CD Player

[Nutul]

I'll post an issue on github

As reported in the threads above, the problem is in cdio_paranoia, which is spectacular for ripping CDs, while not so SOTA for playing them; hence the above configuration for MPD to pass over different parameters to cdio_paranoia to try to mitigate, and possibly eliminate, the problem.

[Yawarakaimono]

(Almost) Solution:

Setting SoX Resampling option 'Multithreading' to 'No' makes the music playing for much longer time.
However, every now and then it stops for a moment.

[Nutul]

(Almost) Solution:

Setting SoX Resampling option 'Multithreading' to 'No' makes the music playing for much longer time.
However, every now and then it stops for a moment.

Have you tried leaving SoX resampling completely out of the picture?

[Yawarakaimono]

Mmm

Not yet. In the same stack is already a normal (rather old) CD player, I was curious if my ears could hear the difference between that one and a modern DAC with resampling. If I can't use that, I will stick with the player.

[TheOldPresbyope]

I haven't paid much attention to my old moOdeCDPlay  repo in many moons. The mention of SoX, though, has triggered a memory. Sure enough, the last section of the ReadMe of my repo says

SoX

A problem experienced in early testing with audible dropouts during CD playback was traced to SoX resampling at 32 bit/384KHz (inadvertently left on after a stress test of the RPi4B). Disabling resampling eliminated the dropouts. Your mileage with SoX resampling may vary.

I suggest the maintainer of the repo you're working from (which began as a fork of mine) say something similar.

As always, enjoy the music.

Regards,
Kent

[Nutul]

Mmm

Not yet.  In the same stack is already a normal (rather old) CD player, I was curious if my ears could hear the difference between that one and a modern DAC with resampling. If I can't use that, I will stick with the player.

Resampling a CD has no other benefit than that of easing the work of the DAT (if it is a resampling one, which almost all are, leaving the R2Rs out of the equation, of course).

A fair comparison, anyway, should be between an old and a new CD player, at the same sample-rate... I mean, without cheating (although in the end, as I said, the DAC will do the up-sampling anyway). ;-)

[Yawarakaimono]

Back to moOdeCDPlayer

When a CD is paused, the disc head can be heard moving, about every 3 seconds.
A verbose mpd log shows many occurences of:

----------------------------------------------------------------------------
Checking /dev/cd rom for cdrom...
CDROM sensed: HL-DT-ST DVDRAM GH22NS50 TN02 SCSI CD-ROM
Sep 17 14:37 : cdio: drive returns audio data Little Endian
Sep 17 14:37: client: [38] command returned 3
Sep 17 14:37: client: [38] process command "close"
Sep 17 14:37: client: [38] command returned 4
Sep 17 14:37: client: [38] closed
Sep 17 14:37: client: [39] opened from 127.0.0.1:38854
Sep 17 14:37: client: [39] process command "status"
Sep 17 14:37: client: [39] command returned 0
Sep 17 14:37: client: [39] command returned Sep 17 14:37 : client: [39] process command "currentsong"
Sep 17 14:37: client: [39] command returned 0
Sep 17 14:37: client: [39] process command "lsinfo "cdda:///3""
.
.
.
Checking /dev/cd rom for cdrom...
CDROM sensed: HL-DT-ST DVDRAM GH22NS50 TN02 SCSI CD-ROM
----------------------------------------------------------------------------

And so on. Why is this activity running when the CD has not be changed?
It also continues when music is playing.

Can it interfere with the timely delivery of blocks of audio data?

When ripping with abcde or playing audio from cd with CD the drive is quietly humming along.
All that extra disc head moves seem abnormal, or at least unnecessary behavour to me.

[Yawarakaimono]

I'll post an issue on github

As reported in the threads above, the problem is in cdio_paranoia, which is spectacular for ripping CDs, while not so SOTA for playing them; hence the above configuration for MPD to pass over different parameters to cdio_paranoia to try to mitigate, and possibly eliminate, the problem.

[Nutul]

Sep 17 14:37: client: [39] command returned Sep 17 14:37 : client: [39] process command "currentsong"

That is the process for letting the "now plying" information available in the file /var/local/www/currentsong.txt, in case some external program needs to know it.
It's weird that it causes the head to move somewhere, though...

You can disable it in Settings => Audio, under MPD Options, uncheck "Metadata file".

Although we may spend some time to investigate this, and maybe (if it can affect the audio reproduction) disable / think of an alternative way in case of CD playback.

[Yawarakaimono]

Without 'updating the Metadata file' CD's play without pieces of silence.