Why third party players?

[Miska]

But could the c48 be so inferior that is could mask sonic differences between iTunes and third party players?

 

Or you are lucky to have good enough system that the same data sounds the same - just like it should. Don't try to "fix" it if that's the case.

 

Different data, like upsampled or otherwise processed should naturally sound different. Same data shouldn't (just like two pieces of the very same CD).

 

 

[thrang]

Or at the price of the McIntosh you could buy the Bridge or V-Link and look around on dealer's websites, Audiogon, etc., for top-end used pre-amps, e.g., Spectral.

 

Yes, at the point the C48 is going back -

 

How does the v-Link/M1Dac combo differ from something like the HRT Streamer II +, which is spec'd as asynchronous USB transfer?

 

Thanks

 

 

[thrang]

Or you are lucky to have good enough system that the same data sounds the same

 

Sacrilege...!

 

[4est]

IME, if everything sounds the same it is from something masking the differences. Put another way, there appears to always be differences, even in digital!

 

[Jud]

thrang asks:

 

How does the v-Link/M1Dac combo differ from something like the HRT Streamer II +, which is spec'd as asynchronous USB transfer?

 

I've never heard the M1DAC or the HRT, so I wouldn't know. I run an inexpensive (~$50) Audioquest Emerald USB cable from a MacBook Pro to the V-Link, then through an Omega Mikro Zephyr S/PDIF cable to my (very) old reliable Theta Pro Basic II DAC. So I am just playing ripped 16/44.1 CDs, but it still sounds better than my Theta Data playing the original CDs did. (In my computer audio system before the V-Link, I ran Toslink from an Airport Express through an optical-to-coax converter before it went through the Zephyr and into the DAC. The original CDs were superior to that setup.)

 

Something else I don't know, since I've never been in position to make use of it, is what effect upsampling may have. Upsampling vs maintaining a "bit perfect" signal through the digital path is a matter of some controversy here (imagine that!;-). I believe IIRC that the M1DAC upsamples; don't know about the HRT. And it looks like the C48 upsamples. If the HRT doesn't upsample, and you liked the sound as much as the more expensive C48, you might want to look at other DACs along that line rather than the M1DAC.

 

BTW, if you don't hear differences between players, I would sincerely doubt it's because they're all doing a perfect job. OTOH, if you try a range of equipment and still can't distinguish among players, great, you can use whatever's most convenient. :-)

 

[Miska]

if everything sounds the same it is from something masking the differences

 

Yeah, I prefer that equipment and listening room rejects (masks) interferences! Including sun spots, full moon, rain/shine and neighbors' peeing...

 

I still want to listen to the data stored in files, not to the CPU/GPU/HDD/whatever activity.

 

Some good tests to listen and measure:

1) Play a WAV/FLAC of digital silence, check the noise floor

2) Play a sine sweep and check for any spurious things

 

Repeat above with full computer activity see if everything is still perfectly clean. If not, there's something to fix in the hardware...

 

 

Edit: For (1) you may need to use dithered silence, in case your DAC has silence detection and engages mute circuit on series of digital zeros.

 

 

[4est]

That is if the OP's system if of sufficient quality in regards to equipment AND set up. Here is where we see a basic difference in our methods from what I have read. You are instantly going under the assumption that the digital domain is where the problems are whereas I am convinced that most problems occur in the analog domain, and need to be addressed there.

I realize we are discussing the differences in the sound of digital software, but I feel you are conflating the OP's issue with not being able to hear a difference as there being none between player programs. I realize you are a programmer, and the thought of player programs having a sonic signature w/o a hardware problem bothers you. Yet, many here and elsewhere have noted them. If it is a hardware issue with all of us, please, please inform us as to what is going on instead of telling us we are stupid enough to think our neighbors urination is effecting our sound systems.

 

[Miska]

You are instantly going under the assumption that the digital domain is where the problems are whereas I am convinced that most problems occur in the analog domain

 

No, I'm specifically saying that these specific problems are NOT in digital domain - thus in playback software. If both read "Hello world!" from the WAV file and send "Hello world!" to the DAC, the message should be unchanged. And yes, of course you could write two books of text (UTF-16) in parallel into 16-bit stereo WAV or FLAC and having 44100 characters per second transfer speed. You could even transfer it between two computers using S/PDIF. This is similar to how DACs with "bit perfect" -check perform the test, just looking if the predefined string of values is received. Regardless which "bit perfect" software is used, you should be able to receive the books intact on the other side.

 

Since it never entered analog domain through DAC, S/PDIFs jitter wouldn't have impact unless it's so high that the receiver side loses lock.

 

If it is a hardware issue with all of us, please, please inform us as to what is going on instead of

 

I've been telling it several times already in the past. Different software components cause different activity patterns in noise within and by the computer (or any other device that has different activity patterns). Interference sensitive devices that are within/connected (or even nearby) to the computer can be affected by this. This is all analog domain activity and doesn't affect the digital domain or the data unless the hardware is broken.

 

Properly constructed computer playback hardware is immune to any interference and only deal with the data sent to it, buffered and clocked out independently. DAC is dual-domain component with digital and analog domains. DAC's clocking is analog domain activity. Analog interferences existing in digital domain signals (note that these don't affect correctness of the data unless hardware is broken) must not enter the analog domain - including the clock.

 

Now, continuing in spirit of the previous message, if we think the digital domain being the neighbor, none of the neighbor's activity is supposed to interfere with our listening in analog domain, regardless if they lended us a CD or are even streaming the music to us over WLAN. If sound proofing (interference rejection) is not good enough, their activity could interfere. But it wouldn't make the listening room more "resolving", actually less.

 

If you don't take the data out of the digital domain and copy it around you can algorithmically prove if it changes or not. The problems may begin when it enters the analog world. "Bit perfect" software just copies the data around and you can verify it with checksums or plain comparison, just like any other data in computer.

 

In short - digital and analog domains are separate and shouldn't be mixed, otherwise problems will arise. And still the data (but nothing else) should be able to cross the boundary in carefully controlled way. A bit like DHS' arrival checks... Interference can be fixed by better shielding the noisy computer parts, improving interference rejection on the analog circuitry side and properly isolating these two things from each other. Everything else in this scope is just treating the symptoms, not the disease.

 

 

[4est]

Thank you for your explanation. So then, essentially you think it comes down to noise. Where does one find computer hardware that is properly built for these purposes? As with most things audio and otherwise, I suppose there is give and take, but where does a layman start with this?

 

[PeterSt]

I feel it is "ridiculous" to state that the OP suffers from a noiseless system (PC) and thus no differences *can* be perceived. Ehm, might someone try to state that.

 

It is the most (MOST) obvious that indeed something is seriously and heavily masking, so no differences ever can be perceived.

 

Take out your pre-amp. Never mind it works out for the better (for SQ in general) or not. Now try again ...

 

Peter

 

[PeterSt]

I said that because only Miska has such a system.

Not saying that that ain't true, but I do say that you having such a system has a chance of exactly zero.

Proove me otherwise, and show me a person or system who/which achieved that. Miska is counted out. Haha.

 

[Miska]

Where does one find computer hardware that is properly built for these purposes?

 

There are two sides to this, computer that doesn't generate lot of noise towards the audio device, and DAC or sound card that is immune against what is left. If there is galvanic isolation between the two, even better. Location and isolation of the audio clock is important. In well-built computer and card it can be within the computer too.

 

I don't know many good ready made computers from the big traditional computer manufacturers, it's just outside of their range of interest. Maybe Mac Pro for example. DIYer can definitely build one. Then there are those various "music servers" and network players. And there are some smaller local companies that will build these on request, customized.

 

Third important thing is grounding layout of the system, especially in cases with mixed two/three-pin mains connections. Ground currents traveling in cable shields between the devices can carry noise...

 

And no, I'm not against dealing with the symptom using software tweaks, as long as it is recognized as such. What I'm specifically against is claim that a system that doesn't exhibit this symptom would be unresolving...

 

 

[Miska]

I said that because only Miska has such a system.

Not saying that that ain't true, but I do say that you having such a system has a chance of exactly zero.

 

Hmmh, I remember many others saying the same. Maybe it's just that those who don't have a problem don't make so much noise?

 

What I cannot hear with my AKG and Sennheiser headphones connected to the audio interfaces, nor measure, doesn't exist for me. I make no claims about others. Yeah, and sometimes I listen through the "main system" loudspeakers too, but less frequently. Just for practical reasons, 00:00 when others are sleeping is not the perfect time to start playing loud music...

 

But it all didn't happen just out of pure luck, every system component was selected with care. Including computer case, MB, PSU, which card in which slot, etc.

 

P.S. And if my system is then "unresolving" then I definitely prefer it just that way. I'm happy as long as I can hear all my tiny processing algorithm tunings and differences in electronics while having otherwise repeatable results and no ambiguity.

 

 

[Jud]

@Miska -

 

What I got from a previous post was that (1) players themselves should not sound different due to differences in how they treat an identical digital file; (2) it is the effects of players on susceptible analog circuitry of the playback hardware that causes the differences in sound among players; and (3) thus players won't sound different on properly shielded, grounded, etc., hardware:

 

(1) No, I'm specifically saying that these specific problems are NOT in digital domain - thus in playback software.

 

(2) Different software components cause different activity patterns in noise within and by the computer

 

(3) Properly constructed computer playback hardware is immune to any interference and only deal with the data sent to it

 

But now you say:

 

And if my system is then "unresolving" then I definitely prefer it just that way. I'm happy as long as I can hear all my tiny processing algorithm tunings and differences in electronics while having otherwise repeatable results and no ambiguity.

 

So you purport to have the type of properly constructed "unresolving" system described in (3) above, yet on this system you hear "tiny" differences in a single software player, surely far smaller than the gross differences between players from different programmers - the exact sorts of differences you should *not* be able to hear on such a system, according to (1) and (2).

 

Is there some logic I'm missing here?

 

[PeterSt]

I can't speak for Miska all the way, but I am fairly sure that with :

 

I'm happy as long as I can hear all my tiny processing algorithm tunings

 

Miska means the changes he applies which we tend to call "DSP".

 

So, this may come across a little confusing, because following from the logic of it all, we can say this :

 

1.

If your computer is not under the best control for power management, even 100% bit perfect output from one and the same player may incur for different sound. Of course this player has to be made for this, hence anticipates on just these "power failures".

This is what PeterSt says. Well, seems to say. See 3 below.

 

2.

If your computer -but the complete network of interconnecting devices- is under the best control for power management, only then you can perceive very well what DSP like appliances in the player incur for, because everything is now audible the best, and not influenced by sh*t out of your control underway.

This is what Miska says.

 

[intermezzo]

Miska will correct this where needed, and may add a couple of nice "theories" which are no theories at all (meaning it's practice), or may turn out to be "theories only" lateron (next year, whatever).

In the mean time, I will -in advance- add that I don't believe any such "environment under good control" exists, PLUS I don't think it is even relevant. So :

 

3.

PeterSt never has said (though may have implied it by not being very concrete always) that a noisy PC is the cause of playback software influencing. He does say though, that the behaviour of the D/A converter can be influenced by means under his control (XXHighEnd) and which is exactly where his concreteness stops.

He also adds (but said that often) that while knowing what is happening, he himself is not even able to create a D/A converter that prohibits this influence.

In the end this is a good thing, or otherwise a DAC would sound as is, and which is always "wrong". So now it can be tweaked to some degree, which is beyond its inherent design.

 

4.

Because PeterSt can not achieve that immune DAC (ref. 3), he can easily be prooven wrong. Here too, this, or the opposite may reach some final workout at some stage. Next year, whatever. Could be the same date as implied under [intermezzo] above.

 

I think this is logic (once it can be followed).

Haha.

 

Peter

 

 

 

 

 

[Miska]

Is there some logic I'm missing here?

 

I didn't make claims about unresolving systems, so just the conditional beginning "if my system is then" ... [considered by others] ...

 

Mostly trying to point out lack of logic in claim of systems not exhibiting the symptom being unresolving, maybe in a slightly sarcastic way. Since these are two different things which may or may not exist in the same system.

 

So sometimes system may not be resolving enough to show the difference when it exists, and sometimes system may be very resolving while not having the problem of same data being sent out by different pieces of software sounding different. I did not try to deny the possibility of the difference, just trying to talk about the known factors why it can happen. It is also important to consider correct causality of the factors. So that we can try to start curing the disease instead of treating the symptoms (which may still be the only choice in case of terminal illness).

 

 

[Miska]

If your computer -but the complete network of interconnecting devices- is under the best control for power management, only then you can perceive very well what DSP like appliances in the player incur for, because everything is now audible the best, and not influenced by sh*t out of your control underway.

 

Sort of, but the level of audibility is quite easy to find out by trying. If one can hear differences between different resampling algorithms and dither/noise-shapers, then the system is quite resolving. But what I have to wonder is how bad the system would have to be, for something like digital room correction not being audible...

 

At least I would be trying to get rid of all the annoying randomness from the system while preserving that "resolving" ability.

 

By the way, one my tests is quite good stress test:

1) Start playing music

2) cd /usr/src/linux ; make clean ; make -j4

3) Observe any changes in player output

 

 

[Jud]

...for this further helpful discussion, and my apologies if my last comment sounded a little, er, "abrupt."

 

Mostly trying to point out lack of logic in claim of systems not exhibiting the symptom being unresolving, maybe in a slightly sarcastic way.

 

Sure, though I'd suppose(?) there are enough differences between players in actual intended effect on the sound to account for many of the differences people hear. Of course there are also the unintended effects you've mentioned. I think I am really saying much the same thing as you do below:

 

So sometimes system may not be resolving enough to show the difference when it exists, and sometimes system may be very resolving while not having the problem of same data being sent out by different pieces of software sounding different.

 

- though, like Peter (or at least as I understand him to be saying), I would think "not resolving" would account for the overwhelming majority of cases of players sounding identical.

 

I did not try to deny the possibility of the difference, just trying to talk about the known factors why it can happen. It is also important to consider correct causality of the factors. So that we can try to start curing the disease instead of treating the symptoms

 

Makes sense. Let me say, though, that assuming we know a lot or even nearly everything about such a complex subject (audio, electronic, and software engineering) in such a chaotic environment (our homes, the insides of the hardware, the connections between them) makes me uncomfortable, so I nearly always prefer leaving lots of room beyond what we do know for causes of what we hear that haven't been fully worked out yet.

 

(which may still be the only choice in case of terminal illness).

 

Whew, that sounds a bit dire! ;-)

 

[Miska]

Sure, though I'd suppose(?) there are enough differences between players in actual intended effect on the sound to account for many of the differences people hear. Of course there are also the unintended effects you've mentioned.

 

OK, if those are the unintended effect. What is the technical mechanism of this intended effect if the players are sending the same data to an async interface?

 

If my technical explanation is about something unintended, I would like to hear the technical explanation of the intended part. Fair enough? Hands-waving doesn't suffice.

 

I would think "not resolving" would account for the overwhelming majority of cases of players sounding identical.

 

That's quite easy to test out by using different levels of modifications in the actual data. Or is the equipment somehow intellectually selective to know when to stop being resolving?

 

 

[Jud]

Miska asks:

 

What is the technical mechanism of this intended effect [difference in sound between players] if the players are sending the same data to an async interface?

 

A very fair question. Let me clearly state why I won't answer it! :-) I'm not an engineer or scientist (you're shocked, right?), just someone who has listened to lots of live and recorded music over many years. (But I will eventually circle back to your question, OK?)

 

Though I'm not a professional in the field, natural curiosity started me more than 40 years ago trying to figure out how I could best hear the music I loved. I tend to be fairly systematic and thorough when making large purchases, so when I finally started to make a good living and could afford a nice system (which was around the time people began to offer the first modified CD players and separate DACs), I took almost a year traveling most weekends to audio stores and engaging in long listening sessions. From Eastern Pennsylvania, where I still live, it was easy to find New York and Philadelphia-area stores offering any equipment within my budget (and lots that was vastly over).

 

My collection was pretty much all LPs at the time, so I did a lot of listening to vinyl to select speakers, cables, amp, preamp. Also, since vinyl was becoming less and less available, I knew I'd have to get something to play CDs. I began listening to CD players and DACs with the equipment I'd assembled so far - Vandersteen 2Ci speakers, PSE Studio SL preamp and PSE Studio IV amp, mostly Audioquest cables. (I first was made aware of PSE and Audioquest when I called Vandersteen, which I preferred to any other speakers in my price range, and Richard Vandersteen himself answered the phone. Ah, things were simpler then.

 

In my vinyl listening sessions I'd often played "Catembe," the first track from Miles Davis' "Amandla" album. The opening seconds in particular have emphatic percussion and plucked electric bass, followed by an ethereal scream from Davis' trumpet that would make the hair stand up on the back of your neck. So in preparation for listening to CD playback, I bought the CD of "Amandla."

 

No hair standing up. Here I was listening to $2000 CD playback systems in a format that was supposedly "perfect sound forever," and vinyl playing on my $400 Systemdek XII at home blew them out of the water. So I began paying attention to people who were complaining about the sound of CDs.

 

I have no doubt the sound of the "Amandla" CD was mainly the victim of the quick-and-dirty digital transfers common at the time. But something else I heard in the meantime that absolutely blew my mind was the difference in digital cables! I mean, "bits is bits," right? I had no technical knowledge or background to disagree with all the audio engineers and other self-styled experts who were saying that a mere wire changing the sound of digital playback was impossible if it didn't alter the 1s and 0s, but on the other hand, I was hearing differences that were absolutely not subtle. Just as important, so were my then-girlfriend-now-wife and anyone else who happened to be in the room. It was only after reading whatever I could get my hands on (pre-Internet days, so this took some time and effort), phone calls to people in the industry, and lots of listening that I believed "jitter" was something other than a marketing term used by people who wanted you to buy too-expensive hardware.

 

OK, enough of my old stories, let's get back to your question. Let me tell you my setup: MacBook Pro -> Audioquest Emerald USB -> V-Link (which has async USB transfer mode) -> Omega Mikro Zephyr S/PDIF -> Theta Pro Basic II -> Omega Mikro analog -> Spectral DMC-12 -> Omega Mikro analog -> PSE Studio IV -> bi-wired Audioquest Cobalt speaker cables with WBT terminations -> Vandersteen 2Ce speakers on Sound Anchors stands. I'm playing 16/44.1 material ripped from my CDs to ALAC or AIFF, no upsampling in ripping or playback. I'm using Audirvana, Decibel (was Ayrewave), and Pure Music as players, no other "applications" up (though whatever services normally run on Snow Leopard, tweaked to run exclusively in 64-bit, are running).

 

And on this setup, I am hearing absolutely consistent differences between players. I know Pure Music is proprietary, not sure about Decibel, but Audirvana is open source (though Mac). So perhaps you can look at the code, or borrow a Mac and run tests (or if there are tests I can run on the Mac using free software, let me know), and tell me whether Audirvana at least is altering the data in this situation, or if the differences must be coming from somewhere else.

 

If somewhere else, of course my imagination is a candidate, but I doubt my imagination would be so reproducible, or transferable to my wife. The differences are so evident that when I ask my wife which she prefers, she is able to indicate her choice within the first few seconds of hearing a selection through a second player. (And my wife, as I like to point out, is no audiophile - she puts up with my "testing," but she's really just waiting for me to be done so she can get back to watching Oprah.)

 

[Miska]

A very fair question. Let me clearly state why I won't answer it! :-) I'm not an engineer or scientist (you're shocked, right?), just someone who has listened to lots of live and recorded music over many years. (But I will eventually circle back to your question, OK?)

 

Well you didn't return to the question. But intended effect cannot be technically unknown effect? Unintended effect could be. So we have to start with assumption that known effects == intended effects. And unknown effects == unintended effects. (?)

 

Now, instead of just waving my hands and talking about what is not known, I've been trying to lay out what I know to effect digital to analog playback. And I welcome all the others to join. It all may not be nice and comfy. Or is this some kind of "thou shall not know why it happens, it shall be upon your ears for eternity"?

 

Note that I'm not claiming that there wouldn't be differences in some cases. What I'm trying to emphasize is that even though something partially manifests itself through playback software (and all other used software components) it may not be due to what the software is specifically and intentionally doing. You can play with a rays of light using lenses and mirrors, but it still doesn't change amount of light nor it's source. I'm chasing the source.

 

What I'm also trying to state is the goal - that even EMP from a nuclear blast cannot change the fidelity of the sound the slightest.

 

 

[Paul R]

I can 100% assure you the differences you hear are not imaginary, but I cannot explain, with any confidence, why. Other than as you say, they are changing the bits around.

 

I don't have an analysis rig good enough to get a definitive answer from. But I am looking to borrow the use of one. Have to scratch the itch to know!

 

On the other hand, those differences do not show up in every system, and some of the systems that have no audible difference between players, are very resolving systems indeed.

 

It's a mystery, with facts that apparently contradict each other. When we finally get the answers though, I bet it turns out to be something relatively simple.

 

-Paul

 

 

[Jud]

Miska writes:

 

Well you didn't return to the question.

 

? I thought when I wrote later in that message "OK, enough of my old stories, let's get back to your question," it would indicate that what followed was my attempt to respond. But if I wasn't clear, let me try to say it another way.

 

- I provided my hardware setup in case that would be at all helpful in determining what might cause the differences between players, such as noting there's an async USB transfer mode interface in there. I also talked about the software environment. I didn't note, but will now, that I do no digital frequency EQ. I have the attenuation on all the players and the computer itself set to 0 db. So far as I'm aware, I've got the Mac, the players and the hardware set not to change bit depth or resolution. And as I also mentioned, the source files are all straight ALAC or AIFF 16/44.1 rips of my CDs.

 

- Now as I said, I'm not qualified to say with any assurance what is causing the differences I'm hearing between the 3 players I've talked about.

 

- In an effort to see if we can learn any more about what the cause might be, I mentioned that Audirvana is open source, so perhaps (I don't know for sure) you can look at the source code of that particular player and tell me if it is doing anything to change the bits, knowing I've set it not to upsample and to 0 db attenuation. If you might know of anything I could look at or any tests I might run to see if the data is being changed by the player(s), I would be happy to do so.

 

- If we were able to determine by one or more of these methods that the players, or at least Audirvana, wasn't altering the bits, then we might look to other reasons for differences. Audirvana loads the music files into memory; I believe Decibel does; and Pure Music can be set to do so (that's how I have it set in tests). So we won't find our differences there. Perhaps it's in unintended hardware effects as you've mentioned might be possible, or perhaps it's in something else I haven't thought of.

 

So - to sum up, I hear differences, don't know what's causing them, and am happy to do some guided amateur research into finding out what the cause(s) may be.

 

 

[Miska]

I thought when I wrote later in that message

 

True, sorry, I somehow missed that sentence.

 

V-Link

 

At least this has some potential being sensitive. It takes a while to arrive here, but now I just got my hands on latest Stereophile for the image of the inside.

 

AFAIK, it uses power from USB and seems to run it through just a normal regulator which also means it has USB ground connected. I don't see galvanic isolation of the data lines either. And I don't see isolation transformer for the coax output either, so this device should be used only with DACs having transformers on input or even better over optical connection. So there's a possibility of clock poisoning from USB or coax cable.

 

USB cable with ferrite beads may improve things with this gadget.

 

I did not find good enough image of your DAC to tell what kind of input construction it has.

 

For USB -> S/PDIF converters so far the best I've seen has been M2Tech hiFace Evo. And for USB-input DACs the most elegant Ayre QB-9 (image) if those chips before the ribbon cable are opto-isolators. Only caveat with the Ayre is that I cannot tell from the images if the USB and DAC side have ground connection between the two (ie. does it isolate the two grounds too?).

 

No I don't have Ayre and V-Link has not arrived yet here at the backyard of the Europe. Everything with the disclaimer that of course I may have guessed something wrong from the images...

 

 

[Terry]

I love these arguments it kind of reminds me of the one is the earth round or is it flat.