Home of the Squeezebox™ & Transporter® network music players.
Page 30 of 32 FirstFirst ... 202829303132 LastLast
Results 291 to 300 of 318
  1. #291
    Senior Member
    Join Date
    Apr 2005
    Posts
    8,081
    Quote Originally Posted by DanBa View Post
    What is the smaller parameter?
    Thanks.

    Apparently, the Android Linux-powered Samsung Galaxy S3 has the same interworking issue with the CEntrance DACmini:
    http://www.head-fi.org/t/595071/andr...0#post_8827852
    I'm not convinced this is the same thing. The change I made for rDAC is in post 219:
    http://forums.slimdevices.com/showth...l=1#post702222

    This is basically ensuring the interval requested for the syncpipe (the feedback connection) is not too long for the size of the ehci schedule maintained by the linux kernel - essentially it can't schedule periodic intervals which are too long as the size of the schedule held in memory means that scheduling packets into the future goes beyond the size of the schedule held.

  2. #292
    Senior Member
    Join Date
    Aug 2011
    Posts
    144
    I updated touch to 7.8. Installed TT 3.0, installed EDO. I use HRT II+ USB DAC with hub. I am getting clicks and pops. Commented the priority line in rcs.local file as suggested by John Swenson. Still I get the clicks. Any suggestions?.

  3. #293
    Junior Member
    Join Date
    Jan 2012
    Posts
    2
    @ Triode
    Thank you very much!


    By the way, is there any chance that you will tackle Android USB audio in the future?


    As far as I know, there are very few experts working on Android (standard) USB audio for the time being: some Samsung and Archos engineers, likely some engineers from other Android device manufacturers, and likely no third party / open source developer.


    Personally, I think Android USB audio will trigger a new era of Hi-Fi for the masses with affordable rig:
    . affordable (subsidized) smartphone or digital audio source
    . affordable common USB DAC/amp compatible with PC or Mac
    . affordable headphones


    Some Android USB audio rigs:
    stock Samsung Galaxy S3 > 50us$ Stoner Acoustics UD100 > Sennheiser IE80




    stock Samsung Galaxy S3 > HeadAmp Pico > Beyerdynamic T5p




    stock Samsung Galaxy S3 > Epiphany Acoustics EHP-O2D (i.e. ODAC + O2) > Denon AH-D2000




    A list of standard USB DAC reportedly interworking with the Android-powered smartphone Samsung Galaxy S3 / Note2:
    http://www.head-fi.org/t/595071/andr...0#post_8820225

  4. #294
    Senior Member
    Join Date
    Apr 2005
    Posts
    8,081
    Quote Originally Posted by lake_eleven View Post
    I updated touch to 7.8. Installed TT 3.0, installed EDO. I use HRT II+ USB DAC with hub. I am getting clicks and pops. Commented the priority line in rcs.local file as suggested by John Swenson. Still I get the clicks. Any suggestions?.
    Any clicks without TT3.0? EDO is really designed to work without it, so can you confirm this works?

  5. #295
    Senior Member
    Join Date
    Aug 2011
    Posts
    144
    With only EDO there is no clicks or pops. When combining EDO and TT 3.0 is when the clicks come in.

  6. #296
    Senior Member
    Join Date
    Apr 2005
    Posts
    8,081
    Quote Originally Posted by lake_eleven View Post
    With only EDO there is no clicks or pops. When combining EDO and TT 3.0 is when the clicks come in.
    Suggest using with EDO only then

  7. #297

    clicks pops crackling Teac UD-H01

    Hello there,

    here is a genius at work. Thank you for your efforts, triode.

    I┤ve recently acquired a Teac UD-H01 dac, but unfortunately I got no SBT, so I cannot test your kernels numbered 9-12.

    I have massive problems with playing back audiofiles beyond 48kHz. The playback of high-res audio files clicks/crackles/pops.

    I┤ve tried it with 2 newer Ubuntu kernels (for x86 based machines): 3.2.23 and 3.5.0, same behaviour as mentioned above while playing back files with aplay.

    You┤ve stated that your "Tenor" kernels were designed to work around the frequency corruption seen with the Tenor chip. Is this frequency corruption the reason for the the crackling playback or are there maybe any other factors that could disturb the USB audio stream?

  8. #298
    Senior Member
    Join Date
    Apr 2005
    Posts
    8,081
    Quote Originally Posted by jollyjoggler View Post
    Hello there,

    here is a genius at work. Thank you for your efforts, triode.

    I┤ve recently acquired a Teac UD-H01 dac, but unfortunately I got no SBT, so I cannot test your kernels numbered 9-12.

    I have massive problems with playing back audiofiles beyond 48kHz. The playback of high-res audio files clicks/crackles/pops.

    I┤ve tried it with 2 newer Ubuntu kernels (for x86 based machines): 3.2.23 and 3.5.0, same behaviour as mentioned above while playing back files with aplay.

    You┤ve stated that your "Tenor" kernels were designed to work around the frequency corruption seen with the Tenor chip. Is this frequency corruption the reason for the the crackling playback or are there maybe any other factors that could disturb the USB audio stream?
    Well if its the same tenor chip as we were testing, it corrupts the feedback messages so if you look at the momentary frequency in /proc/asound/<card>/stream0 then it will probably look wrong?

  9. #299

    Teac UD-H01 has the TGenor TE8802 chip for USB

    Quote Originally Posted by Triode View Post
    Well if its the same tenor chip as we were testing, it corrupts the feedback messages so if you look at the momentary frequency in /proc/asound/<card>/stream0 then it will probably look wrong?
    The teac UD-H01 has the Tenor TE8802 chipset.

    Have a look here

    http://pastebin.com/hiqpra58

    the frquency is jumping in between 183994 Hz and 199994 Hz while playing back a 192kHz file. So I assume this is the frequency corruption you are talking about.

    Could your patch be applied to other kernels which are not for ARM based machines like the SBT?

    Thank you!

  10. #300
    Senior Member
    Join Date
    Apr 2005
    Posts
    8,081
    Quote Originally Posted by jollyjoggler View Post
    The teac UD-H01 has the Tenor TE8802 chipset.

    Have a look here

    http://pastebin.com/hiqpra58

    the frquency is jumping in between 183994 Hz and 199994 Hz while playing back a 192kHz file. So I assume this is the frequency corruption you are talking about.

    Could your patch be applied to other kernels which are not for ARM based machines like the SBT?

    Thank you!
    Yes this looks similar. I did email Tenor, but they did not appear keen on fixing Linux for some unknown reason.. Anyway as they've not responded for a while, I thought I'm post my observations and the hack thats in kernel 11:

    Here's some logging we did with a Tenor device recording the feedback frequencies:
    Code:
    Example at 44.1kHz sample rate
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058340 freq: 44102 Hz set
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    feedback raw: 00058340 freq: 44102 Hz set
    feedback raw: 00048330 freq: 36100 Hz ignored<---- this looks wrong raw value should be 00058330
    feedback raw: 00048330 freq: 36100 Hz ignored<---- this looks wrong raw value should be 00058330
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058330 freq: 44100 Hz set
    feedback raw: 00058340 freq: 44102 Hz set
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    feedback raw: 00068330 freq: 52100 Hz set <---- this looks wrong raw value should be 00058330
    Example at 48kHz
    feedback raw: 00050000 freq: 40000 Hz ignored<---- this looks wrong raw value should be 00060000
    feedback raw: 00050000 freq: 40000 Hz ignored<---- this looks wrong raw value should be 00060000
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00070010 freq: 56002 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00070010 freq: 56002 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00070010 freq: 56002 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00070010 freq: 56002 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060010 freq: 48002 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00070000 freq: 56000 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00070000 freq: 56000 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00070000 freq: 56000 Hz set <---- this looks wrong raw value should be 00060000
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00050010 freq: 40002 Hz ignored<---- this looks wrong raw value should be 00060000
    feedback raw: 00050010 freq: 40002 Hz ignored<---- this looks wrong raw value should be 00060000
    feedback raw: 00060000 freq: 48000 Hz set
    feedback raw: 00060000 freq: 48000 Hz set
    Example at 96 KHz
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0010 freq: 96002 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000c0020 freq: 96004 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    feedback raw: 000d0020 freq: 104004 Hz set
    Essentially this mean that the 3rd byte of the feadback value is getting corrupted and probably changing when it should not.

    I suspect this can be made more general, but this is essentially what kernel 11 is doing - it reads the feadback frequency in /sound/usb/urb.c retire_playback_sync_urb and then modifies it to try to get around the corruption after f is read:
    Code:
    if (IS_TENOR) {
     
      // automatic frequency detection can get this wrong at start of a new sample rate ľ so force to 0 which is the value for uac2
      subs->freqshift = 0  
    
      // Tenor 8802 corrupts high order feedback bytes ľ assume only low order 2 bytes change and use nominal frequency for high order bytes
      f &= 0x0000ffff;
      f |= (subs->freqn & 0xffff0000);
      if (unlikely(f > subs->freqn && f - subs->freqn > 0x8000)) {
       f -= 0x10000;
      }
    
    }

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •