[racerxnet]

Sasha:

Did you test jitter at the Word Clock using a Wavecrest DTS? What DBT did you use and files to test the audibility of jitter?

Sometimes the discussion gets derailed by naysayers who want to impose their view on me and convince that what I am hearing is my imagination.

 

In scientific applications, the burden of proof is in your court, do you have proof regarding your statements. You drive over bridges and look at skyscrapers based on solid engineering, but then discount the valid testing methods when it comes to PC playback and subjective/objective testing methods. Go figure....

I do not know, and mainly because you cannot trust published figures, if you can find them in the first place.

You can find PDF files on quite a few DAC chips from the manufacturer. I can look up the specs on the ESS  Sabre 32 chipset in less than a minute. What specs are companies lying about? Care to provide some published specs that validate the position.

Changing speakers is likely to have a larger affect on the fidelity than the source.

AMEN to that statement. I posted this same suggestion at AA and NOT to try and shape sound through cable impedance mismatching and other goofy ideas. If the cable is of quality construction that passes the signal properly, there will be no audible difference.  I do recommend that if one dislikes the sound, consider room treatment and/or speaker replacement/placement. 

Sasha, if you could provide the specs of you PC based system, maybe  someone could help you in your quest for the sound you desire.

Here a jitter, there a jitter, everywhere a jitter, seems to be today's flavor for digital sonic problems. I'm just enjoying the music and show, while some try to define everything they hear from the system.

MAK

[Sasha]

I really do not get it, doesn?t anyone read what is being said in these discussions?
I have no intention of proving anything to anyone.
Wavecrest DTS? Are you out of your mind? You want me to spend money on it, do testing to prove something on these forums?
Chipset specs? What in the world chipset specs have to do with the performance of the final product?
Read what is being discussed, I will not repeat it.
Let me make it very simple, if you are convinced that 4ns or whatever silly level is inaudible, that all sources sound the same due to some miraculous jitter attenuation techniques, that some archeological artifact of a PC you put together in an ad-hoc manner sounds the same or better as best CDP, then enjoy the music and let those who seek better performance exchange ideas and suggestions.

[srb]

Hi All,

I have been searching hi and low for any scientific info on the delectability of Jitter in digital music playback

Well James, with all this debate on jitter, you are the first to say that it might be "delectable"    .  Gimme some of that delicious jitter.  Makes the music all mushy and tubey!  Yum!
 
Seriously, I wish there were some kind of accurate Jitter Meter (like $100 from Radio Shack, with coax, toslink, usb inputs/outputs that could measure this distortion anywhere you plug it into the chain!).
 
The reality is it takes either a mondo-expensive dedicated test instrument or a digital storage oscilloscope + a good dose of user know-how.
 
Still, I am passionate about reducing it, even though I don't really know what it sounds like!
 
Steve

[racerxnet]

 

let those who seek better performance exchange ideas and suggestions.

You appear to be a bit annoyed when questioned for factual data regarding the matter. And the last time I looked, this is a forum to exchange everyone's thoughts, whether you agree or not. If you think you are the judge , jury , and executioner on these matters you are deluded.

Chipset specs? What in the world chipset specs have to do with the performance of the final product?

Why not just use any old DAC chip from your perspective. And yes, the overall sound is dependent on many factors, including the DAC chip used.. The specs are useful in engineering the PCB and components mounted on the board for final use.. including the final output to the pre amp/amp.  http://www.twistedpearaudio.com/digital/buffalo.aspx

Let me make it very simple, if you are convinced that 4ns or whatever silly level is inaudible, that all sources sound the same due to some miraculous jitter attenuation techniques, that some archeological artifact of a PC you put together in an ad-hoc manner sounds the same or better as best CDP, then enjoy the music

You can learn some electronics applications here that others have tried in the past.
http://www.diyaudio.com/forums/forumdisplay.php?forumid=9



There are many trying to close and surpass the sonic qualities of PC playback; Cic's, Amarra, Foobar and many others. Some will say they have already accomplished this. Steve Nugent just one name off the top of my head. Your opinions are one of millions, but back up your rhetoric with facts please. Cic's just posted a synopsis of PC jitter at AA if you are interested.
http://www.audioasylum.com/cgi/vt.mpl?f=pcaudio&m=56101

Sasha, if you are in the Chicagoland area, take your CDP over and I would enjoy the company to compare the differences, if any on my 2 systems. They are both well above average fidelity wise.

What can we do to help. Are you wanting PC playback to rival your CDP? Are you suggesting that the PC is inferior for playback due to jitter?

MAK

[James Tanner]

Hi All,

I have been searching hi and low for any scientific info on the delectability of Jitter in digital music playback

Well James, with all this debate on jitter, you are the first to say that it might be "delectable"    .  Gimme some of that delicious jitter.  Makes the music all mushy and tubey!  Yum!
 
Seriously, I wish there were some kind of accurate Jitter Meter (like $100 from Radio Shack, with coax, toslink, usb inputs/outputs that could measure this distortion anywhere you plug it into the chain!).
 
The reality is it takes either a mondo-expensive dedicated test instrument or a digital storage oscilloscope + a good dose of user know-how.
 
Still, I am passionate about reducing it, even though I don't really know what it sounds like!
 
Steve

Hi Steve,

Thanks for catching that - boy do I look dumb 

[racerxnet]

Hi Steve,

Thanks for catching that - boy do I look dumb 

I think it tastes great,,,











Less filling.

[srb]

James, of course I knew you meant "detectability" !  (Funny how one just one letter can change everything).
 
Just trying to lighten up this sometimes tense thread.
 
Seriously, a lot of us have the Kill-A-Watt tester for $30 to measure the current draw of our equipment.  Now I just want the Jitt-No-Mo tester to troubleshoot my digital chain!  I am willing to pay three times the price.
 
Steve

[racerxnet]

Just trying to lighten up this sometimes tense thread.

I think most people do not thrive on challenge to ones beliefs. It becomes personal for no reason, and the mind becomes closed at that point. Call me skeptical, I'd rather base my knowledge on proven concepts in the field. This forum is a cake walk compared to Hydrogen Audio. Most people not having data to back up assertions are handed their a$$ on a plate. There are many very qualified people who frequent many forums, I just prefer to digest information with a GRAIN of SALT all the time.

It's about enjoying the music.

MAK

[Sasha]

Mak,

I am annoyed because I keep repeating one the same thing but it seems to fall on deaf ears.
If you read this entire thread you will find why I consider some of published information questionable, you will find details on my PC setup, you will find out how I compared PC to CDP performance and how I arrived to my conclusions, and finally what my conclusions were.
DC and I exchanged quite a few ideas, lots of information, a few others jumped in.
Why is it so hard to read all that was said before barging in?
I have read everything in the links you posted, I know what Steve Nugent does and what products he offers, and again, if you read what has been said you will understand I was not interested in USB in general. And I do not want to purchase anything, including Steve?s pace car, without getting some firsthand information from someone who shared my experience and my views.
I have read and applied cics? suggestions, I have participated for a long time in Computer Audio Asylum.
I do not want to waste time on ?backing up? my rhetoric, what is relevant was already said, it is absolutely clear why I do not consider PC audio to be all what is claimed to be.
Finally, yes, I want PC playback to rival CDP, not just any CDP but top CDP, and yes, PC is inferior due to jitter.
Majority of posts turn to pure waste of time because of fruitless pondering not related to the subject, I do not want to preach anything to anyone, be preached to or prove anything.
And then those baseless suggestions, it goes like this:
You ask specific question on comparative performance between A and B, someone barges in and says ?I just got B, could not be happier?.
You ask how A was compared to B, and the answer is ?never compared, I got B and never looked back?.
WTF?

[Crimson]

Sasha,

I believe it's the absolutist statements ("PC audio is not as good as the 'best' CDP") that can be somewhat annoying. You compared your CDP with your PC setup in your system and came to your conclusion. Others may have reached a different conclusion, myself included. 

[Sasha]

Sasha,

I believe it's the absolutist statements ("PC audio is not as good as the 'best' CDP") that can be somewhat annoying. You compared your CDP with your PC setup in your system and came to your conclusion. Others may have reached a different conclusion, myself included.

Right, and what would be accomplished if I try to convince you otherwise, or you try to convince me?
My views are shared by many, as are yours.
You are happy with the performance of your system, I am aspiring for better one.
Is something wrong with that?
And if I seek opinions and suggestions from those that may share my views because they shared my experience, had opportunity to take further steps in their quest for performance improvements and are willing to share their knowledge, is that wrong?
Is this not the place to seek such information?
Do I have to listen to continuous drivel about something that is completely irrelevant, or that I know is completely wrong because my experience thought me otherwise?
I do not understand why it is so hard to comprehend.

[racerxnet]

 Sasha:

Fair enough on your reply that I re-read the 4 pages and come to the conclusion that the optimized PC and Lynx card did not meet your sonic desires. I use the RME 9652 and prefer to use an external DAC for playback. One reason is the jitter rejection within the ESS Sabre DAC. Others reasons are more obvious; external DAC environment from the PC, better PS and so forth.

Other things I have done for myself are to externally locate the sound card outside the PC with a Adex riser.
http://www.adexelec.com/


Most important for me was not using the DACS on the RME at all and just transfer the signal to the external DAC. The ESS chip has very good jitter rejection per the data sheet and had been confirmed in testing. I believe in engineering the best solution I can afford before assembly, to limit my dissatisfaction. Still, the speakers and room will have the most impact on SQ of all. Unless the DAC is a real piece of crap, the sonic differences I hear are minimal at normal listening levels.

I read that you are not a believer in USB DAC's, but maybe you can audition one of Gordons USB implementations and decide after the trial. Optimising per Cic's paper has been tried with 2 PC's using the same MB and associated parts to my system and I did not detect any difference in SQ. One optimised and the other with carefull driver selection and both with a stable OS loaded.

Just something to ponder.

MAK

[Sasha]

Sasha:

Fair enough on your reply that I re-read the 4 pages and come to the conclusion that the optimized PC and Lynx card did not meet your sonic desires. I use the RME 9652 and prefer to use an external DAC for playback. One reason is the jitter rejection within the ESS Sabre DAC. Others reasons are more obvious; external DAC environment from the PC, better PS and so forth.

Other things I have done for myself are to externally locate the sound card outside the PC with a Adex riser.
http://www.adexelec.com/


Most important for me was not using the DACS on the RME at all and just transfer the signal to the external DAC. The ESS chip has very good jitter rejection per the data sheet and had been confirmed in testing. I believe in engineering the best solution I can afford before assembly, to limit my dissatisfaction. Still, the speakers and room will have the most impact on SQ of all. Unless the DAC is a real piece of crap, the sonic differences I hear are minimal at normal listening levels.

I read that you are not a believer in USB DAC's, but maybe you can audition one of Gordons USB implementations and decide after the trial. Optimising per Cic's paper has been tried with 2 PC's using the same MB and associated parts to my system and I did not detect any difference in SQ. One optimised and the other with carefull driver selection and both with a stable OS loaded.

Just something to ponder.

MAK

OK, now we are having meaningful and very interesting discussion.
First question that pops to my mind is if you ever had a chance to compare RME to Lynx in your PC/system.
I know of one person that did, and actually preferred RME, despite the fact that published specs favor Lynx and that Lynx is more $. This person is as anal as I am, so his word has some weight. This is yet another reason I am skeptical about published specs.

For clarification, I do use an external DAC.
My beef is the amount of jitter that sound card outputs on its digital interface (and I see no other reason for poor performance in comparison to CDP tray but jitter, as you could have seen from my comparison with Wadia player being both the disk player and DAC).

Adex riser is very interesting piece, I have been considering an external PCI bus expansion enclosure, but the problem is that Lynx is specific in requirements and the compatible enclosures such as Magma are very expensive, I was not willing to spend such $ on trial.
Can you tell me more info on Adex piece, what specific one did you get, and what improvements have you noticed?

Have you done any modifications to RME card, things such as improvements in interface impedance matching (I am not happy with those DB connectors on Lynx, no way I can have good impedance whether I select SPDIF or AES/EBU, and there must be lots of interference on such closely located leads)? Or powering RME up from an independent more stable heavily regulated/filtered PSU, since you had RME outside?

Speaking of speakers, I agree mostly but not entirely, I sold my PMC IB2 and right now I am in the process of building a pair myself (presently trying to decide on finish of assembled enclosure before mounting drivers). But good source can have tremendous impact on sound quality.

I am not believer in USB based on my limited exposure to USB, I would love to audition anything, but I am not willing to spend big $, I already spent too much on blind fate, one of reasons why I am such skeptic. And frankly, I am leaning away from sources that employ tubes and transformers in analog section, but that is yet another long subject.
Over the time I switched to manufacturers that have proven record in professional world (thus Bryston, PMC, speakers I am building sport quite similar drivers, ATC dome mid, Volt bass), I found such pieces exceptional.

I do not understand what exactly you mean by ?Optimizing per Cic's paper has been tried with 2 PC's using the same MB and associated parts to my system and I did not detect any difference in SQ.?, can you elaborate? If you optimized both systems, why would you expect to hear the difference?
And what exactly is the difference between ?One optimized and the other with careful driver selection? I am puzzled with this one.

[racerxnet]

Sasha,

I did not compare the Lynx to the RME, but the AudioTrak 7.1, 7.1 Hi Fi, 1212, and a modded AV710 card. Using the sound cards onboard DAC for playback their was a difference in sound from the cheap AV710 to the RME. That's due to the poor construction of the 710 card. Nothing I can do about that.  Not much difference between the Audiotrak 7.1 and RME.

My testing the Audiotrak and RME to the external Tri Vista 21 and Buffalo DAC revealed little if any difference. I do not have golden ears, so mileage may vary.

I do not understand what exactly you mean by ?Optimizing per Cic's paper has been tried with 2 PC's using the same MB and associated parts to my system and I did not detect any difference in SQ.?, can you elaborate? If you optimized both systems, why would you expect to hear the difference?
And what exactly is the difference between ?One optimized and the other with careful driver selection? I am puzzled with this one.

I'll rephrase this. I had 2 computers assembled with the same MB, power supply, sound cards etc. One PC was set up per Cic's recommendation and the other set up as a normal PC feeding the same external DAC. I tested to see if there was any difference in SQ based on the optimizations vs. a normal PC. I could not discern any difference in SQ between the 2 different configurations. I've tried Cic's latest player to my outboard DAC's and quite frankly, I'm not a fan of it. I am not knocking it, but it is not my cup of tea. At this moment I am using Foobar .083, mega nerds src @ 88.2 to the DAC. I have also tried the SOX resampler for .083 but do not remember any differences at the time. This was tested over 2 years prior. I've had my head under the hood until I was satisfied with the sound quality. This works for me, so take it with a grain of salt.

As for the riser card, this is just an extension cable to move the card out of the PC. The box to mount it in can be pine if you choose. Nothing special about machining a case from aluminum, or buying a small box to mount the card. You'll need a pinout of the PCI slot if you want to self power the card. I don't remember the part number, but if you look at the link provided you can see that they sell many different lengths and styles of risers. Many are used in Mil Spec apps where space constraints require placement in oddball locations.

The only mods I did to the RME was to replace the crap dongle with a 75 ohm BNC and cable, as well as BNC on the Tri Vista 21 and Buffallo DAC.

The only other suggestion is to buy used from audiogon and sell if you do not like to SQ of said component. There should be little loss in resale value if you choose carefully. I bought my 4 7BST amps and many other components from there and am happy with the purchases.

My speakers are the RS1 B's and Genesis 200's at this time. So I feel I have no need for other auditions at this time. Also have the Kappa 9's for sale and are no match for either speaker. I have listened to others and walked away unimpressed many times from the showroom floor.

On a last note; the ESS Sabre DAC chip does a commendable job of jitter reduction, you can read about it at the DIY audio link I provided. The RME allows an external clock to slave to. This is my preferred method at this moment. You can use the Superclock, Tent clock, or others which will lower the jitter compared to others as long as the implementation is done correctly.

[Sasha]

MAK,

Thanks for good info, I appreciate it.
Have you noticed any improvement when you moved RME outside of the case on the riser card, and if so, what was the magnitude of improvements, can you try to describe it?
I am wondering if it is worth pursuing or not on its own without playing with power as well.

How did you wire 75 ohm BNC on your card, just an ordinary copper hook-up wire for such short distance, or something else?
What improvements did this bring you?

I see you like arrays. I have not heard yours, but did McIntosh XRT2K. How do they compare if you had a chance to hear them?

[bob stern]

jitter is only introduced into the signal chain during the A-D (recording) and D-A (playback) processes at the hardware level, i.e. during a physical medium transition. No more, no less.

Jitter is not a function of the other factors inherent in the signal chain that could potentially add noise and/or artifacts to the signal, yet we audiophiles have become obsessed with............jitter.

The term "introduced" in the above quote is erroneous; it should be "present".

In other words, it is true that only the jitter "present" at the A-D and D-A stages affects the audio output.  However, jitter "introduced" at every point in the signal chain from the source to the D-A is likely to result in jitter "present" at the D-A stage.

As discussed in another thread on this forum that cited a mathematical paper, sample rate converters and other re-clocking schemes between a DAC's digital input and the D-A stage can greatly attenuate jitter, but they do not completely eliminate it.  The term jitter "rejection" is misleading if it is interpreted as "elimination" rather than "attenuation".
http://www.audiocircle.com/index.php?topic=59224.msg594529#msg594529

Stereophile's John Atkinson describes his jitter measurement methodology here, including excellent measurements of the BCD-1:
http://www.stereophile.com/features/1208jitter/

At the risk of discrediting the goldenness of my ears, my comparison of CD transport versus Mac connected to my BDA-1 is here:
http://www.audiocircle.com/index.php?topic=53416.msg645615#msg645615

[werd]

Hi All,

I have been searching hi and low for any scientific info on the detectability of Jitter in digital music playback and so far other than the Dolby study (see link above) there is not much out there. If anyone knows of any valid research on this subject please post.

The Dolby study claims anything below 4NS is not detectable. That seems a little on the high side to me but Dolby is no feather weight when it comes to research.. Most sources I have spoken too (all though antidotal) seem to feel below 400PS is required for most not to hear the negative affects of Jitter?

Also, there are many ways of measuring jitter. What's the measurement bandwidth? Is the jitter "white" or signal-correlated? Is the measurement RMS or peak? There's no standard spec.
 
james

Hi James

I think what your asking for is unattainable. Its my estimate that all measuring of jitter is purely done by forecast. This means all data is  subjective based. The type of attention needed to measure distance/speed/timing in the pico  range accurately  is likely only found in science workstations such as this http://www.lightsource.ca/education/material_pdf.php. Millions of dollars spent on equip along with attention to
the design of the building are imperative not to skew data. Not normaly found around the hifi community or the people who build and design it.

[racerxnet]

Sasha,

Sorry I did not post sooner, my wife and I took a day to see our son at scouts camp. A 750 mile round trip, but worth it!

Regarding sound differences/descriptions, I am not the best to describe them. I thought the BNC connections added some base depth on songs that had that in the passage. Other songs sounded the same on the top end. It just depended on the particular passage on each song.

I did not compare the sound with the card removed from the PC vs. internally. As we know, audio memory is short lived. At least in my case it is. The assembly for the external card took about 2 days of putzing around. By then I could not give a accurate comparison. Buying an extension from Adex is cheap. You can lay the card on a piece of wood or off to the side to see if you can hear any difference outside the computer. It will still be powered from the PC unless you cut the wires powering the card and use another source.

The cable is only 6ft long soldered together with a new DB9 housing and a BNC connection on the other. You can buy the connectors at DIGI KEY.

I have not heard the Mac stuff. I wish I had more opportunities to listen to other equipment, but I live in the sticks West of Chicago. I get the chance to visit a shop every few months when I am in town. I feel I have no need to look further for new speakers. The Genesis 200 were a good amount of coin and the Infinities are able to hold their own on anything below $25,000.00 IMO.

You might try to load the OS without all the Cic's suggestions as a starting point and the BIOS settings per default, and see what you think. Underclocking and volting adds nothing to the sonics with my set up. I use the Asus P5W-DH MB, Intel 6320 processor, and a good PS for the PC. I tried AMD with the Via chipsets and had some success with Soyo MB's at the time. I will stick with the Intel based MB from now on. ASIO is based on SSE 3 compilations.

[doctorcilantro]

Sorry I have been out of the loop; just finished grad school the other day with an exam.

I'm curious as to the merits of using an external clock.

I was going to try an RME FF400 given my issues with my EMU, but after reading many comparisions of the two units, I'm keeping the 1616M; the converters (I also need decent ADC) are in its breakout box and it is powered by it's own adapter. I'm going to try a "EDI" (which I never realized is not a proprietary cable, just "cat5") shorter cable to the PC and test using an external clock.

Racer, you seem to be one extolling the possible benefits, whereas Sasha is not.

My thoughts on configuring a PC, or how a PC contributes to jitter via software: your test speaks volumes Racer. I can see how running a multi-threaded audio analysis on 4 cores while gaming could cause your system to become "dirty", but with common use and buffers (I don't agree with his assessment on shorter path is better) and a quality audio engine (J. River, Amarra, Foobar, etc.) things will be fine. I see what, 1% cpu usage playing back 24/96 on my 45w AMD cpu on mini-ITX board with a 110w Pico-psu.

Anyway, I'm keen on trying an external clock, thoughts again? Sasha noted on the possibility of making things worse.

DC

[racerxnet]

 Hey Doc,

I think that a better clock might improve things; when everything else is equal (cables, PS etc). I have not had the time to test different clocks from the card vs. the DAC. My thought is that wherever the better clock, your best place to start is with said clock; be it the PC card, DAC or reclocker (digital lens - Big Ben).

Another factor is that I strive for a quality constructed MB to start with. There are to many variables to state this is fact. But, I try to start out with a stable OS and quality components, then I focus on drivers for the RME,  and other chipsets on the PC. My processor is only a dual core, but has 4MG cache on die. I prefer this faster memory on die than 2 Mg cache processors. It seems to me that to many changes are made on the OS and bios per Cic's posts, and maybe people are not familiar enough with what they are doing and actually degrade performance. Start with the basics, because audio playback is trivial on today's PC's and MAC's.