Home of the Squeezebox™ & Transporter® network music players.
Page 6 of 7 FirstFirst ... 4567 LastLast
Results 51 to 60 of 66
  1. #51
    Senior Member
    Join Date
    Jun 2006
    Posts
    1,690
    Quote Originally Posted by marcoc1712 View Post
    John,


    could I ask You witch USB DAC are you using? If is kind of a KIt or a commercial product, could you point me to the source?

    is a kind of the one you'll be using in CSP1?

    thx a lot.

    p.s.

    I could not understand why if you upsample at i.e. 352.8 you still need a software filter, is not assured this way that aliasing is far away from the audible spectrum (reason why the hardware filter is disabled, I suppose)?


    Marco.
    The DAC I am using is a custom design I have done so you can't get it anywhere. Speaking of kits, I am designing a somewhat similar DAC for Bottlehead, it will be a kit, but all the hard work is already done on one PCB. It's not available yet, but hopefully will be out sometime early next year.

    Or just get a CSP!

    As to upsampling, all (usefull) upsampling has to use a filter, that is just part of the upsampling process. If you just resample the data without applying the filter you haven't actually done anything. This is best with a picture, I don't have one handy (I'm on vacation), maybe someone else can post a picture of this.

    For the wordy explanation, lets take the output of a basic DAC chip, without any digital or analog filtering you get a stairstep. Upsampling without the filter is just subdividing each of those stairsteps into smaller pieces, but keeping the same values. Lets say you have a sample at .5V level, 8X upsampling gives you 8 shorter samples all at the same .5V, you haven't changed anything.

    What you want is the upsampler to make a "guess" at filling in the intermediate values between the .5V sample and the next sample. The filter defines exactly how that is done. The human hearing system seems to be quite sensitive to exactly what that filter does, slight differences in that filter can make significant differences in what we hear.

    All modern DAC chips do this filtering process builtin to the chip, and every one of them does it in such a way that does not sound so good. This thread is all about doing that filtering externally in order to bypass the not so good version in the DAC chip.

    John S.

  2. #52
    Senior Member
    Join Date
    Mar 2007
    Location
    UK
    Posts
    1,304
    Quote Originally Posted by JohnSwenson View Post
    What you want is the upsampler to make a "guess" at filling in the intermediate values between the .5V sample and the next sample.
    John,
    I've no doubt you know more about the subject than me, and that you meant something else, but I think your language, namely the appearance of "want" and "guess" in the same sentence, might fan prejudices about digital audio!

    Do you mean DAC designers use cheap short-cuts that aren't in fact even trying for the correct sampling-theory ideal? Perhaps you're suggesting your way is actually a theoretically more correct way. I am not here arguing with these assertions.

    All I'm worried about is the wording of the above sentence, since guessing is not something we want, if we are trying to implement sampling theory faithfully (I am sure you agree!) because the theory says a unique, completely correct and completely smooth analogue band-limited waveform is implied by the original samples.

    Regards, Darren
    Last edited by darrenyeats; 2013-11-02 at 07:35.
    Check it, add to it! http://www.dr.loudness-war.info/

    SB Touch

  3. #53
    Senior Member
    Join Date
    Nov 2009
    Location
    DUS
    Posts
    879
    Quote Originally Posted by bennyboyph View Post
    Cheers - it doesn't want to work at 352.8 or 384kHz, which is what I need to bypass the digital filters of my PCM5102 chip :-(
    OK. Sox maxrate is 192khz afaik. And your Squeezebox environment (server and client) would have to support 384khz too.

  4. #54
    Junior Member
    Join Date
    May 2013
    Posts
    7
    Hi John,

    I've been following all this and understand where you're coming from. But what is a CSP?
    Murray (N.Z.)

  5. #55
    Senior Member ralphy's Avatar
    Join Date
    Jan 2006
    Location
    Canada
    Posts
    2,471
    Quote Originally Posted by murrays View Post
    Hi John,

    I've been following all this and understand where you're coming from. But what is a CSP?
    Community Squeeze Project
    Ralphy

    1-Touch, 5-Classics, 3-Booms, 1-UE Radio
    Squeezebox client builds donations always appreciated.

  6. #56
    Junior Member
    Join Date
    May 2013
    Posts
    7
    Quote Originally Posted by ralphy View Post
    Ah, thanks. That's something interesting to look into. Strange it didn't come up in the search.
    Murray (N.Z.)

  7. #57
    Not really. Nothing comes up in the search, in the majority of searches that I've conducted.

  8. #58
    Junior Member
    Join Date
    Dec 2015
    Posts
    12

    Further understanding of upsampling options, esp. in piCorePlayer

    I know this thread died out a long time ago but I think that there are still many of us who would like a deeper understanding of the upsampling options offered in piCorePlayer and I assume other linux-based players. My linux skills are close to zero. My hope is to help others get the most from the options at hand.

    If there is a better thread, forum or site to follow for this please let me know.

    I have found these two links that have some detailed information:

    http://manpages.ubuntu.com/manpages/...ezelite.1.html
    http://manpages.ubuntu.com/manpages/...an1/sox.1.html

    Two specific questions, both related to John Swenson's recipe of "mI:::28" (post #5)
    1) If I leave off the "28" what happens? What's the default? IIRC piCorePlayer uses 32 bit processing - If we specify the 28, does that actually reduce the processing from 32 to 28 bit?

    2) If I want to use a filter option different from M(inimum), I(ntermediate) or L(inear), (lets say "17" on the scale from 0 to 100) do you leave out the M, I or L from the recipe and use "m:::28:::17" or mI:::28:::17" - Does the "17" supercede the "I" ?

    Thanks for any advice

    -Mike

  9. #59
    Member
    Join Date
    Mar 2015
    Location
    Frankfurt
    Posts
    86

    Upsampling Rules

    Quote Originally Posted by JohnSwenson View Post
    ....
    The current implementation in Squeezelite does upsampling to the highest interger rate your DAC cupports. Thus if your DAC's maximum rate is 192, it will upsample to 44.1 to 176.4. If your DAC does not support 176.4, you can use the -r option (or max sample rate in the gui) to set the max rate to 96, then squeezlite will upsample to 88.2. Even upsampling to 88.2 will usually make a significant improvement. There has been talk about giving the upsampling more flexibility so you could choose 192 in this case.
    ....
    John S.
    I found here some descriptions on upsampling rules using Squeezelite.
    However, there is one thing I am not able to get to work as described above.
    A rule that makes Squeezelite to multiply in full numbers, only.

    44.1 -> 176.4
    96 -> 192
    192 -> 192 (no conversion)

    My DAC is able to handle up to 192 and I have Squeezelite and LMS on pCB. - great solution

    My current settings:
    max sample rate: 176.4
    upsample setting: vX::3:28:70:110:50
    Result: Everything is upsampled to 176.4 by Squeezelite & 192 files are downsampled by LMS

    When I change max sample rate to 192 or blank everything is upsampled to 192.


    Is Squeezlite (or LMS) able to upsample to the highest interger rate, only?
    Living Room: piCorePlayer 5.0.0 on rPi 3B+ & Allo DigiOne & 1TB USB (LMS 8.0.0 & Squeezelite) & Keces DC-116, Mutec MC-3+ Smart Clock, Rega DAC R, Rotel RB-1070, Rotel RC-1070 or Fezz Titania Signature, Klipsch Forte III
    Bedroom: Boom / Workspace: Boom / Kitchen: Radio

  10. #60
    Senior Member
    Join Date
    Dec 2009
    Location
    Albinea (RE) Italy
    Posts
    809
    Quote Originally Posted by schiff1108 View Post
    I found here some descriptions on upsampling rules using Squeezelite.
    However, there is one thing I am not able to get to work as described above.
    A rule that makes Squeezelite to multiply in full numbers, only.

    44.1 -> 176.4
    96 -> 192
    192 -> 192 (no conversion)

    My DAC is able to handle up to 192 and I have Squeezelite and LMS on pCB. - great solution

    My current settings:
    max sample rate: 176.4
    upsample setting: vX::3:28:70:110:50
    Result: Everything is upsampled to 176.4 by Squeezelite & 192 files are downsampled by LMS

    When I change max sample rate to 192 or blank everything is upsampled to 192.


    Is Squeezlite (or LMS) able to upsample to the highest interger rate, only?
    No, you need to investigate the input file (or stream) to determinate the sample rate. I implemented this in C-3PO plugin, you could have a look at it, is in the third party plugin lis in LMS.
    __________________________________________________ ______________________
    Author of C-3PO plugin, Squeezelite-R2, Falcon Web interface - See www.marcoc1712.it

Posting Permissions

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