Thanks @bpa. More to digest. The tee command can prove useful to compare
Ogg/FLAC stream to ffmpeg capture at alsa output. Ahh of course,
directing the convertion to an application of own making opens up many
possibilities. Thanks, inspires to think of new ways of using LMS.

With data corruption I'm thinking at byte level not interupting the
stream. As in why does DeltaWave find 1 dB difference in RMS level and
still return 40% bit accuratsy? I would expect it to be null when
amplitude differs. And why would higher frequencies be less accurate?

THE BIG ELEPHANT IN THE ROOM ARE OF COURSE IF DELTAWAVE RESULTS ARE TO
BE TRUSTED [/B]WITH CURRENT SETUP. 

NOW BANGING AT THIS FOR A COUPLE OF DAYS AND NOW WITH SEVERAL DIFFERENT
REFERENCE FILES LATER A PATTERN SEEM TO EMERGE. TODAY I WHERE ACTUALLY
ABLE TO HAVE DELTAWARE SOMETIMES REPORT BIT PERFECT CAPTURES FROM FLAC
FILES. 
WHEN THIS HAPPENS DELTAWARE USE MUCH LESS TIME TO COME TO THIS
CONCLUSION. OFTEN BUT NOT ALWAYS IT WARNS IT CHANGE METHODS TO ALIGN THE
TRACKS FOR COMPARISON AND WOULD SPEND LONGER TIME BEFORE FINISHING. AND
WHEN THIS HAPPEN IT IS MORE LIKELY THE COMPARANCE WILL SHOW LESS
ACCURACY.

NOW IF DELTAWAVE FAILS TO ALIGN THE TWO FILES IT COMPARES THIS COULD
HELP EXPLAIN THE STRANGE RESULTS MENTIONED ABOVE. ALTHOUGH MY THINKING
PCM DATA SHOULD BE EXACTLY THE SAME WHERE ONLY TIME DOMAIN WITH VARYING
CLOCK ON THE CAPTURE COULD ADD TO THE COMPLEXITY.


[B]IF OTHERS WOULD TAKE PART IN DEBUNKING MY FINDINGS HERE IS MY SETUP.

RPi3 running lateset RPi-OS and Squeezelite v1.9.8-1317 installed from
'apt'. 
LMS 8.3.0 on dedicated RPi3 with RPi-OS

Tools used:
DeltaWave 2.0.4 https://deltaw.org/

To capture the stream I use
~$ ffmpeg -f alsa -channels 2 -sample_rate 44100 -i loopout
fromFLAC.wav
with /etc/asound.conf copied from bottom of this page
https://trac.ffmpeg.org/wiki/Capture/ALSA

Squeezelite output 
/etc/default/squeezelite
SL_SOUNDCARD=default

I record 2 minuttes and use the original FLAC file as Reference in
DeltaWave where I calculate the End for this file to be approx same
lenght as the recording.


------------------------------------------------------------------------
zzzap's Profile: http://forums.slimdevices.com/member.php?userid=31947
View this thread: http://forums.slimdevices.com/showthread.php?t=117152

_______________________________________________
Squeezecenter mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/squeezecenter

Reply via email to