Posts: 13,391
Threads: 304
Joined: Mar 2018
Reputation:
542
12-13-2018, 07:42 PM
(This post was last modified: 12-13-2018, 07:42 PM by Tim Curtis.)
I can't really help with the tweaks and the audio glitches but wrt the aiff files, I'm not seeing any issues on my end.
I added a NOTE to the 0.21.3 recipe that mentions that MPD database update has to be run after installing this version.
I would check the MPD log to see if any errors were reported when trying to play the aiff file.
Code: cat /var/log/mpd/log
Posts: 107
Threads: 15
Joined: Oct 2018
Reputation:
9
(12-13-2018, 05:35 PM)yaslam Wrote: Edward's Tweaks works well with moode 4.4 with MPD 0.20.20.
However, in DSD 128 and DSD 256, playback is interrupted every few seconds.
Also, I updated MPD to 0.21.3 by the following procedure.
http://moodeaudio.org/forum/showthread.p...82#pid6182
With MPD 0.21.3, certain PCM files can not be played correctly.
Is there a way to identify the cause of these problems?
Or are there any possible causes?
My information
Model: Raspberry Pi 3 Model B +
HAT: Allo DigiOne
Hi Yaslam
This sounds like overruns....too much data trying to fit through a small whole at the same time. Try adjusting the MPD fragment size and/or period time.
/etc/mpd.conf
Look for the audio_output line, you might want to also add memory mapping so MPD has a dedicated part of RAM to play in/with.
audio_output {
type "alsa"
name "my ALSA device"
device "hw:1,0"
mmap_emulation "true"
fragment_size "4096"
}
Posts: 13,391
Threads: 304
Joined: Mar 2018
Reputation:
542
I think those MPD options were deprecated quite some time ago. More recently, options including gapless_mp3_playback, buffer_before_play, id3v1_encoding and others are also deprecated. There are various reasons for example memory mapping and gapless mp3 playback are automatically handled as are certain types of buffering.
The code fragment below from MPD sources shows some of the options that have been deprecated. The indication is { param, repeatable, deprecated } for example { "buffer_before_play", false, true }.
https://github.com/MusicPlayerDaemon/MPD...plates.cxx
Code: const ConfigTemplate config_param_templates[] = {
{ "music_directory" },
{ "playlist_directory" },
{ "follow_inside_symlinks" },
{ "follow_outside_symlinks" },
{ "db_file" },
{ "sticker_file" },
{ "log_file" },
{ "pid_file" },
{ "state_file" },
{ "state_file_interval" },
{ "restore_paused" },
{ "user" },
{ "group" },
{ "bind_to_address", true },
{ "port" },
{ "log_level" },
{ "zeroconf_name" },
{ "zeroconf_enabled" },
{ "password", true },
{ "local_permissions" },
{ "default_permissions" },
{ "audio_output_format" },
{ "mixer_type" },
{ "replaygain" },
{ "replaygain_preamp" },
{ "replaygain_missing_preamp" },
{ "replaygain_limit" },
{ "volume_normalization" },
{ "samplerate_converter" },
{ "audio_buffer_size" },
{ "buffer_before_play", false, true },
{ "http_proxy_host", false, true },
{ "http_proxy_port", false, true },
{ "http_proxy_user", false, true },
{ "http_proxy_password", false, true },
{ "connection_timeout" },
{ "max_connections" },
{ "max_playlist_length" },
{ "max_command_list_size" },
{ "max_output_buffer_size" },
{ "filesystem_charset" },
{ "id3v1_encoding", false, true },
{ "metadata_to_use" },
{ "save_absolute_paths_in_playlists" },
{ "gapless_mp3_playback", false, true },
{ "auto_update" },
{ "auto_update_depth" },
{ "despotify_user", false, true },
{ "despotify_password", false, true },
{ "despotify_high_bitrate", false, true },
};
The links below have more information on the current list of available MPD options.
https://github.com/MusicPlayerDaemon/MPD...mpd.conf.5
https://github.com/MusicPlayerDaemon/MPD...nf.example
https://www.musicpd.org/doc/html/user.html
Posts: 27
Threads: 2
Joined: Oct 2018
Reputation:
1
12-14-2018, 07:06 PM
(This post was last modified: 12-14-2018, 07:07 PM by yaslam.)
Hi tim, Edward
I assigned only one CPU to the MPD by taskset, but I changed it to three CPUs.
As a result of that change, DSD128 and DSD 256 files are playback without problems.
Thank you.
Also, I tried MPD 0.21.3 without setting Sound Tweaks. In that case as well, I can not play a specific AIFF file.
Since this problem has nothing to do with Sound Tweaks, I write it on another thread.
Posts: 107
Threads: 15
Joined: Oct 2018
Reputation:
9
(12-14-2018, 07:06 PM)yaslam Wrote: Hi tim, Edward
I assigned only one CPU to the MPD by taskset, but I changed it to three CPUs.
As a result of that change, DSD128 and DSD 256 files are playback without problems.
Thank you.
Also, I tried MPD 0.21.3 without setting Sound Tweaks. In that case as well, I can not play a specific AIFF file.
Since this problem has nothing to do with Sound Tweaks, I write it on another thread.
Hi Yaslam
Glad everything worked out for you
Posts: 27
Threads: 2
Joined: Oct 2018
Reputation:
1
Hi Edward,
The sound quality of your Sound Tweaks is wonderful.
There is one question.
Is there any difference between your Sound Tweaks and the Sound Tweaks of following posted by mactolinux?
http://moodeaudio.org/forum/showthread.p...55#pid4355
Posts: 107
Threads: 15
Joined: Oct 2018
Reputation:
9
(12-16-2018, 08:47 AM)yaslam Wrote: Hi Edward,
The sound quality of your Sound Tweaks is wonderful.
There is one question.
Is there any difference between your Sound Tweaks and the Sound Tweaks of following posted by mactolinux?
http://moodeaudio.org/forum/showthread.p...55#pid4355
Hi Yaslam
They are very similar in the ram settings and over/undervoltage. Both pay attention the FIFO and priority too. With mine I include some kernel level changes which seeks to make the Linux OS perform as a 'whole' with less latency...again this is from RedHat so if a company that size with over 10k employees did the research on this and published the results I tend to believe it. I've been using these and other settings for must be more than 10years +/- with pretty good results. Glad you too and others find some improvements.
Posts: 41
Threads: 3
Joined: Oct 2018
Reputation:
1
12-16-2018, 08:42 PM
(This post was last modified: 12-17-2018, 08:05 AM by mactolinux.)
Funny thing: I did many tests (looong tests) with i2s recording to SSD/HDD, RPi as a slave, in real-time modes (at high rates, 192-24/32, with SOX, arecord, Jack, whatever), and discovered that real-time tuning is something that put linux (RPi are not power heroes) into a very stressed situation... As result – recording has lot of mistakes. Instead, regular, almost out-of-the-box Lite Raspbian with basically generic multi-tasking setup is exact in what is it recording (bit to bit, literally). With some very light config tuning.
How is this linked to a regular playback procedures? I don't know, but I would take RT/low latency tuning with care, again comparing before and after, thanks it is easy to do by changing SD-cards.
P.S. You may find sometime even a bit curious stories, like "why am I getting overruns during recording? – you shall go to RT settings and use jackd, no overruns there". Almost true, but... when you look inside the data flow and compare it with the source, you see more problems than before...
Posts: 27
Threads: 2
Joined: Oct 2018
Reputation:
1
(12-16-2018, 05:35 PM)Edward Wrote: (12-16-2018, 08:47 AM)yaslam Wrote: Hi Edward,
The sound quality of your Sound Tweaks is wonderful.
There is one question.
Is there any difference between your Sound Tweaks and the Sound Tweaks of following posted by mactolinux?
http://moodeaudio.org/forum/showthread.p...55#pid4355
Hi Yaslam
They are very similar in the ram settings and over/undervoltage. Both pay attention the FIFO and priority too. With mine I include some kernel level changes which seeks to make the Linux OS perform as a 'whole' with less latency...again this is from RedHat so if a company that size with over 10k employees did the research on this and published the results I tend to believe it. I've been using these and other settings for must be more than 10years +/- with pretty good results. Glad you too and others find some improvements.
Hi Edward,
Thank you for explaining about Sound Tweaks.
I was doing PC audio for a long time at Audirvana on Mac, but recently introduced Raspberry Pi.
Therefore, I'm unfamiliar with Linux OS. It was very helpful.
Posts: 107
Threads: 15
Joined: Oct 2018
Reputation:
9
(12-17-2018, 11:11 AM)yaslam Wrote: (12-16-2018, 05:35 PM)Edward Wrote: (12-16-2018, 08:47 AM)yaslam Wrote: Hi Edward,
The sound quality of your Sound Tweaks is wonderful.
There is one question.
Is there any difference between your Sound Tweaks and the Sound Tweaks of following posted by mactolinux?
http://moodeaudio.org/forum/showthread.p...55#pid4355
Hi Yaslam
They are very similar in the ram settings and over/undervoltage. Both pay attention the FIFO and priority too. With mine I include some kernel level changes which seeks to make the Linux OS perform as a 'whole' with less latency...again this is from RedHat so if a company that size with over 10k employees did the research on this and published the results I tend to believe it. I've been using these and other settings for must be more than 10years +/- with pretty good results. Glad you too and others find some improvements.
Hi Edward,
Thank you for explaining about Sound Tweaks.
I was doing PC audio for a long time at Audirvana on Mac, but recently introduced Raspberry Pi.
Therefore, I'm unfamiliar with Linux OS. It was very helpful.
So coming from Audirvana how are you liking MoOde?
I'm sure you know this DIY world can be unpredictable at times.
|