Go Back   Home > Forums > Source & Line > PC Based

PC Based Computer music servers, crossovers, and equalization

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 25th October 2011, 02:28 PM   #291
diyAudio Member
 
soundcheck's Avatar
 
Join Date: Mar 2005
Location: DUS
Quote:
Originally Posted by rrrremus View Post
What is the difference between manually decoding a flac to wav and a player application applying the exact same decoding algorithm ?

Most players i use run in high priority anyways.
The load during realtime decoding of flacs is causing extra load. Not much, but sufficient to cause more physical distortions.
Somehow this is causing more jitter/timing variations resp. extra common-mode noise.

It's important where the decoding takes place though.

On a Squeezebox system it is sufficient to let the server decode the flacs and to stream PCM data to the client which feeds the DAC. Decoding processes are very well isolated
due to the server/client structure.

On PC based transports with local playback you decode the flac prior to playback.

Even if phofman won't believe it an never did, I reported these effects years ago.


Nowadays all kind of so called audiophile players, decode flacs prior to playback and do ramplayback to name a 2nd very important feature.

If audio interfaces would be able to cope with those little distortions we wouldn't have that discussion and all these players as well as my SB modifications would be obsolete. Unfortunately I'm not aware of any device which can cope with those distortions. That'll
make every PC based transport sound different on the very same DAC device.

Enjoy.

Cheers
__________________
::: Squeezebox Touch Toolbox and more ::: by soundcheck
  Reply With Quote
Old 25th October 2011, 07:39 PM   #292
phofman is offline phofman  Czech Republic
diyAudio Member
 
Join Date: Apr 2005
Location: Pilsen
Quote:
Originally Posted by soundcheck View Post
The load during realtime decoding of flacs is causing extra load. Not much, but sufficient to cause more physical distortions.
Somehow this is causing more jitter/timing variations resp. extra common-mode noise.
...
Even if phofman won't believe it an never did, I reported these effects years ago.
Well, you argue that the extra CPU load induced by FLAC decoding will cause "physical distortions", causing deterioration of the sound quality, while at the same time you argue the extra CPU load induced by your recommended playback buffer minimization to reduce playback latency (more frequent IRQ servicing, more frequent data transfers, more frequent player wakeups), and the extra CPU load induced by inserting additional item into your "highly recommended" playback pipeline - ecasound, yield improvement of the sound quality.

And these obviously contradicting arguments are based on what you and reportedly a few others have heard. No proper statistically objective blind listening tests were conducted, at least not a single one was reported. In a professional community this kind of subjective statement would not ever be taken seriously. But I understand unsupported claims are good enough for some audiophiles who just live to believe...

Last edited by phofman; 25th October 2011 at 07:43 PM.
  Reply With Quote
Old 26th October 2011, 09:37 AM   #293
diyAudio Member
 
Join Date: Sep 2009
Location: Timisoara
I posted in this thread before and claimed that i could not distinguish pureplayer from foobar. However, a problem i encounter these days with Windows OS made me think again about PC playback and i will post my problem here in order to find a solution and also to better understand what is happening with audio under Windows

My setup is the following one :

M-Audio Delta 192 PCI sound card
3.5 meter coaxial cable running from sound card to DAC.
DIY DAC based on SRC4382 and PCM1792A

No matter what player i use ( or sound generating program, could be Firefox for example ) it sometimes happens that the music is distorted beyond clearly distinguishing the original sound.
To better describe the behavior imagine a playlist loaded in foobar and clicking next button... everything works as supposed to until one track will be distorted in the way described above... after clicking next again or replaying the same track, the playback will resume under normal conditions. This only happens i would say once at a hundred tracks or so and i believe it to be a software problem that appears only under Windows XP. Somehow the data sent to the sound card is scrambled, or maybe the soundcard scrambles the data.

Now, i do not want to post offtopic here but this makes me believe that there are more important factors at play that affect the audio data than the software player, CPU load etc.

How can i further investigate this problem ?
My intuition tells me that this could be caused by a bad timing somewhere on the way. But if so, pureplayer or any player still do not resolve timing problems.
Joking a bit : Or maybe these are exactly the "physical distortions" discussed before.
  Reply With Quote
Old 26th October 2011, 10:10 AM   #294
phofman is offline phofman  Czech Republic
diyAudio Member
 
Join Date: Apr 2005
Location: Pilsen
Quote:
Originally Posted by rrrremus View Post
M-Audio Delta 192 PCI sound card
The card is based on Envy24 which does no DSP, no data transformations.

Quote:
Originally Posted by rrrremus View Post
No matter what player i use ( or sound generating program, could be Firefox for example ) it sometimes happens that the music is distorted beyond clearly distinguishing the original sound.
This sounds like overruns issue - the soundcard was starved of fresh data. It can happen when your system gets hogged by another driver.

Did you try increasing the DMA buffer size on the soundcard's control panel?

The weird thing is that you can fix it by skipping to next track. But who knows what the windows driver does when opening/closing the soundcard.

Perhaps we should move this discussion to a new thread

Last edited by phofman; 26th October 2011 at 10:15 AM.
  Reply With Quote
Old 26th October 2011, 10:18 AM   #295
diyAudio Member
 
Join Date: Sep 2009
Location: Timisoara
Quote:
Originally Posted by phofman View Post
The card is based on Envy24 which does no DSP, no data transformations.
I know there is no DSP but there is the option to re-sample the data.

Quote:
Originally Posted by phofman View Post
This sounds like overruns issue - the soundcard was starved of fresh data. It can happen when your system gets hogged by another driver.

Did you try increasing the DMA buffer size on the soundcard's control panel?

The weird thing is that you can fix it by skipping to next track. But who knows what the windows driver does.

Perhaps we should move this discussion to a new thread
when opening/closing the soundcard.
I tried to modify the buffer size with no effect. The only effect i can clearly tell is when playing 24bit / 192Khz with a small buffer there are many drop-outs present related of course to CPU load.
  Reply With Quote
Old 26th October 2011, 10:34 AM   #296
phofman is offline phofman  Czech Republic
diyAudio Member
 
Join Date: Apr 2005
Location: Pilsen
Quote:
Originally Posted by rrrremus View Post
I know there is no DSP but there is the option to re-sample the data.
The card itself cannot do any resampling. I did find such option in the description of the control panel either http://www.m-audio.com/images/global...192_UG_EN1.pdf . What option do you mean?
  Reply With Quote
Old 26th October 2011, 10:40 AM   #297
diyAudio Member
 
Join Date: Sep 2009
Location: Timisoara
Quote:
Originally Posted by phofman View Post
The card itself cannot do any resampling. I did find such option in the description of the control panel either http://www.m-audio.com/images/global...192_UG_EN1.pdf . What option do you mean?
I have the updated drivers with the new control panel.
Under hardware settings you can select codec sample rate. Under normal operation there is no re-sampling being done, but if you manually select a different sample rate, the card will re-sample to the desired rate.
  Reply With Quote
Old 26th October 2011, 01:51 PM   #298
phofman is offline phofman  Czech Republic
diyAudio Member
 
Join Date: Apr 2005
Location: Pilsen
Quote:
Originally Posted by rrrremus View Post
I have the updated drivers with the new control panel.
Under hardware settings you can select codec sample rate. Under normal operation there is no re-sampling being done, but if you manually select a different sample rate, the card will re-sample to the desired rate.
Are you really sure the resampling occurs? The card itself cannot resample. It could be the driver, but I would be surprised if the windows driver developer implemented the resampling functionality. It is not generally the case for Envy24-based cards, at least those I have seen. Of course resampling can occur in the upper layers of windows audio stack, but I assume you are using ASIO or equivalent interfaces which should not resample on their own.

The sample rate switch on the control panel is just for switching the card clock. I do not know the latest drivers, but the functionality described in the user manual does not resample, it is just default samplerate the card runs on when no music is being output. It should switch automatically to the samplerate of the track, unless the "lock the rate" option is checked. In such case it should throw error and refuse playback since the only other way to satisfy this combination would be to resample. Which is what I would guess does not happen.
  Reply With Quote
Old 26th October 2011, 05:03 PM   #299
diyAudio Member
 
Join Date: Sep 2009
Location: Timisoara
Quote:
Originally Posted by phofman View Post
Are you really sure the resampling occurs? The card itself cannot resample. It could be the driver, but I would be surprised if the windows driver developer implemented the resampling functionality. It is not generally the case for Envy24-based cards, at least those I have seen. Of course resampling can occur in the upper layers of windows audio stack, but I assume you are using ASIO or equivalent interfaces which should not resample on their own.

The sample rate switch on the control panel is just for switching the card clock. I do not know the latest drivers, but the functionality described in the user manual does not resample, it is just default samplerate the card runs on when no music is being output. It should switch automatically to the samplerate of the track, unless the "lock the rate" option is checked. In such case it should throw error and refuse playback since the only other way to satisfy this combination would be to resample. Which is what I would guess does not happen.
No, the resampling does not occur under normal operation. However as you also said and also as written in the operating instructions the card is able to re-sample during playback. I just noted that re-sampling is possible ( and with the new drivers i am able to down-sample during music playback right down to 16Khz ) and do not think this has anything to do with the problem discussed. It would be very interesting to find out what is really happening and how the card is initialized for playback.
  Reply With Quote
Old 26th October 2011, 07:48 PM   #300
phofman is offline phofman  Czech Republic
diyAudio Member
 
Join Date: Apr 2005
Location: Pilsen
Quote:
Originally Posted by rrrremus View Post
No, the resampling does not occur under normal operation. However as you also said and also as written in the operating instructions the card is able to re-sample during playback. I just noted that re-sampling is possible ( and with the new drivers i am able to down-sample during music playback right down to 16Khz )
I said the card can change samplerate - the frequency of its clock, the speed at which it operates. That is what could be written in the operating instructions. Is that what you call resampling?

I am afraid there are no decent debugging facilities for this soundcard in windows. If you experienced this problem in linux, we would have numerous powerful tools available, such as the capability to tap the stream at any point between the playback application and the actual driver, listing of immediate status of all Envy and codec registers, detailed statistics of eventual underruns, and lastly the source code of the driver, allowing us to easily make any modifications we would deem useful (e.g. printing debug messages at any place of the driver source code). But your OS does not provide you with such luxury.

At least you could inspect your kernel drivers latency with the latency checker and try to find any correlation with your problem.
  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
Pure digital crossover cka3o4nuk Digital Line Level 14 11th September 2010 11:08 PM
Pure CD Player and DVD Transport Mechanism inside - what about this ?? tiefbassuebertr Digital Source 2 23rd February 2010 08:36 PM
Eminence Pure rossco Car Audio 2 28th March 2004 10:40 PM
Pure Delight Alain Dupont Pass Labs 107 8th February 2004 02:48 PM


New To Site? Need Help?

All times are GMT. The time now is 05:56 AM.


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