Re: [Musicpd-dev-team] Current mpd-git DSD playback problems/slowdown - possible clues
On 2012/04/04 21:50, Jurgen Kramer wrote: > Yes, that is correct. I was referring to the sample size not the enum values > for the different types of formats. OK, then it might be fallout from another bug, because setting a sample size of 4 bytes is wrong. DSD has 8 bit-samples per channel in each "frame", and since MPD does not know about bit samples, it considers each octet as one sample, i.e. sample size must be 1 (byte). Max -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev ___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] Current mpd-git DSD playback problems/slowdown - possible clues
On 4 apr. 2012, at 21:31, Jeffrey Middleton wrote: > Maybe Jurgen meant the value returned by sample_format_size(), which is > currently (on master) 1 for SAMPLE_FORMAT_DSD? The numeric value of the enum > is not explicitly set, and I doubt it's 1. Yes, that is correct. I was referring to the sample size not the enum values for the different types of formats. > > On Wed, Apr 4, 2012 at 11:18 AM, Max Kellermann wrote: > On 2012/04/02 18:23, Jurgen Kramer wrote: > > After much digging around I managed to get sound working but the timing > > issue remained. When the value for SAMPLE_FORMAT_DSD is changed from 1 > > to 4 (src/audio_format.h) I get proper sound when playing DSD-over-USB. > > This ratio of 1 to 4 makes sense but unfortunately is not the (complete) > > solution. For example with SAMPLE_FORMAT_DSD set to 4, DSD2PCM still > > works fine but now it's timing is off. > > No, I can't imagine how changing the (meaningless) numeric values of > symbolic enums can make a difference for anything. > > Max > > -- > Better than sec? Nothing is better than sec when it comes to > monitoring Big Data applications. Try Boundary one-second > resolution app monitoring today. Free. > http://p.sf.net/sfu/Boundary-dev2dev > ___ > Musicpd-dev-team mailing list > Musicpd-dev-team@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team > -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] Current mpd-git DSD playback problems/slowdown - possible clues
Maybe Jurgen meant the value returned by sample_format_size(), which is currently (on master) 1 for SAMPLE_FORMAT_DSD? The numeric value of the enum is not explicitly set, and I doubt it's 1. On Wed, Apr 4, 2012 at 11:18 AM, Max Kellermann wrote: > On 2012/04/02 18:23, Jurgen Kramer wrote: > > After much digging around I managed to get sound working but the timing > > issue remained. When the value for SAMPLE_FORMAT_DSD is changed from 1 > > to 4 (src/audio_format.h) I get proper sound when playing DSD-over-USB. > > This ratio of 1 to 4 makes sense but unfortunately is not the (complete) > > solution. For example with SAMPLE_FORMAT_DSD set to 4, DSD2PCM still > > works fine but now it's timing is off. > > No, I can't imagine how changing the (meaningless) numeric values of > symbolic enums can make a difference for anything. > > Max > > > -- > Better than sec? Nothing is better than sec when it comes to > monitoring Big Data applications. Try Boundary one-second > resolution app monitoring today. Free. > http://p.sf.net/sfu/Boundary-dev2dev > ___ > Musicpd-dev-team mailing list > Musicpd-dev-team@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team > -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] DSD over USB
On 2012/04/01 13:22, Bobby Beever wrote: > You probably don't care, and probably too late.. but I do not agree > with the addition of the added DSD sample type. I've attached my > patch to give you an idea where I was heading to with the added > FRAME_TYPE which in my opinion is the cleanest. Yes, I think that is the right direction. It's obviously not done yet, and you should define the meaning of sample_format for non-PCM frames. I guess sample_format could be ignored for DSD frames. Meanwhile, I have applied a few more changes to the MPD core that may be relevant for you: I have removed some audio_format settings are somewhat redundant or lead to redundant code. For example, pcm_export is now responsible for packing 24 bit samples and reversing the byte order, and the code has been moved out of the pcm_convert library, which grew too complex after all these years. Max -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev ___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] Current mpd-git DSD playback problems/slowdown - possible clues
On 2012/04/02 18:23, Jurgen Kramer wrote: > After much digging around I managed to get sound working but the timing > issue remained. When the value for SAMPLE_FORMAT_DSD is changed from 1 > to 4 (src/audio_format.h) I get proper sound when playing DSD-over-USB. > This ratio of 1 to 4 makes sense but unfortunately is not the (complete) > solution. For example with SAMPLE_FORMAT_DSD set to 4, DSD2PCM still > works fine but now it's timing is off. No, I can't imagine how changing the (meaningless) numeric values of symbolic enums can make a difference for anything. Max -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev ___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] [PATCH] Add support for DSD-over-USB version 1.0, remove pre-v1 support - updated docs
On 2012/04/02 18:36, Jurgen Kramer wrote: > Hi, > > Attached patch updates the documentation. Modified the links to the > documentation of the DSD-over-USB standard version 1.1. > JR will make the link live later today. This one and the DSD1.0 patch merged to master. -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev ___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] [PATCH] update: properly skip symlinks in path that is to be updated.
Merged to v0.16.x -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev ___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
Re: [Musicpd-dev-team] Link broken?
On Wed, Apr 04, 2012 at 04:23:53PM +0800, zw g wrote: > Hi guys, > > > The following URL seems broken: > > http://www.musicpd.org/doc/protocol/ > > Someone confirm for me? > Thanks. I can confirm that it's broken in that it's a redirection to the wiki, but I can't fix it. Jonathan Neuschäfer -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev ___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
[Musicpd-dev-team] Link broken?
Hi guys, The following URL seems broken: http://www.musicpd.org/doc/protocol/ Someone confirm for me? Thanks. -- Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev___ Musicpd-dev-team mailing list Musicpd-dev-team@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/musicpd-dev-team
