If a dac fed bit perfect data ten times outputs ten different analog signals, the dac is not very good. Maybe one of the early dacs that were 16-18 bit but delivered 12-14 bit resolution. The errors generated in the AD process should be repeatable.

I can understand digitizing an analog signal ten times may yield ten different digital results. The errors get smaller with higher sampling and bitrate. But never go to zero. Just close to zero.

 

George

 

 

[Paul R]

You really should look at Barry's website and work. Better yet, listen to it - as it surely speaks for itself.

 

For what it is worth, I do not have the chops to speak authoritatively about music based audio engineering, but Barry certainly does. I do find that his investigative technique is sound and I have pretty good confidence in the results he reports.

 

Honestly, where are you coming from with that really arrogant attitude? Or is it just that your written communications are not clear and produce unintended assumptions when read?

 

-Paul

 

 

[audiozorro]

Perhaps I am mistaken, but when I read a title that says “food for audio thought” I can only presume the author intended to provoke thought. I believe that I am well within the bounds of providing my thoughts politely, as we are all free to do here.

 

I believe Barry is attempting to see how playback via iTunes compares to other software players. I believe I asked Barry the following two fair questions that I believe anyone reading this post could respond:

 

1. whether you have ever heard any sonic differences in bit-perfect software players?

2. whether you believe that sonic differences can exist in software players that are bit-perfect?

 

When I read that Barry would not name the products, other than iTunes, he was testing, until at the end of the test, I assumed that this was his personal decision and not that he was being restrained by some manufacturer. If my assumption is incorrect, this is my mistake.

 

I believe that this post is an open discussion and I have no battles to fight, but I have expressed my beliefs and observations that I thought relevant to the topic. I have no hidden agenda, no services to sell and no products to champion.

 

But my observations and the observations from others that I value led me to question the test. Of course Barry is free to conduct any test, in any way he wishes, but I believe I see a flaw in the test that many here may agree or disagree with.

 

Many of us have been able to hear the different sonics through:

 

1. different software players like iTunes, Amarra, Pure Music or XXHighEnd

2. different digital cables with the same bit-perfect data going from Point A to Point B

3. different types of digital transmission or different digital interfaces

4. playback from solid state drives versus spinning hard disk drives

 

I could continue the list but my point is that if bit-perfect digital data can sound different from changes in the digital realm that do not affect the bit-perfect data, then any conversions from digital to analog to digital are just as likely to sound different.

 

These are not preconceptions, these are observations and beliefs. These sonic differences are real enough to those of us who hear these or similar differences and judging from some of the posts, perhaps not real enough for those who don’t. It is perhaps my mistake in thinking that, at least here at CA, many if not most believed in the possible sonic differences of bit-perfect data in the digital realm.

 

I thought most of us believe these differences are due to jitter or timing differences. Some believe that there are additional or unknown factors at play. If we just focus on jitter, then why wouldn’t a toslink cable result in different sonics from a coaxial or USB cable, why wouldn’t two coaxial cables transmitting identical bit-perfect data sound different, and why wouldn’t a software player like XXHighEnd that I believe has at least 4 adjustable parameters for modifying jitter sound different with different bit-perfect settings?

 

My apologies to anyone who took offense or perceived arrogance to any of my comments or questions. That was not my intent.

 

 

[manisandher]

"And I'm sorry Mani decided to take only part of my sentence instead of the whole thing, in full context. (Perhaps that would have made the argument less pointed?)"

 

Barry, I wasn't trying to make my argument more pointed, I was trying to be as brief as I needed to be with the quote.

 

So, for the record, here's what you actaully said:

 

"I also have to double check the nulls with either the volume wide open or by capturing the nulls and increasing the gain. In my studio, at a rather healthy playback volume, the power amps might as well have been turned off. When I nulled the files, the result was silence.

 

If the silence remains after the double check on the nulls, I would have to conclude that any talk of iTunes "messing up the sound" is unfounded. And that anything that sounds different could well be altering the sound from the original, i.e. "enhancing" it but in doing so, taking it away from, not closer to, the sound of the original."

 

My interpretation of this was: subject to double checking the nulls [verifying part of the methodology], I conclude that...

 

Now, without meaning to pre-emp anything, I strongly suspect that you will indeed find that the silence remains after the double check on the nulls. This is to be expected, after all, your methodolgy tests whether these players are bit-perfect, and they certainly should be.

 

Whatever you choose to conclude after that is up to you...

 

Mani.

 

[Mahesh]

audiozorro,

 

"When I read that Barry would not name the products, other than iTunes, he was testing, until at the end of the test, I assumed that this was his personal decision and not that he was being restrained by some manufacturer. If my assumption is incorrect, this is my mistake."

 

Look, you have a software company taking $$ or $$$ for an application that is supposed to sound a lot better than iTunes. Then, a sound engineer with 35+ years of experience does some testing, and, initially, all his findings suggest that the super app does nothing more for sound quality than Apple's free iTunes. He wants to do some more testing, and double check everything, just to be absolutely sure that his findings are correct, especially since he actually had the preconception that the super app would, indeed, sound better than iTunes. He knows that a public statement from him about the super app not delivering what it is marketed as doing, might have negative implications for the software company, considering his reputation in the music industry.

 

At the same time, he thinks that his experiments might be interesting for his fellow music hobbyists, and he decides to write a preliminary report. He decides to not reveal the name of the super app until he is absolutely sure that everything is double checked. In addition, he knows about this other company out there, that makes another super app, charging money for it. So to make things fair, maybe he should test both super apps before going public with the names.

 

I consider this a plausible explanation for not revealing the name of the app.

 

Mahesh

 

[Mahesh]

Mani,

 

">> [barry:] And I'm sorry Mani decided to take only part of my sentence instead of the whole thing, in full context. (Perhaps that would have made the argument less pointed?)"

 

"> [Mani:] Barry, I wasn't trying to make my argument more pointed, I was trying to be as brief as I needed to be with the quote."

 

By omitting that part of the sentence, you completely changed the meaning of it. You may omit irrelevant parts when citing someone, but you should ensure that the meaning of the original statement is preserved.

 

Mahesh

 

[manisandher]

Well that certainly wasn't my intention... and I hope I've put that right now.

 

Mani.

 

[manisandher]

I think what many here are missing is that Barry's test validates the bit-perfectness (before the DAC) of the software.

 

I think drawing any conclusions wrt the sound quality of the software (after the DAC) are flawed.

 

'Bit-perfect' does NOT equal 'sound quality'.

 

Mani.

 

[PeterSt]

Warning : Final attempt post, trying to correct things, diplomacy not in my vocabulary, double dutch to be expected, possible Explicit Lyrics, ban alert !

 

But I'll try my best.

 

Generally I can't take it well when things are so much interpreted wrongly, especially with all the good intentions from everyone (Barry ahead in this case !!), while as it turns out now, ehh ... almost nobody seems to understand the merits of bit perfect ?

 

And no, I certainly don't want to feel *anyone* addressed explicitly, but if I may and FWIW, it is my conclusion from this thread that only few people seem to have gotten what this is all about. I mean, from the -say- past year and hundreds of posts.

 

Let's start with the last posts, in order to an attempt of making clear what's happening with the communication here. Btw Mani, wasn't this your job ? haha.

 

Now, I read through the last 40 posts or so since yesterday, and this starts with one from Audiozorro. I'm addressed here and there, and felt the intention to reply to it by something like "if you're talking like that, don't expect any answers from me".

 

With this in my head (shotgun still on the shoulder), I continue reading, and AZ starts to explain why Barry shouldn't be doing what he is doing, or should do otherwise;

 

Please try to understand this :

 

After reading the first post of AZ I completely agree with AZ on his next posts. why ? because AZ *does* understand the merits of all this (bit perfect and blahblah). He thinks in the line of how things are BUT forgets that Barry approaches things not along this line. I tried to explain this per already my first post, and I'm sure I failed, even after the third.

Of course, Barry doesn't understand the "attack" (which is no attack at all !!) from AZ, because he doesn't think along the lines ... which should be logical lines after these hundreds of posts about it in here, and which to me never prooved that many didn't get it. And of course, all those rejecting AZ's posts seem to be on this same path (and please don't tell me that I'm suddenly against someone like Clay or anyone else ...).

 

Is this a problem ? is this bad ? hell no, why would it. It may be a pitty for good understandings, but otherwise ...

 

Well, otherwise it creates these communication problems, which is bad for good atmosphere the least.

I hope this post does *not* contribute to that ...

 

Well, what actually is happening, as how I see it ?

 

Barry, by the way you wrote your first post, and the further responses, there are many things which are completely unclear to those who, say, know already. Whether we should call this known facts or science or whatever, they are commonly known to some. What you do, however, is dispute those facts (very positively, I can't emphasize this enough), and start work, done 30 years ago, all over. Notice that this is the merit of your writing, and not the merit of your intention. You, without realizing it, try to proove that computer data is not changing along the way of transfer.

 

I cannot help it that you don't see that it is actually that what you are doing (please trust me, you are !).

I also can't help it that 90% of readers find it interesting all over. They learn, like you do. But :

 

What's wrong with it (no, what I feel wriong with it) is that this same 90% has been part of discussions for a very long time and in many many posts, but apparently never understood everything about it. Of course, by itself this is not wrong at all, but it *is* very disappointing to those who try hard at explaining, and out of all those are the outsiders at the moment.

 

It seems that CA by now has obtained quite some community with knowledgeable people, but ... they learned too much by heart and never really understood perhaps ?

It feels so strange. But also hard to deal with.

 

If I read a post from I_S, like his last one in here, I read that with very much interest, try to learn everything what's to be learned from it in my case, but I *ALSO* keep track of what you get from such a post. And you know, since there are no confirmations, no questions either, BUT many of these posts came by in the past, I will ASSUME you got the messages from it. As I assume you'll get them from my posts.

ANYONE's post !

 

The latter is dangerous, if it actually is difficult for many to understand what someone like me tries to tell (yea, me especially of course with my stupid dutch).

 

Should I care ?

Well, yeah, I think I should. Look, this thread started beautifully, and if more people would have the material, time and sense to do these kind of things, audio would proceed faster, I guess. Or anyway it would make clear better what is going on.

But,

If someone like me tries to correct what is "wrong" in the setup, it shouldn't be so that 90% disagrees with that. Mind you, this isn't in the posts, but what is in nearly all the posts is that Barry will continue, and 90% is very excited to learn the results.

 

Results of which I can promise you they are known in advance, because, well, it is known already. I mean, is there someone on this globe still trying to proove that 1+1=2 ? Sorry for the stupid example, but it is in the same line of what is happening here. One difference, apparently is is not accepted yet that 1+1=2 so we're still trying to proove it. Ok ...

 

Sadly this distracts from the real matter, and this is what happens *after* what is known to be bit perfect. So, we KNOW (do we have this ?) that all players saying that they are, indeed produce the same bits and bytes and everyting, but after that has been produced, still the sound can be different.

"We" can do two things : stick with prooving which has been proven long time ago and don't proceed, or go on to the obvious step : find out why sound changes after step 1.

 

Maybe it is good to point out that by a thread like this, many people will support the "I don't hear a difference either" which, sadly, is NOT important. All those not perceiving a difference between players are just lucky because they can spend their time with listening to music. What IS important, is when difference can be pereceived clearly while at the same time (!!) the digital data is the same.

Allright, I am not going to tell you (Barry) what you should or should not do, but it is the only sensible thing to do when you want to explore in this area in the first place.

... As it is so that everybody's interest should be there, instead of prooving 1+1=2.

And again, I am NOT telling everynody what to do, but it would be the only logic thing for someone like me to "perceive from you" while nowhere (I mean NOWHERE) so far has been any proof of doubts from people what it actually means that a player is bit perfect. Whether by means of HDCD light, by means of ever extensive post and work from Juergen, or my own. So I sure hope this is clear : by the activity of this thread you will be doing nothing more than prooving that Juergen (etc.) is correct.

Wow.

 

Back to the beginning of my post, it is the way all is written and explained that makes some people totally confused. So Barry, what do you actually mean by comparing the master with iTunes ?

But merely, why does nobody else ask this question ?

Don't we all see that comparing a master audibly is just comparing a player again (like SoundBlade) with iTunes ? or do we perhaps now doubt that a master may have different digital data than the rips we make ? (yes, I know, some even doubt that, but I don't recognize this is the subject (of testing) here).

 

And Barry, I think it is 100% allright to check for yourself whether a player is bit perfect, or check whether the way you thought to do it, is the right way (it 100% is). But here it should stop. Here, IMHO, you could say "okay, I can do it, now let's stop with confusing".

Did it occur to you how confusing it might me to kind of state you hear no differences between the players under test (to me clearly suggesting players in general will sound the same), while later the context changes into something like "but SoundBlade sounds the best" (I don't like literal quoting, but this is in my mind).

Or maybe it is so that indeed something like SoundBlade is not considered a player by you ? ... that would explain A LOT (and your next step would be applyig your test to SoundBlade vs iTunes etc. :-))

 

Let me end with once again referring to the first lines of this post, and by saying that I'm only trying to get noses in the same direction, while many probably are already.

I sure hope no one is offended, and by any means this was not intended.

 

All the best to everyone,

Peter

 

 

PS: This doesn't incorporate the last 5 or so posts.

 

[manisandher]

I'd like to hightlight something that I_S put forward as something to think about (I'll try not to miss anything out ;-))

 

"Now, modern CPUs gobble up enormous amounts of current due to the huge number of transistors running at very high speeds. It is not an enormous stretch to believe that playback software ( which much bear some relation in terms of load to the audio out ) could draw current at different times that cause e.g. the supply to oscillators to have a very small amount of noise on them.

 

So how come Barry can't hear this? Because the only point at which small amounts of jitter are critical is at the DAC clock, which would appear to to be separated from any PC clocks by operating in an asynchronous type mode ( firewire, but could equally be true for USB/ethernet )"

 

I think this is something worth exploring further... although I regularly use an AFI1 and FF800 (both of which are async firewire devices) and still hear differences between bit-perfect software players. Also, I'm pretty certain Peter used a FF800 for his tests, although I'm not sure exactly how he was using it (firewire or spdif).

 

IMO, here's some real food for thought...

 

Mani.

 

 

 

 

 

 

[PeterSt]

Hi Barry,

 

Try to keep in mind, I am trying to constructively helping you, and in this post not dismissing what you're doing.

Maybe it gets you thinking, maybe not. All is ok. This time I do quote lterally, or otherwise it won't "work". The below is only one concatenated sequence out of a reply to AZ. It is full of potential misunderstandings ...

Ok, to my poor brain it is.

 

I am simply seeking to find if the apps I am testing sound different.

 

It will be symantics, but I would say that you then would have had to have two examples sounding different. But you have ... SoundBlade vs the rest.

 

If they sound different, it would seem to be reasonable to expect that captures of the output of these apps will also reflect differences.

 

Don't laugh, don't cry, but this is NOT reasonable per the *fact* that software has been sounding different for, say, 5 years, and the whole world talks about it for the last year, especially in the ream of "bit perfect". Search for the phenomenon in Google or in CA and try to dismiss that ...

I have been talking to my self somewhat longer. :-)

 

(Some here seem to confuse the capture prior to the DAC with the file that is input to the program.

 

It would be a kind of fun(ny) when you could quote that. Because you know, if it were for me you are the only one saying this. Also, I think this is a response to my suggestion that no such thing exists as capturing earlier than the output of the program (plus soundcard etc.), unless *you* start to talk about files.

Or ... unless you indeed were referring to the master (yes, you did that) and now .. oh well :-)

Or ... you thought it wasn't important to compare the file ...

 

I know, this time *I* will be the confusing one. I mean, did you ever realize that the captured data you compare will be 100% the same as comparing the captured data with the file ?

Now, assuming you see that I am right on this, do you see how much confusing it is to state that out of all "we should not think you compared with the source file" ?

We should, because it shouldn't make a difference !!! (but do you know ?)

 

There also appears to be some mistunderstanding of the difference between a "loop back" and the way the MIO Record Panel works in association with the MIO hardware.)

 

Well, with the big smile On, what would you suggest it is what you are doing then ?

Here too, to me it only shows that (softly said !) you don't know what it is you are doing on this subject (please don't take this as an offence, I don't know how to say it better). So :

If we are talking about a loop back, this -in its *obvious* context- is about digital loopback. This can be done by cable or by software. It is your choice how to do it, but tap it (Y) from the input of the DAC would be the most honest.

In all cases you ARE looping back, unless you captured by microphone (kidding here).

If you think otherwise, let me know how or what. It would be the most interesting.

Might you have thought that someone like me meant analogue loopback, then I wonder how you think someone like me had let that working. This can not be a mistake, in the context of me explicitly telling about the *analogue loopback*, or didn't I ?

 

All 'n all, you create huge confusion here. Everybody who was sure what a loopback is, now doesn't know anymore, because it is done by some secrecy in a MIO thing.

 

In other words, if I hear a difference, I would expect the input to the DAC is different, since all apps are feeding the same DAC, using the same clock.

 

Again, please don't perceive my words wrongly, but why bring up a clock ? because Eloise brought it up ?

The clock in the DAC (or a clock in the PC if you want) is irrelevant. It can have jitter, you could even let it run faster between two takes, the data you are capturing will be the same always. It may depend a bit on what you capture from, but SPDIF assumed (or ADAT etc.) won't resample or anything or in other words, I don't think you will be able to find a way that the capturing clock will capturing wrong slopes or anyhthing subject to the change of data. I must add though, that letting the software capture the data internally (by which I refer to your MIO thing ... yes, I know that is hardware) is prone to just copying the data, which is why I said earlier that it is more honest to use a cable for the digital loopback. And even then ...

 

I have deliberately not introduced any variables that are not present when I'm listening and in fact I was listening while the captures were taking place.

To be clear, each capture was a recording of the output of each app, played in real time, as I listened to the music. There was no loop back and no additional conversion to AES to get in and out. It was the direct Firewire feed that was captured; via the same hardware, just ahead of the DAC section.

 

You may not realize it, but this suggests a few things exactly opposite. Look :

 

1. This quote would apply a 100% if you had measured differences.

2. Thus, while you are prooving the other way arond, this suggests importance all over, while (I say) it is not at all.

 

Besides there *was* loopback (digital) :

a. putting up the volume really doesn't matter a thing (hence, what is the intended message really ?);

b. a conversion to AES in/out/back/up can (I think) differ in jitter, while your capturing mechsnism doesn't allow to record jitter as such. Thus, it can't make a difference for your result.

c. mentioning a "direct firewire feed" is ? I don't know. By putting that like you did you can only mean data. However, this isn't about normal data, it is about audio data (which also is data in the realm of not subject to jitter), but you used a protocol. It can have been SPDIF, ADAT, MME emulated or not, and some more Pro protocols.

 

Am I nitpicking ? actually yes. But I hope it makes clear that it is confusion all over, although maybe I'm the only one suffering from it. Maybe one more. ;-)

 

I again hope you take this as intended ... : Well.

Peter

 

 

 

[bdiament]

Hi Sunil,

 

"...a) Is the recorded output truely from the respective players ? A control here could be in the form of a comparism of an altered file (from within the player via EQ/gain, whatever) compared to an unaltered file.

 

To my knowledge, and from what the designer himself told me, yes. If I change iTunes' volume during the capture, this is heard in the captured file. If I insert EQ in the other app, this is heard in the captured file.

 

 

b) Since most report subtle / slight changes in using various players, could it be the (in)sensitivity of the playback equipment in playing just the difference between the 2 players ?

 

Not sure how to answer this. It could be that I simply don't hear it. Personally, I would not be heavily on this but that is just my opinion. It is also my opinion that the playback system is quite good at revealing very small differences (this is why I use it for my daily work) and the DAC and software (not the server software) are, in my experience, unmatched.

 

At this point, I have many questions of my own. I hope to get some answers along the way.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[PeterSt]

Hopefully this isn't too much off topic (but at least I consider this the 180 degree different subject from Barry's) ...

 

I think this is something worth exploring further... although I regularly use an AFI1 and FF800 (both of which are async firewire devices) and still hear differences between bit-perfect software players. Also, I'm pretty certain Peter used a FF800 for his tests, although I'm not sure exactly how he was using it (firewire or spdif).

 

IMO, here's some real food for thought...

 

Mani.

 

Yes, I used a Fireface800 for capturing, but remember I talk about the analogue domain. Thus, SPDIF to the DAC, An-out to the FF. In this case the SPDIF to the DAC came from the FF because it is also the capturing device in such a setup, and the clocking device must be the same for that (or otherwise the capturing will de-rail hence captures different slopes (the point on them) a next time.

For others : you may attempt to do the same of course, but please notice it took me a couple of months to write the software in order to let startpoint calibration be workable (which is 100% essential here, different to capturing digital data).

 

Allright, the more food for thought comprises of this :

 

What I have proven, and which is partly by means of applying logic, and partly because of physical work on the (Phasure NOS1) DAC, is that no matter what we try, jitter increases by means of exactly what I_S said.

What some know, is that at a certain stage I start to build that DAC to create the opposite of what XXHighEnd works with : make it immune for that.

I always failed ...

 

Currently the DAC uses a 4ps clock, and this is net output at the same time (meaning : if you'd measure this DAC, it measures 4ps of total jitter at the analogue outputs). BUT THIS IS MEASURING.

Measuring is a constant flow of the same sequency (two if you want) with always the same load on the current. It just can't be used to measure jitter, because of this and the assumption I_S made (which is actually mine). So my THEORY is this :

 

No matter what we do, the jitter in the DAC, originally 4ps whatever, grows into 800ns or whatever it takes to create the HUGE audible differences by means of software only (those who don't perceive any difference at all, please skip, haha).

So, what I am currently hunting for is attacking the beast before it can harm. I want something like a 40fs jitter clock, that turns into 4ps at playing music, and THAT won't bring audible differences anymore.

 

But where to obtain such a clock ...

I, so far, can't get further than 300fs or so, if it were to be a clock which comes with all the specs needed (and don't think a 100ppb oscillator should have good jitter specs too).

 

Anyway, and to be clear : the 4ps clock is as influenceable as all what I tried before.

 

Peter

 

PS: Any tips are welcome !! but with jitter specs please, not stability only.

 

[manisandher]

"Yes, I used a Fireface800 for capturing, but remember I talk about the analogue domain. Thus, SPDIF to the DAC, An-out to the FF."

 

So, you're connecting to the PC via async firewire.

 

And when I_S says...:

 

"... the only point at which small amounts of jitter are critical is at the DAC clock, which would appear to to be separated from any PC clocks by operating in an asynchronous type mode ( firewire, but could equally be true for USB/ethernet )"

 

... this is taken into account - you're not using the PC clock, but the FF800 clock.

 

And yes, I totally inderstand why the clocking device must be the same for the spdif feed to the DAC and the analogue capture from it.

 

Mani.