parsing tags from WAV files seems time consuming

[dean]

After the initial scan, the server shouldn't need to access the files
again, except to play them.

Can you give me exact steps (button presses) that you take to make this
happen? Is this in the web interface or with the remote....



On Jan 16, 2004, at 4:06 PM, John Snell wrote:

> I have a slimserver 5.0.1 running on linux with a library of ~1800
> .wav files.
> Most of these tracks are of the Classical genre. The first time one
> selects
> 'Classical' from the 'Browse by Genre' list, the server grinds the
> hardisk
> and becomes unresponsive for the better part of an hour. It recovers
> from
> this and then works normally and quickly.
>
> Looking through the code, it looks like the Perl module Audio::Wav is
> the
> culprit. Slimserver seems to use this to parse tag values (duration,
> etc.)
> from the wav files. Adding a few logging statements revealed that this
> process takes 1-2 sec per track. Presumably the entire wav file is
> being read
> end-to-end? Is there anything that can be done to speed this process
> up?
>
> -John
>
>
>

[John Snell]

I have a slimserver 5.0.1 running on linux with a library of ~1800 .wav files.
Most of these tracks are of the Classical genre. The first time one selects
'Classical' from the 'Browse by Genre' list, the server grinds the hardisk
and becomes unresponsive for the better part of an hour. It recovers from
this and then works normally and quickly.

Looking through the code, it looks like the Perl module Audio::Wav is the
culprit. Slimserver seems to use this to parse tag values (duration, etc.)
from the wav files. Adding a few logging statements revealed that this
process takes 1-2 sec per track. Presumably the entire wav file is being read
end-to-end? Is there anything that can be done to speed this process up?

-John