Re: Convert mp3 enbulk

[Jon Dowland <jmtd@debian.org>]

On 20/02/12 04:39, Carl Fink wrote:
> 	find . -name "*mp3" -exec oggenc<options>  {} \;

On a multi-core system, for a small-ish number of files, I would suggest

find . -name "*mp3" -exec oggenc<options>  {} +

which will run the oggenc processes in parallel.

(Of course you have two distinct lossy compression steps doing ogg → 
mp3, but that's another matter).

For a larger number, I'd decide on a number of parallel encoders I'd 
like, usually <number of cores>*2 + n or similar; and do

find . -name "*mp3" -print0 | xargs -r0 -P $number oggenc <options>

As for metadata, I use "eyeD3" and short-ish shell script snippets,
e.g.

for i in *mp3; do title=$(echo "$i" | sed 's/\.mp3$//' | sed 's/^[0-9] 
//'); eyeD3 --title="$title" "$i"; done

and

for i in *mp3; do image=$(echo "$i"|sed 's/mp3/jpg/'); eyeD3 
"--add-image=$image:FRONT_COVER" "$i"; done

etc.