[shriramosu]

I understand that one of the main advantages of PC audio over CDP's is reducing/eliminating jitter (others are error correction, ability to upsample, apply cheaper room correction).

Specifically with respect to jitter it seems like the higher end devices like Empirical audio's solution, using battery supply, doing DAC separately etc are ways to reduce jitter.

My question is what points in the system for PC audio can jitter enter the data stream and how? (starting from say whats high jitter, from a soundcard out, to less using USB to Coax to receiver in, then even less using USB to I2S to DAC) etc.

Thanks

Shriram

 

[Logistics]

If you want to minimize jitter, one thing which is important is having clean power. This is much easier with batteries. But on a PC you'll need to clean things up. Jitter can happen in part from oscillations in the IC's. Often these oscillations are due to dirty or inadequate power... or both. Ways to remedy this could be recapping your hardware, power conditioning, etc.

This is why it irks me when people start going on about switching to better DAC's on their sound card or moving to an entirely different card. Sure, you might notice some improvement, but it seems a little like someone suggesting you swap an NSX engine into your stock Honda Civic. While it may be faster in some ways, the car would have been a whole lot more fun by simply upgrading to sporty suspension, better/lighter wheels and good, sticky tires... and it would make through a road-course a whole lot better than the stock car with the NSX engine. (this is hypothetical, I don't care that an NSX engine won't fit in a Civic, hehe)

...uhh, the illustration is supposed to be representative of a system with cruddy power and hardware from a component level standpoint, but with an expensive, renowned card vs. a system with clean power and hardware from a component level standpoint, but with a lesser card.

 

[003]

Ok to clarify, PCs have MORE jitter than CDPs.

Empirical audio's products are designed to bring PCs to the almost non existent levels of jitter found in high end CDPs.

 

[Gatticus]

No, PC's have less jitter than CDP's. I have an article bookmarked on my XP system that proves it and will post it later. I'm in Vista right now so can't. If you want the least amount of jitter use an external soundcard with it's own PSU and use digital coax connection from the PC soundcard. Optical digital connection has more jitter than coax. Doing that will give you the least amount of jitter, less than CDP for sure.

If I put an 8 cylinder 300hp engine into my 99' Oldsmobile Alero 2 door Coupe it will beat a Mustang GT. I've got the sticky and wider low profile tires, aluminum wheels and it has Mcpherson suspension system. Porsche 911 uses the same suspension system as well as all BMW cars except one. Mustang GT suspension would only be better if it uses double wishbone. I don't know what the Mustang GT uses for suspension though but based on thrust to weight ratio it will beat it easily.

 

[Redo]

Quote:

Originally Posted by Gatticus /img/forum/go_quote.gif No, PC's have less jitter than CDP's. I have an article bookmarked on my XP system that proves it and will post it later. I'm in Vista right now so can't. If you want the least amount of jitter use an external soundcard with it's own PSU and use digital coax connection from the PC soundcard. Optical digital connection has more jitter than coax. Doing that will give you the least amount of jitter, even less than CDP.

That's all dependent on the external sound card. I doubt most external sound cards can match the nicer dedicated CDPs for jitter, though I don't really have any proof one way or the other.

 

[Dept_of_Alchemy]

This whole jitter thing is so silly sometimes. Back in the day, improvement in DAC design has actual meaning like flater FR, lower noise floor, lower distortion and better filters. Nowadays all that has been fixed and it's about this whole jitter game, which we don't even know for certain that it matters (certainly not any rigorous science has demostrated picosecond jitter makes a difference). Sorry to thread crap, but it's just so frustrating sometimes to see people worrying about (not to mention spending hundreds to fix) something that very well may not even matter.

Alright, time to contribute.

I think CDPs generally have lower jitter than PC audio due to the fact that CDPs have clock(s) that are made for one thing only. Products from Empirical Audio sometimes claim excellent jitter performance but you can hardly consider that as PC-audio per se, it's more like a clock upgrade for your transport.

 

[003]

Quote:

Originally Posted by Gatticus /img/forum/go_quote.gif No, PC's have less jitter than CDP's. I have an article bookmarked on my XP system that proves it and will post it later. I'm in Vista right now so can't. If you want the least amount of jitter use an external soundcard with it's own PSU and use digital coax connection from the PC soundcard. Optical digital connection has more jitter than coax. Doing that will give you the least amount of jitter, less than CDP for sure.

That may be true if you're using an Empirical Audio digital output solution, but if not, no... a really good CDP would make a computer pale in terms of jitter. If there is an article that says otherwise, it's either referring to using an Empirical Audio digital output, or it's just wrong (maybe even made up)...

Coaxial is better than optical, but in reality they are both poor digital in/out standards. Direct I2S (not from USB), BNC, or some kind of bit or clock sync connection destroys both optical and coaxial.

 

[cotdt]

Use PC -> DAC via I2S connection. It has lower jitter than CD Players. Or even with SPDIF, you can add a clock. However, after looking into the subject I'm not sure how much jitter really matters. Most likely it is the clock noise that is causing the sound degradation, so a less noisy clock would improve sound greatly.

 

[Gatticus]

Quote:

Originally Posted by Redo /img/forum/go_quote.gif That's all dependent on the external sound card. I doubt most external sound cards can match the nicer dedicated CDPs for jitter, though I don't really have any proof one way or the other.

I don't really care about jitter anyway as I can't hear a difference. Just pointing out that what some of guys are saying is wrong.

http://www.positive-feedback.com/Issue22/nugent.htm

 

[bigshot]

I've concluded that my system only has jitter when a salesman is trying to sell me new equipment.

See ya
Steve

 

[Logistics]

Quote:

Originally Posted by Gatticus /img/forum/go_quote.gif No, PC's have less jitter than CDP's. I have an article bookmarked on my XP system that proves it and will post it later. If you want the least amount of jitter use an external soundcard with it's own PSU and use digital coax connection from the PC soundcard. Optical digital connection has more jitter than coax. Doing that will give you the least amount of jitter, less than CDP for sure.

The article is about how a PC can supposedly be a better transport than "optical disc systems." It is not about how PC's can play CD's better than stand-alone CD players. This suggests the author feels that hard-drive based audio is better. He does, however warn against it in the case you are using USB to connect to an external DAC. He also states that only a "well implemented" PC can be better, whatever that may be. He makes mention of how PC's suffer, typically from dirty power. That's a very vague article.

In any case, I think it's dumb the way people worry more about attaining bit-perfect audio when they have junk for electronic components in their hardware which will have a much larger impact on the sound.

 

[Gatticus]

Did you even read the opening paragraph? Here, lets make sure you did.

"I am convinced that computers are the future of high-performance audio, and not optical discs. Besides having a number of ergonomic advantages, such as the ability to build play-lists of favorite tracks, organize music like a file system and download tracks and artist information from the web, there are a number of technical reasons why this technology is inherently superior in performance, particularly when compared to the current generation of optical players. The current generation of CD and DVD players cannot compete with computer systems that are well implemented, but future generations may as they start to resemble computer-driven systems. The primary audio quality improvement that is possible with computer driven audio is a significant reduction in jitter over that of optical disk systems."

Whether you think being concerned about jitter is dumb or not is irrelevant to the discussion at hand. The fact is that PC as source has less jitter than CDP. Some people were saying otherwise. They are simply wrong. End of discussion.

 

[LeonvB]

Quote:

He also states that only a "well implemented" PC can be better, whatever that may be.

Please do note the article is somewhat dated. Todays dual core PC's have PLENTY processing power to make sure data arrives on time at the USB DAC. I've tried overrunning my PC with 4 simultaneous FLAC encoding tasks, a rip and playing music to my Stello. It never popped or stopped playing (to my surprise I might add). And to add: even the rip didn't suffer too much from the processing being done by the other tasks. To compare: on my "old" AMD 3500+, doing this would result in a rip being done at 0.2x realtime with even only 1 FLAC encoding task.

On the jitter: aside from the problems in proving jitter is actually a noticable problem in todays electronics, in a PC to USB DAC setup there never is any jitter coming from the PC. The clock used to create the music is the one inside the DAC itself. So this type of setup has the lowest amount of jitter, simply because in a CDP the drive unit and the electronics on it add jitter.

 

[audioengr]

Quote:

Originally Posted by LeonvB /img/forum/go_quote.gif Please do note the article is somewhat dated. Todays dual core PC's have PLENTY processing power to make sure data arrives on time at the USB DAC. I've tried overrunning my PC with 4 simultaneous FLAC encoding tasks, a rip and playing music to my Stello. It never popped or stopped playing (to my surprise I might add). And to add: even the rip didn't suffer too much from the processing being done by the other tasks. To compare: on my "old" AMD 3500+, doing this would result in a rip being done at 0.2x realtime with even only 1 FLAC encoding task. On the jitter: aside from the problems in proving jitter is actually a noticable problem in todays electronics, in a PC to USB DAC setup there never is any jitter coming from the PC. The clock used to create the music is the one inside the DAC itself.

Not true. Some DAC's that that asynchronously upsample have a clock inside, but even these are not insensitive to incoming jitter. Most DAC's that do not upsample have no clock inside.

There is a clock required for USB interface and in some chips, such as the TAS1020 and the TUSB3200 the incoming USB jitter can be reduced by this clock. The jitter reduction from these chips depends entirely on the jitter of the clock that is driving them. A low-jitter clock such as a Superclock yields a good result. These chips are also somewhat sensitive to incoming jitter on USB.

There can be a LOT of jitter in the USB signal from a computer.

Three techniques can be used to reduce it:

1) asynchronous upsampling - good (because the upsampling limits your choices and you are stuck with the upsampling algorithm)
2) low-jitter clock in the USB conversion - better
3) reclocker - best

And BTW, when I say "well-implemented", I'm talking about the USB interface, not the PC.

Steve N.
Empirical Audio

 

[bigshot]

Quote:

Originally Posted by audioengr /img/forum/go_quote.gif There can be a LOT of jitter in the USB signal from a computer.

And an awful lot of angels can dance on the head of a pin. Jitter is not audible in properly operating equipment of modest quality. No need to worry about whether one system has more or less when it's not anywhere near the level of audibility anyway.

See ya
Steve