LAME mp3 downsample
September 28, 2006 2:16 PM Subscribe
I need to batch process mp3 files from high bitrate to lower bitrate. LAME can do that - I think, but can anyone help me optimize for spoken mp3 files? I need a command line example and it might need a lowpass filter setting.
Thanks
BB
Best answer: lame has a voice preset that's not bad.
For example:
lame --preset voice oldfile.mp3 newfile.mp3
may do what you want. If you need additional
lowpass filtering:
lame --preset voice --lowpass 10 old.mp3 new.mp3
for 10K filter cutoff.
posted by spasm at 4:28 PM on September 28, 2006
For example:
lame --preset voice oldfile.mp3 newfile.mp3
may do what you want. If you need additional
lowpass filtering:
lame --preset voice --lowpass 10 old.mp3 new.mp3
for 10K filter cutoff.
posted by spasm at 4:28 PM on September 28, 2006
I don't think the "voice" preset is officially supported, is it? It doesn't appear in the longhelp.
I have had good luck with:
-V3 --vbr-new --lowpass 8
You can try different settings, but the "new" vbr routine is very fast.
e.g. "./lame --V3 --vbr-new --lowpass 8 song-hi.mp3 song-lo.mp3
Do you need shell script to do the batch, too?
posted by bafflegab at 4:54 PM on September 28, 2006
I have had good luck with:
-V3 --vbr-new --lowpass 8
You can try different settings, but the "new" vbr routine is very fast.
e.g. "./lame --V3 --vbr-new --lowpass 8 song-hi.mp3 song-lo.mp3
Do you need shell script to do the batch, too?
posted by bafflegab at 4:54 PM on September 28, 2006
Response by poster: Hi
Yes, I need a script too, I you have one handy ... otherwise I have to call Larry Wall again ... ;-(
thx
BB
posted by bright77blue at 6:31 PM on September 28, 2006
Yes, I need a script too, I you have one handy ... otherwise I have to call Larry Wall again ... ;-(
thx
BB
posted by bright77blue at 6:31 PM on September 28, 2006
There's a tool that is already in existance to do this that I use. Its not commandline based, theoretically. Switch should do what you're looking for.
What I do is take a pod... er... netcasts which may be anything from 128kbps to 64kbps and put it into Switch, set the output to mp3 - 16kbps - mono output. This will usually drop down a 2 Hour mp3 to something in the range of 20mb depending on the quality I use. So I can put several hours of audio onto my cheap 128mb mp3 player.
You CAN do 8kbps, but I find the tinniness of the output to be very distracting. Go with 8kbps only if you really need to size to be miniscule and you're not overly concerned about being annoyed by listening to it.
posted by gregschoen at 12:58 AM on September 29, 2006
What I do is take a pod... er... netcasts which may be anything from 128kbps to 64kbps and put it into Switch, set the output to mp3 - 16kbps - mono output. This will usually drop down a 2 Hour mp3 to something in the range of 20mb depending on the quality I use. So I can put several hours of audio onto my cheap 128mb mp3 player.
You CAN do 8kbps, but I find the tinniness of the output to be very distracting. Go with 8kbps only if you really need to size to be miniscule and you're not overly concerned about being annoyed by listening to it.
posted by gregschoen at 12:58 AM on September 29, 2006
I don't think the "voice" preset is officially supported, is it? It doesn't appear in the longhelp.
It's mentioned in "--preset help".
posted by spasm at 1:21 PM on September 29, 2006
It's mentioned in "--preset help".
posted by spasm at 1:21 PM on September 29, 2006
This thread is closed to new comments.
posted by gergtreble at 4:14 PM on September 28, 2006