iTunes or Looney Tunes? The great music server debate.

[MikeSp]

Being new to this forum (and feeling lucky to have found it), I read all of this thread with interest since I would like to build a dedicated media server. The topic appeared to have turned into a PC vs. Apple debate which was inevitable. I build my own PC's since I have total control over component choices and purchasing -- am still running Vista 32 and never had a crash or driver issues and have some potent software running to support my digital darkroom hobby.

I must ask a really dumb question -- WHY is it preferred or considered necessary to have a dedicated DAC between the music server and the preamp? In my case, I am using a pre-pro that has both analog pass-through as well as digital DSP and Lyngdorf Room Perfect. IF I want to use the Room Perfect corrections for the room, it must occur in the digital domain and it seems like a lot of conversions to go from digital to analog to digital and back to analog out to the amps when the pathway could have stayed totally in the digital domain until the final conversion to analog that is sent to the amps:

digital (from music server) --> analog (in DAC) --> digital (in pre-pro for Room Perfect) --> analog (in pre-pro) and out to the amplifers [so many conversions??]

thanks,

MikeSp

 

[amirm]

I must ask a really dumb question -- WHY is it preferred or considered necessary to have a dedicated DAC between the music server and the preamp? In my case, I am using a pre-pro that has both analog pass-through as well as digital DSP and Lyngdorf Room Perfect. IF I want to use the Room Perfect corrections for the room, it must occur in the digital domain and it seems like a lot of conversions to go from digital to analog to digital and back to analog out to the amps when the pathway could have stayed totally in the digital domain until the final conversion to analog that is sent to the amps:

digital (from music server) --> analog (in DAC) --> digital (in pre-pro for Room Perfect) --> analog (in pre-pro) and out to the amplifers [so many conversions??]

MikeSp

Not a dumb question at all. First, let me welcome you to the forum.

Now to your question. You most definitely can go through your processor. Many of the people in this forum have dedicated 2-channel systems with no signal processing whatsoever so for them a DAC is the only thing they need to feed the rest of their analog system. Your case is different and you can certainly take the S/PDIF coax or Toslink optical from the PC and run it into your processor and use that for the rest of the chain.

That said, a dedicated DAC is likely to have better performance than IC DACs in a processor of reasonable cost. So there may be slight improvement there. Unfortunately if you go back to digital domain in your processor, that improvement will likely be lost so in your case, I would say with the processor.

 

[amirm]

BTW, while it is generally understood that optical TosLink has lower fidelity than Coax S/PDIF, I recommend per above to try both connections if they are both available. Optical link isolates the PC from the audio equipment and that might have a more positive effect than lower jitter on Coax connection.

It is also eye opening to have both connected at once and switching between them and hearing a marked difference! Definitely makes the camp who thinks "digital is digital" think!

 

[JackD201]

Who from PNWAS can get John Tucker in here? We had a discussion, well I tried to keep up but couldn't, about why he thinks iTunes is not up to scratch so he made his own. My feelings were hurt but what I heard at FIM's room at CES '08 with speakers, pre and amps I'm familiar with, made me think. If I wasn't such an analog hound lusting for the Technics R2R I might have paid closer attention, not that it would have helped me understand all the rocket scientist speak. I believe he did lots of work at JPL. He makes a darned good DAC too.

 

[flez007]

BTW, while it is generally understood that optical TosLink has lower fidelity than Coax S/PDIF, I recommend per above to try both connections if they are both available. Optical link isolates the PC from the audio equipment and that might have a more positive effect than lower jitter on Coax connection.

It is also eye opening to have both connected at once and switching between them and hearing a marked difference! Definitely makes the camp who thinks "digital is digital" think!

Amir - i also thought that this was the case (toslink/coax) but in my system, the toslink route yields a more defined presentation, image and real-sized instruments compared to the coax output which is more warm but tends to colorate the signal... I wonder why?

 

[Vincent Kars]

First of all if we talk about differences between SPDIF coax or Toslink we talking a bit of nonsense. We never hear the protocol, we always hear the implementation.
Results therefore might vary depending on the components used.
A badly implemented electrical SPDIF can generate tons of input jitter.

However, these pictures demonstrated what is believed to be the case

Obvious the jitter level of the Toslink is about 7 times higher than the coax but even 70 ps is a rather low value.
Next question is how does our DAC copes with the input jitter of these connections.
Often DACs do have their means like PLLs to reduce the input jitter.
It is possible that both connections remain below the audible threshold of jitter.
http://thewelltemperedcomputer.com/KB/BitPerfectJitter.htm

If this is the case, the complete galvanic isolation as offered by Toslink might have an advantage as no stray signals can disturb the DAC
Source: http://www.hometheaterhifi.com/blu-...edition-universal-blu-ray-player.html?start=1

 

[Phelonious Ponk]

First of all if we talk about differences between SPDIF coax or Toslink we talking a bit of nonsense. We never hear the protocol, we always hear the implementation.
Results therefore might vary depending on the components used.
A badly implemented electrical SPDIF can generate tons of input jitter.

However, these pictures demonstrated what is believed to be the case

Obvious the jitter level of the Toslink is about 7 times higher than the coax but even 70 ps is a rather low value.
Next question is how does our DAC copes with the input jitter of these connections.
Often DACs do have their means like PLLs to reduce the input jitter.
It is possible that both connections remain below the audible threshold of jitter.
http://thewelltemperedcomputer.com/KB/BitPerfectJitter.htm

If this is the case, the complete galvanic isolation as offered by Toslink might have an advantage as no stray signals can disturb the DAC
Source: http://www.hometheaterhifi.com/blu-...edition-universal-blu-ray-player.html?start=1

One might even go so far as to say it is likely that both connections remain below the audible threshold of jitter. If one happened to be just a bit skeptical of things heard by audiophiles straining to hear things that are only heard by audiophiles, one might say that unless the design or the execution is pretty screwed up, it is almost certain that both connections remain below the audible threshold of jitter. Then one might say buy decent digital equipment with the connections that are convenient, but I digress....



Tim

 

[amirm]

Amir - i also thought that this was the case (toslink/coax) but in my system, the toslink route yields a more defined presentation, image and real-sized instruments compared to the coax output which is more warm but tends to colorate the signal... I wonder why?

I touched on the reasons why. PC typically has a horrible ground. It has a processor running at microwave frequencies pounding on it. It has untold number of clocks running, banging on the same rail. And of course, lots of RF radiation, all of which bleeds through that ground connection to your audio system. Some of these also act as jitter while others are analog interference screwing up reference circuits in your DAC, etc. So it is not surprising that optical which isolates this connection, helps a lot.

Another solution if you have a pro card and DAC is to use AES balanced digital connection. Even here, you want to do an A/B to be sure you pick the one you like.

 

[amirm]

One might even go so far as to say it is likely that both connections remain below the audible threshold of jitter.

Tim

Tim, as I just described, there are lot more reasons for distortion other than jitter. I won't rehash everything discussed in the jitter thread either. I will say this though: you have a hypothesis that these two links sound the same. Why not go and see if you can prove it to yourself? Connect both to a PC and switch between them. Then come back and report what you heard. I have done this many times and the differences can be surprising. It shocked the hell out of me the first time I heard it as I wasn't prepared at all for what I heard. This doesn't mean I didn't have bias but it was an educational experience. I am suggesting you go to the same school .

 

[amirm]

First of all if we talk about differences between SPDIF coax or Toslink we talking a bit of nonsense.

We are not talking about that topic in general. We are talking about them in the context of a PC being the source. It makes a difference.

We never hear the protocol, we always hear the implementation.
Results therefore might vary depending on the components used.

Right and hence the reason if people have both connections, they should simply test them and not make assumptions that are not transportable to their situation. To wit, there is not a single measurement in the world for your PC driving your DAC. It might have 1000 times more jitter than Oppo and you wouldn't know it.

However, these pictures demonstrated what is believed to be the case

This is the case of one test and one player. It has no bearing on all cases and those involving PCs.

Next question is how does our DAC copes with the input jitter of these connections.
Often DACs do have their means like PLLs to reduce the input jitter.

PLLs are not a panacea. The analogy in the article is correct but doesn't work in digital world as a system. A digital playback system must maintain the timing of the samples as they arrive from the source. For example, if I send 41,001 samples/second because my clock runs that little bit faster, the DAC better reproduce that many samples or else, it gets out of sync with the source. If the DAC has to allow such timing change to be preserved, then how on earth does it know that deviation was not created due to jitter timing error? In real life, PLLs filter some amount of jitter but let others through. For example, they can opt to filter high frequency jitter because that is less likely to be created by the source on purpose.

It is possible that both connections remain below the audible threshold of jitter.
http://thewelltemperedcomputer.com/KB/BitPerfectJitter.htm

Please read the jitter debate thread where I showed that there is no published paper to give you comfort as far as audibility of jitter. See the second paragraph in the first post: http://www.whatsbestforum.com/showthread.php?1151-Audible-Jitter-amirm-vs-Ethan-Winer

 

[Vincent Kars]

We are not talking about that topic in general. We are talking about them in the context of a PC being the source. It makes a difference.

It doesn’t. Even in the context of a ‘PC'‘ implementations vary.

This is the case of one test and one player. It has no bearing on all cases and those involving PCs.

Do you really think I believe a n=1 case conclusive?
Do you really think I can't tell the difference between a PC and a Blue-ray player?
What I’m referring to is the believe in audiophile circles that Toslink sucks and Coax is so much better.

PLLs are not a panacea.

That is exactly why I use the word ‘reduce’ instead of 'eliminate'

Please read the jitter debate thread where I showed that there is no published paper to give you comfort as far as audibility of jitter.

As you can read on my website (http://thewelltemperedcomputer.com/KB/BitPerfectJitter.htm) I’m pretty aware that opinions about the audibility vary

 

[amirm]

Now you have me confused. I said to OP to try both links to see if he hears a difference. What was the point you were making? It sure sounded like you were saying the difference is too small to normally hear. That is what I challenged in the context of PC sources. Let me know your position first and then we can figure out if we should dig deeper.

 

[Vincent Kars]

My position is a very simple one.
If you hang around to long on to many audio forums you simply start to wonder.
All those tweaks, fanless, low power, memory playback,etc, etc. do they make a differences?
Are the differences reported (blown away seems to be a minimum) real or are they audiophile differences?

That’s why I’m delighted with the measurements of the Oppo.
You can see the difference between running and stopped mode
You can see the difference between Toslink and coax.
For me this is the first time I see some measurements demonstrating the impact of electrical activity on jitter performance. Likewise Toslink – coax.

Next: what do these measurements mean?
Will they map into a audible difference?
If I apply what I know to be the most severe (Julian Dunn, 20 ps ast 20.000) the difference is small but should be audible say between 18.000 and 20.000
If we believe Rumsey & Watkinson (1995) anything below 10 ns goes.
As delighted as I am with at least some measurements, how to interpreted them is another.

Of course you can’t extrapolate the Oppo to a PC but at least this measurements probably show on a small scale what happens in a PC on a large scale (bit more clocks, motors, fans, etc)

It is also eye opening to have both connected at once and switching between them and hearing a marked difference! Definitely makes the camp who thinks "digital is digital" think!

Total agreement here.
If you happen to have different digital connections, do compare them!
I did compare Toslink and USB as both the PC (iMAC/Win7) and the DAC (Benchmark) do have USB and Toslink.
Much to my surprise I do hear a differences (the DAC uses ASRC, supposed to eliminated all the input jitter) and in favour of Toslink

Hope to be less unclear this time

Vincent

 

[amirm]

Hope to be less unclear this time

Vincent

Mission accomplished. Little worthwhile to add or critique .

 

[flez007]

I touched on the reasons why. PC typically has a horrible ground. It has a processor running at microwave frequencies pounding on it. It has untold number of clocks running, banging on the same rail. And of course, lots of RF radiation, all of which bleeds through that ground connection to your audio system. Some of these also act as jitter while others are analog interference screwing up reference circuits in your DAC, etc. So it is not surprising that optical which isolates this connection, helps a lot.

Another solution if you have a pro card and DAC is to use AES balanced digital connection. Even here, you want to do an A/B to be sure you pick the one you like.

Sorry Amir, I forgot mentioning this is happening while using my Oppo as a transport...

 

[Phelonious Ponk]

Tim, as I just described, there are lot more reasons for distortion other than jitter. I won't rehash everything discussed in the jitter thread either. I will say this though: you have a hypothesis that these two links sound the same. Why not go and see if you can prove it to yourself? Connect both to a PC and switch between them. Then come back and report what you heard. I have done this many times and the differences can be surprising. It shocked the hell out of me the first time I heard it as I wasn't prepared at all for what I heard. This doesn't mean I didn't have bias but it was an educational experience. I am suggesting you go to the same school .

I've already compared optical and coax on my headphone system, where I have that option. The differences fall in the "I think I might hear..." range, slightly in favor of optical. This is consistent with my belief that jitter is practically inaudible, giving the slight edge of good galvanic isolation to optical. Of course implementations vary, but seriously, if there is a significant difference in favor of the connection capable of carrying the noise of the PC/transport, your other "reasons for distortion" to the DAC, are we not simply talking about poor implementation? Or perhaps a listener's preference for a bit of noise?

I'm not trying to re-live the jitter debate, Amir, I am merely expressing my point of view in the broader dialogue. If it sometimes seems repetitious, my apologies. My only defense is that in the audiophile community, the pragmatic point of view is not broadly repetitious but actually quite rare compared to the views in support of audiophile digital cables, sonic differences between bit-identical files, DACs that charge a premium for reducing jitter that is already sufficiently reduced while connecting the analog system electrically to the noise of the digital system, and the superiority of, of all things, USB. If I seem to repeat myself it is only because there are so few, like Vincent, who speak my mind far better than I can speak it for myself. Expound upon the sonic virtues of a premium USB cable on many computer audio boards, and a few hundred heads will nod in unison. You would probably get more positive feedback from frozen cds than from suggesting that jitter is not really an issue in competently designed and executed gear.

It's not easy being me.

hmmm....I should probably sign this one,

P

 

[Vincent Kars]

Expound upon the sonic virtues of a premium USB cable…

I expect a lot of fun the oncoming years.
A couple of USB audio class 2 DAC’s are now shipping and more to come.
Class 2 runs on 480 MHz instead of 12 MHz (class 1)
I expect a lot of these audiophile USB cables to fail the moment you use them for Class 2
Their build quality is probably so poor they can’t cope with this higher bandwidth.
Might also explain why they do have a sonic signature……

 

[LesAuber]

Just to stir up an old debate a bit. Plugged in the iPhone to sync and recieved a reinstall iTunes message as it was missing files needed to start. The install is on WinXP machine which has been boringly stable. iTunes also had been well behaved on this machine only once in a while needing reminded that its library resided on NAS instead of locally.

Up until today the closest thing I'd had to being an issue was that I needed to do iPhone OS updates on a Win 7 laptop with the DSL modem directly connected to prevent the update from timing out which is what happened on the XP box. Yes the laptop is faster but the XP box is still fast enough to keep up with DSL directly connected. Apple tech support has no idea why either. Just made the suggestion to try a different machine. But this strays rather OT.

So is it Windows or iTunes that is squirrely here? Or is it because I'm mixing dogs and cats er Windows and Apple?

 

[amirm]

Smells like an iTunes problem. Windows wouldn't tell you to re-install iTunes.