Page 2 of 2

Re: Corrupted audio files

Posted: 15 Jun 2020 15:51
by fred48fm
Thank you for your time again !

The backup files and the files used on the computer hosting PF are exactly the same files (no transcoding, etc).
The hard drive seems OK (chkdsk and SMART state) but i'll take a deeper look for bad clusters.

The point about the long use of hard drive and audio drop-out is happening as such :

When setting the option "auto save metadata to audio tag" (with all metadata checked)
1. A WAV file >1 hours is actually playing in PF
2. A WAV file > 30min is enqued in PF (never played before, and never analysed by a PF instance before)
-> Audio drop out during the last minutes of the current playing file and audio drop out in the first minutes of the next audio file
The hard drive use is at maximum usage.

With disabling the "save metadata to audio tag", problem disappeared and disk usage is normal...

Re: Corrupted audio files

Posted: 15 Jun 2020 16:37
by radio42
I guess, this must also have a different reason and coincident!
When the "Auto Save MetaData to Audio TAG" option is selected, the meta data is automatically saved, when the meta data was changed in one of the related Players, e.g. the DJ Player or the PFL Player etc. This requires a manual action, e.g. setting of cue-point or alike.
The meta data will for example then only be saved once a track is unloaded from the DJ Player.
Thus no heavy IO load is generated by just turning this option on.

So the heavy IO load must be related to something different!

Re: Corrupted audio files

Posted: 17 Jun 2020 17:28
by fred48fm

Problem solved...!
I'm deeply sorry for the time you spend on this.

The problem seems to be ... a faulty SATA cable.
SMART monitoring was not working properly. After further analyse with another SMART software, I found a huge amount (billions) of disk timeout errors.

So the huge disk usage for simple task, like writing metadata to file, is mostly a result of these errors.

Best regards,


Re: Corrupted audio files

Posted: 17 Jun 2020 22:06
by radio42
No problem at all - I am happy, that you found the issue!