What happens to an audio signal from media player to dac - Page 2 - diyAudio
Go Back   Home > Forums > Source & Line > Digital Line Level

Digital Line Level DACs, Digital Crossovers, Equalizers, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 10th January 2013, 03:02 PM   #11
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
thats alright, we are running a system here on the forum that is outputting ~1.5ps buffered i2s. measured, proven. well actually Ian had to cripple the unit with a higher jitter generic clock to get a measurable result of 5ps, because his scope has a 2ps noise floor. using the intended 0.5ps clocks it works out to 1.5ps period jitter including the small amount from the flip flops.

so if your one gets people riled up... =) i'm readying for war lol.

if the peachtree uses a similar memory buffer type reclocker, actually your efforts, though understandable, would be in vain. a large amount of jitter may perhaps cause an overrun on the memory. probably it has an isochronous connection to the micro doing the conversion to i2s, but the output of the memory is asynchronous to that input (offset by the time it takes to half fill the fifo buffer) but if it doesnt overrun, I cannot think of any mechanism in the PC that would improve on it.

the measurements of this system were taken with 150ps jitter on the input signal and its the same whether optical, electrical spdif or i2s input. a complete leveller, the dac becomes completely transport agnostic.
  Reply With Quote
Old 10th January 2013, 03:58 PM   #12
n0vtz is offline n0vtz  United States
diyAudio Member
 
n0vtz's Avatar
 
Join Date: Aug 2012
Location: Iowa
qusp;

Well, from now on when they give me headaches about 3 ps at the clock in the Dacit, I'm going to just send them your way, hehe. :0)

I might have to find out more about that Dac. The fact that coax or toslink are giving the same readings is interesting.

God forbid, ps, I wish I could measure at those levels. I have a heck of a time just measuring uv's for the JLH filter. I use VA Analyzer (free software), but trying to calibrate and verify a reference voltage is pretty iffy. My scope goes to 50 mv increments, so uv measurements are; how thick is the trace line. :0)

Thanks for the explanation on the Peachtree and my quest for better clocking. It seems anyone that knows what they're talking about says the same thing. I'm going to have to think through what you said.

The system here now works so well, but you know how it is, we just need to find something to screw with...

Sincerely,

Jim
  Reply With Quote
Old 10th January 2013, 04:32 PM   #13
diyAudio Member
 
Join Date: Mar 2004
Location: At the output stage
Send a message via Yahoo to mr_push_pull
Quote:
Originally Posted by n0vtz View Post
Most of your assumptions are incorrect, but that will happen when you do a lot of guessing. :0)
like I said, my wasted time.

Quote:
Originally Posted by n0vtz View Post
Reason for questions is to get a better understanding of of the internal dymanics and movement of the audio signal path in the pc. Questions are I believe pretty indicative of that.
I'm not sure exactly how it helps. after reading the rest of the replies I'm still not sure what you're getting at.

Quote:
Originally Posted by n0vtz View Post
Clocking as I assume you know has to do with on board XTAL's, and PCM data is "clocked" in the I/O hub controller (ICS?). Your idea of clocking is actually a function of the specs of transistor and it's biasing characteristics.
it has to do with board XTALs, board PLLs and many other things. if it's jitter you're getting at, there are even more factors involved (noise that gets into the circuits involved in many ways, power supply noise, the quality of all the circuits involved and even mechanical vibration which the XTALs are susceptible to etc). I believe the extent of control you have over what's going on inside your PC wrt jitter is very limited but I may be wrong. unfortunately jitter reduction is not a matter of just clocking in the right place. just search for some technical articles on PLL implementation details and unless you're very inclined or an EE you'll likely give up pretty soon as its not exactly science for the masses.

Quote:
Originally Posted by n0vtz View Post
I'd be interested in how your verifying bit perfection, and how your measuring jitter at your analog output.
you have the option to just trust what others have done, for instance Benchmark Media: Main Page - Benchmark
I remember reading there that they test the SW players for bit "perfectness".
there's a setup page for J-River: J River for Windows - Setup Guide - Benchmark maybe they say something about it there.
second option is to do my test. and to make things easier for you, I'm attaching two 16/44.1 files containing a 440Hz sine, one of which has a sample altered (see pic). give them a listen and see if you can detect the clicking sound. ok, maybe you'll say that you listen to music, not sines but whatever artifact of the kind that is audible with music will be even more audible with sines (I assume that you know how low amounts of harmonic distortions are very audible with sines but not so obvious with real music). and unless for some reason it's just one of the least significant bits in the data that is corrupted and it only happens in the most "convenient" moments, you'll hear it. just try it, it takes a few seconds to download and unpack the files.

USB analyzers are expensive, there is no other cheap and quick way to test bit accuracy for yourself.

as for jitter, it's more complicated. search for J-Test and Julian Dunn. Julian Dunn (RIP) was the first to devise the jitter test signal as it is used now in measurement procedures (the plots you see in Stereophile et al. are derived from a digital J-Test signal fed to the D/A converter). you will need rather expensive equipment to do that sort of testing so that you can be sure you're not really measuring its limitations. a very good pro soundcard may do the trick but the cost is again significant.


I'm sorry if I'm stating the obvious and/or it's all old news to you, as you can see I have poor mind-reading abilities
Attached Images
File Type: gif altered.gif (26.8 KB, 105 views)
Attached Files
File Type: zip pack.zip (574.3 KB, 8 views)
__________________
we all love a good ol' stereotype until it's against us
  Reply With Quote
Old 10th January 2013, 04:48 PM   #14
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
its not a dac, its a dedicated i2s Fifo Memory Buffer has modules for spdif input a couple different clock modules, a galvanic isolation module to wall off the PC ground and a PCM adapter daughterboard to allow it to work with different dacs that dont take 32bit PCM, or require special clocking tricks like TDA1541, PCM1704 etc etc.

its a pretty long thread, there is a wiki with a compiled list of important links in the development thread as well as detailing the different modules. its only new so not exhaustive yet.

Its a very well executed build, one of if not the best conceived digital modules on this forum IMO. its not very happy news for some people that just love tweaking their PCs and have invested a great deal of money and time, but that was the point and only recently (a few days ago) did he go through with the in depth measurements, because as you mention its not an easy task and we all were very confident in the result.

all the same to see the objective data to confirm it was a pretty happy moment.

that being said, if the numbers they are quoting are real and you are happy with the way the dac itself performs, I reckon just move on and find something else to tweak, (speakers...) the difference between 3ps and 1.5ps is exceedingly unlikely to be audible. is that 3ps the output of the clock, or where are they measuring that? are they still using ES9018 in that like the Nova?

the fifos power is that it means we can get this same performance with our dacs of choice and which ever type of input we like, but a well executed integrated dac can work just as well. suits me better as I tweak constantly. plus it will handle 44.1->384kHz in synchronous mode with my ESS dacs and it has proper impedance controlled layout and impedance controlled connections on u.fl micro BNC

people will start to think i'm a salesman for Ian hehe, but the truth is, not only would there be little margin for that (the pricing is almost too good) but i'm simply just happy to try and curtail the tail chasing involved in fighting the wars inside the computer when even a perfectly conceived computer transport still leaves issues that need solving; best just tackle it at the dac where its easier to control/ modify. plus with these latest measurements I can be more confident in recommending it, because i'm not just recommending on design and faith in that design functioning as intended.

Last edited by qusp; 10th January 2013 at 04:53 PM.
  Reply With Quote
Old 10th January 2013, 05:25 PM   #15
Julf is offline Julf  Europe
diyAudio Member
 
Join Date: Oct 2011
Location: Amsterdam, The Netherlands
Quote:
Originally Posted by n0vtz View Post
I have been trying to figure out from JRiver via USB to a Peachtree Dacit what is happening to the signal precisely and the places and stages of processes it goes through as it moves through the computer and finally to the Dacit.
Jim,

I guess you will also get a pretty good comparison of the expertise levels and helpfulness of various computer audio forums
  Reply With Quote
Old 10th January 2013, 05:45 PM   #16
diyAudio Member
 
Join Date: Mar 2004
Location: At the output stage
Send a message via Yahoo to mr_push_pull
Quote:
Originally Posted by Julf View Post
Jim,

I guess you will also get a pretty good comparison of the expertise levels and helpfulness of various computer audio forums
I happen to run into CA pages when googling for various stuff. I generally end up closing the tab on my browser after reading no more than 4-5 posts a great, very informative website, in no way full of misleading or downright false info. oh, and obviously in no way plagued by marketing.
__________________
we all love a good ol' stereotype until it's against us
  Reply With Quote
Old 10th January 2013, 05:51 PM   #17
n0vtz is offline n0vtz  United States
diyAudio Member
 
n0vtz's Avatar
 
Join Date: Aug 2012
Location: Iowa
mr_push_pull;

Thanks for the response. Yep, 2nd file you sent, about 3 seconds in. In your opinion, what does your test and the fact that I detected that click mean in terms of jitter and the system here?

Your correct about inner pc exercises toward perfection. I have isolated with an external ps and a JLH filter the 5 v. (<100 uv noise and ripple) the usb implementation to the Dacit here, but the serial data is another story. As you say, there are EMR and RFI contaminates in this signal and there is little I can do about it, except rely on fifo and clocking in the Dac to correct it and it's potential effects.

If you get a chance check out VA Analyzer, a free software that I use in addition to my old fashioned scope, you may find it useful for attempting some low voltage measurements. The generator and spectrum analyzer are quite good as well.

Thanks again for the files, hope to hear from you about what this test indicates.

Jim
  Reply With Quote
Old 10th January 2013, 06:01 PM   #18
diyAudio Member
 
Join Date: Mar 2004
Location: At the output stage
Send a message via Yahoo to mr_push_pull
it doesn't tell one thing about jitter but it pretty much tells everything about bit accuracy. you mentioned bit accuracy in reply #6.
the idea is that any corrupted bit in the data stream would render even more destructive (audible) results, compared to what you can see (and hear) in the attached files. one bit that gets changed from 1 to 0 or vice-versa can give an error close to 100% as long as it's in the most significant position within a word. and also remember that it's precisely one bit changed out of 44100 words / second that you heard.
maybe I'm preaching to the choir but fact is that there are audiophiles who are convinced that with lower quality (read cheap) D/A converters there is data corruption going on.

jitter is either evaluated by listening or objectively with rather expensive equipment, like I said I'm not aware of any cheap and reliable way to assess the jitter amount.

thanks for taking the time to give the files a listen, you are correct about the position of the altered sample.
__________________
we all love a good ol' stereotype until it's against us
  Reply With Quote
Old 10th January 2013, 06:04 PM   #19
n0vtz is offline n0vtz  United States
diyAudio Member
 
n0vtz's Avatar
 
Join Date: Aug 2012
Location: Iowa
Julf;

LOL, good to run into you. BTW, your post at CA was very helpful. Had a computer/electronics repair and build shop (come get a dmitri computer!) a long time ago. Boy, I'm not "with it " anymore. Have to do some serious homework. Your post was very helpful, as some here are. Hope to keep running into you. :0)

Jim

ps If you need any parts or stuff from the usa, let me know.
  Reply With Quote
Old 10th January 2013, 06:42 PM   #20
Julf is offline Julf  Europe
diyAudio Member
 
Join Date: Oct 2011
Location: Amsterdam, The Netherlands
Quote:
Originally Posted by mr_push_pull View Post
I happen to run into CA pages when googling for various stuff. I generally end up closing the tab on my browser after reading no more than 4-5 posts a great, very informative website, in no way full of misleading or downright false info. oh, and obviously in no way plagued by marketing.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Digital Audio Media Player Choices Loren42 Digital Source 0 23rd May 2012 01:28 AM
Compact disc player sonic differences by various signal procesing devices before DAC tiefbassuebertr Digital Source 65 30th January 2012 05:18 PM
Multiple instances of media player to different audio outputs narituba PC Based 8 24th March 2010 10:50 PM
How can I convert SPDIF to EFM (RF) signal, to use a cd player as a DAC ? gaetan8888 Digital Line Level 10 8th February 2009 07:32 AM
Media player click with USB audio jmar Digital Source 1 5th February 2006 04:24 PM


New To Site? Need Help?

All times are GMT. The time now is 10:34 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2