3 Feb
2007
3 Feb
'07
5:39 p.m.
hi,
just starting to use ardour2 for production (thanks everyone for the
help) and running into a little frustration with the mp3 files from my
recorder. ardour of course doesn't import from mp3; I can, I suppose,
convert the files to mp3 one at a time when I want to, but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
thanks,
matt
--
Matt Price
History Dept
University of Toronto
matt.price(a)utoronto.ca
3 Feb
3 Feb
5:58 p.m.
On Sat, 2007-02-03 at 16:39 -0500, Matt Price wrote:
hi,
just starting to use ardour2 for production (thanks everyone for the
help) and running into a little frustration with the mp3 files from my
recorder. ardour of course doesn't import from mp3; I can, I suppose,
convert the files to mp3 one at a time when I want to, but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
this is an area where, unlike all proprietary DAWs, we'd prefer you to
leverage the power of the platform you are running Ardour on.
for example, suppose you have a program called mp3towave and a directory
full of mp3 files, you can create a script that looks like this:
------- cut here ------------
#!/bin/sh
for file in $*
do
mp3towave $file
done
------- and here -------------
the beauty of this approach is that although today you "just" want to do
mp3 conversion, next week, you can adapt this to something else. for
example:
* convert them to aiff instead
* convert and resample them
* convert them into mono files, 2 per mp3 file
* rename them all based on some rule
* convert ogg/vorbis files instead
* convert AAC files instead
* convert them and filter them as well
and so on. if we built this into Ardour, you'd end up with a huge mass
of code duplicating the cool functionality that Unix has offered you for
30 years or more ;)
let me know how this looks. i have not discussed specific tools you
could use for the conversion since it sounds as if you have that
covered.
--p
6 Feb
6 Feb
12:27 p.m.
On Sat, 2007-02-03 at 16:57 -0500, Paul Davis wrote:
On Sat, 2007-02-03 at 16:39 -0500, Matt Price wrote:
> hi,
>
> just starting to use ardour2 for production (thanks everyone for the
> help) and running into a little frustration with the mp3 files from my
> recorder. ardour of course doesn't import from mp3; I can, I suppose,
> convert the files to mp3 one at a time when I want to, but I wonder if
> there's a more fluid way
this is an area where, unlike all proprietary DAWs,
we'd prefer you to
leverage the power of the platform you are running Ardour on.
for example, suppose you have a program called mp3towave and a directory
full of mp3 files, you can create a script that looks like this:
------- cut here ------------
#!/bin/sh
for file in $*
do
mp3towave $file
done
------- and here -------------
the beauty of this approach is that although today you "just" want to do
mp3 conversion, next week, you can adapt this to something else.
let me know how this looks. i have not discussed specific tools you
could use for the conversion since it sounds as if you have that
covered.
thanks paul first of all for these excellent tools which I'm just
starting to enjoy and now for this advice. Also thanks to folks for
the excellent scripts and tips, am using the python script now with a
few mods and it works great.
I notice however that ardour pretty well always converts my .wav files,
no matter what sample rate I set them to. IS there another format to
which I can pre-convert my material so that loading it into ardour is
relatively rapid and painless? Or am I missing something? So for
instance I have a wave file here:
$ file 001_first_entry.wav
001_first_entry.wav: RIFF (little-endian) data, WAVE audio, Microsoft PCM, 16 bit, stereo
44100 Hz
jack is running at 44100 Hz, so I at firstthought the issue had to do
with bit-rate. However, madplay can output 32-bit audio (just 24-bit
with padding, really, as far as I understand), so I tried just now with
a 32-bit file and again had to wait. Am I using the wrong file
format? Or are there any other suggestions? THis isn't a big issue
but it would be nice to havei nstant access to the audio.
thx,
matt
--p
--
Matt Price
History Dept
University of Toronto
matt.price(a)utoronto.ca
2:34 p.m.
On Mon, 2007-02-05 at 21:27 -0500, Matt Price wrote:
I notice however that ardour pretty well always
converts my .wav files,
no matter what sample rate I set them to. IS there another format to
which I can pre-convert my material so that loading it into ardour is
relatively rapid and painless? Or am I missing something? So for
instance I have a wave file here:
just embed them rather than importing them. importing always means that
a new file is written; embedding means that we'll just use the file as
is (if possible).
3 Feb
3 Feb
6:16 p.m.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sat, Feb 03, 2007 at 04:39:31PM -0500, Matt Price wrote:
hi,
just starting to use ardour2 for production (thanks everyone for the
help) and running into a little frustration with the mp3 files from my
recorder. ardour of course doesn't import from mp3; I can, I suppose,
convert the files to mp3 one at a time when I want to, but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
No big deal. Here's the bash one-liner I use:
for i in `ls /path/to/your/mp3s`; do F=`basename $i`; mpg123 -r 48000 -w $F $i; done
Obviously, change the /path/to/your/mp3s to wherever they are, and change 48000 to
whatever rate you have ardour and jack set to.
- -ken
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
iD8DBQFFxQmAe8HF+6xeOIcRAlQYAKDw9HqbwTrXTTxkeJGHXzsBu5eTHwCgp6Qo
LOUaJseOzZ7CkpsDXIZlXhg=
=rOth
-----END PGP SIGNATURE-----
6:33 p.m.
On Sat, 2007-02-03 at 14:15 -0800, Ken Restivo wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sat, Feb 03, 2007 at 04:39:31PM -0500, Matt Price wrote:
a warning to all who write such scripts (including me). they will very,
very often break with mp3 files that are named from CD rips because of
embedded spaces. for example, in the little script i just posted, i
should have used "$@" not "$*" for precisely this reason.
long time *nix users will know that spaces have always been legal in
file names; new *nix users won't know that the oldtimers have never
really used them :)
--p
hi,
just starting to use ardour2 for production (thanks everyone for the
help) and running into a little frustration with the mp3 files from my
recorder. ardour of course doesn't import from mp3; I can, I suppose,
convert the files to mp3 one at a time when I want to, but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
No big deal. Here's the bash one-liner I use:
for i in `ls /path/to/your/mp3s`; do F=`basename $i`; mpg123 -r 48000 -w $F $i; done
Obviously, change the /path/to/your/mp3s to wherever they are, and change 48000 to
whatever rate you have ardour and jack set to. 
6:50 p.m.
On Sat, Feb 03, 2007 at 05:33:42PM -0500, Paul Davis wrote:
long time *nix users will know that spaces have always
been legal in
file names; new *nix users won't know that the oldtimers have never
really used them :)
doesn't everyone have a perl script that fixes spaces in files?
... they should:
#!/usr/bin/perl -w
### fix_fn_keep_ext.pl ###
if( ! defined($ARGV[0]) ){
die "$0 <file>[files...]\n";
}
foreach ( @ARGV ){
if( -f $_ ){
rename( $_ , fix_fn( $_ ) );
} else {
die "$file is not a file(or doesn\'t exist)";
}
}
#########################################
sub fix_fn { # string
my($fn);
my(@fn_arr);
if(defined($_[0])){
$fn = $_[0];}
else{ die"fix_fn() needs 1 arg : $!";}
@fn_arr = split_on_suffix( $fn );
$fn = replace_bad_chars( $fn_arr[0] );
$fn = $fn . '.' . $fn_arr[1];
return( $fn);
}
sub replace_bad_chars { # string
my($fn);
my(@fn_arr);
if(defined($_[0])){
$fn = $_[0];}
else{ die"replace_bad_chars() needs 1 arg : $!";}
@fn_arr = split( /\W+/ , $fn);
$fn = join( '_' , @fn_arr ) ;
return( $fn );
}
sub split_on_suffix { # string
my($fn);
my($suffix);
my(@fn_arr);
if(defined($_[0])){
$fn = $_[0];}
else{ die"strip_suffix() needs 1 arg : $!";}
@fn_arr = split( /\./ , $fn );
$suffix = pop( @fn_arr );
$fn = join( '.' , @fn_arr );
@fn_arr = ( $fn , $suffix );
return( @fn_arr );
}
5 Feb
5 Feb
4:55 p.m.
2007/2/3, paul wisehart <listreader(a)lupulin.net>et>:
On Sat, Feb 03, 2007 at 05:33:42PM -0500, Paul Davis
wrote:
I have a php-script that:
- converts spaces to "_"
- converts to lower-case
- changes " - " ("_-_" after the first step) to "-"
- works recursive
- also renames directories (before or after recursing into them)
Before you ask, its attached...
Arnold
--
visit http://www.arnoldarts.de/
---
Wenn man mit Raubkopien Bands wie Brosis oder Britney Spears wirklich
verhindern könnte, würde ich mir noch heute einen Stapel Brenner und
einen Sack Rohlinge kaufen.
long time *nix users will know that spaces have
always been legal in
file names; new *nix users won't know that the oldtimers have never
really used them :)
doesn't everyone have a perl script that fixes spaces in
files?
4 Feb
4 Feb
12:06 a.m.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sat, Feb 03, 2007 at 05:33:42PM -0500, Paul Davis wrote:
On Sat, 2007-02-03 at 14:15 -0800, Ken Restivo wrote:
Point taken. Corrected version:
for i in `ls /path/to/your/mp3s`; do F=`basename "$i"`; mpg123 -r 48000 -w
"$F" "$i"; done
But I don't have many files with embedded spaces, and I wouldn't be using ripped
(i.e copyrighted) CD tracks in my work anyway, because I like to CC license it.
I occasionally end up with files with %20's in them, but the only things that seem to
consistently end up with embedded spaces are M$ Word documents that people sometimes send
me.
- -ken
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sat, Feb 03, 2007 at 04:39:31PM -0500, Matt Price wrote:
a warning to all who write such scripts (including me). they will very,
very often break with mp3 files that are named from CD rips because of
embedded spaces. for example, in the little script i just posted, i
should have used "$@" not "$*" for precisely this reason.
long time *nix users will know that spaces have always been legal in
file names; new *nix users won't know that the oldtimers have never
really used them :)
hi,
just starting to use ardour2 for production (thanks everyone for the
help) and running into a little frustration with the mp3 files from my
recorder. ardour of course doesn't import from mp3; I can, I suppose,
convert the files to mp3 one at a time when I want to, but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
No big deal. Here's the bash one-liner I use:
for i in `ls /path/to/your/mp3s`; do F=`basename $i`; mpg123 -r 48000 -w $F $i; done
Obviously, change the /path/to/your/mp3s to wherever they are, and change 48000 to
whatever rate you have ardour and jack set to. -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
iD8DBQFFxVuBe8HF+6xeOIcRAoToAJ9tsM3hZe30vo+rduI8Lj72enh5uACfU1vN
Jt7cv+SGoW5xjzwvDla7630=
=RiH8
-----END PGP SIGNATURE-----
3 Feb
3 Feb
6:17 p.m.
Matt Price wrote:
hi,
just starting to use ardour2 for production (thanks everyone for the
help) and running into a little frustration with the mp3 files from my
recorder. ardour of course doesn't import from mp3; I can, I suppose,
convert the files to mp3 one at a time when I want to, but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
thanks,
matt
hmmm... bash script for loop aught to do it. something like
#!/bin/bash
for i in `ls *.mp3`; do
$CONVERTER_PROGRAM $i ${i/mp3/wav}
done
where $CONVERTER_PROGRAM is something that takes two arguments, an mp3
file and a wave file name. this could be done with mplayer like this i
think.
mplayer -ao pcm:${i/mp3/wav} $i
here is more info on bash scripting.
http://tldp.org/HOWTO/Bash-Prog-Intro-HOWTO-7.html
http://tldp.org/LDP/abs/html/parameter-substitution.html
/brian
4 Feb
4 Feb
10:56 a.m.
Brian Dunn <job17and9(a)sbcglobal.net> writes:
for i in `ls *.mp3`; do
That won't work if the filenames contain spaces (since the shell will
split ls's output on spaces), and it's needlessly inefficient. It's
better to say:
for i in *.mp3; do
or if you really do want to use the output of a command, then:
find . -name '*.mp3' | while read i; do
(Not supporting spaces in filenames in the modern world is silly --
particularly for music files! Newlines I'm less concerned about...)
--
Adam Sampson <ats(a)offog.org> <http://offog.org/>
6 Feb
6 Feb
4:05 p.m.
Adam Sampson wrote:
Brian Dunn <job17and9(a)sbcglobal.net> writes:
Well, I stand corrected I've never used that while syntax before. It
looks useful. thanks.
for i in `ls *.mp3`; do
That won't work if the filenames contain spaces (since the shell will
split ls's output on spaces), and it's needlessly inefficient. It's
better to say:
for i in *.mp3; do
or if you really do want to use the output of a command, then:
find . -name '*.mp3' | while read i; do
(Not supporting spaces in filenames in the modern world is silly --
particularly for music files! Newlines I'm less concerned about...)
4 Feb
4 Feb
7:30 a.m.
Matt Price wrote:
convert the files to mp3 one at a time when I want to,
but I wonder if
there's a more fluid way of e.g. converting files when they're uploaded,
or at least converting all files in a particular folder or something.
are there other people out there who have to deal with a similar
situation?
I hacked up this python script to transcode from wav/mp3/ogg/flac to
wav/mp3/flac/ogg.
HTH
R
#!/usr/bin/python
# Copyright Rob Fell, 2007.
import sys
import string
import os
import os.path
import string
import shutil
import tempfile
from subprocess import *
best = False
def GetFlacInfo( meta, f ):
cmd = "metaflac --show-sample-rate --show-tag=artist --show-tag=album
--show-tag=title --show-tag=tracknumber --show-tag=date %s"
p = Popen( cmd % f, shell=True, stdout=PIPE, stderr=PIPE )
artist = album = year = song = tracknum = rate = None
for field in p.communicate()[0].split('\n'):
sep = field.find('=')
if sep > 0:
key = field[0:sep].lower().lstrip()
value = field[sep+1:]
if key == 'artist':
artist = value
if key == 'album':
album = value
if key == 'title':
song = value
if key == 'tracknumber':
tracknum = int( value )
if key == 'date':
year = value
else:
str = ""
for ch in field.lstrip():
if ch.isdigit():
str += ch
else:
break
if str:
rate = int( str )
if song: meta['song'].append( song )
if artist: meta['artist'].append( artist )
if album: meta['album'].append( album )
if tracknum: meta['tracknum'].append( tracknum )
if year: meta['date'].append( year )
if rate: meta['rate'].append( rate )
def GetOggInfo( meta, f ):
p = Popen( "ogginfo %s" % f, shell=True, stdout=PIPE )
artist = album = year = song = tracknum = rate = None
for field in p.communicate()[0].split('\n'):
sep = field.find('=')
if sep > 0:
key = field[0:sep].lower().lstrip()
value = field[sep+1:]
if key == 'artist':
artist = value
if key == 'album':
album = value
if key == 'title':
song = value
if key == 'tracknumber':
tracknum = int( value )
if key == 'date':
year = value
else:
if 'Rate:' in field:
rate = int( field.split(' ')[1] )
if song: meta['song'].append( song )
if artist: meta['artist'].append( artist )
if album: meta['album'].append( album )
if tracknum: meta['tracknum'].append( tracknum )
if year: meta['date'].append( year )
if rate: meta['rate'].append( rate )
def GetID3Info( meta, f ):
p = Popen( "id3info %s" % f, shell=True, stdout=PIPE )
artist = album = year = song = tracknum = rate = None
for field in p.communicate()[0].split('\n'):
if 'Frequency' in field:
freq = field.split(':')[-1].lstrip().lower()
if '44khz' == freq:
rate = 44100
elif '48khz' == freq:
rate = 48000
if not field[0:3] == '===':
continue
key = field[4:8]
if key == 'TIT2':
song = field.split(':')[-1].lstrip()
elif key == 'TPE1':
artist = field.split(':')[-1].lstrip()
elif key == 'TALB':
album = field.split(':')[-1].lstrip()
elif key == 'TRCK':
tracknum = int( field.split(':')[-1] )
elif key == 'TYER':
year = int( field.split(':')[-1] )
if song: meta['song'].append( song )
if artist: meta['artist'].append( artist )
if album: meta['album'].append( album )
if tracknum: meta['tracknum'].append( tracknum )
if year: meta['date'].append( year )
if rate: meta['rate'].append( rate )
def GetMetadata( ftype, f ):
meta = { 'artist' : [],
'album' : [],
'date' : [],
'tracknum' : [],
'song' : [],
'rate' : [] }
if ftype == 'riff':
return None
if ftype == 'flac':
GetFlacInfo( meta, f )
if ftype == 'ogg':
GetOggInfo( meta, f )
GetID3Info( meta, f )
if not meta['tracknum']:
tracknum = ""
for ch in os.path.basename( f ):
if ch.isdigit():
tracknum += ch
else:
break
if tracknum:
meta['tracknum'].append( int( tracknum ))
return meta
def DetermineFileType( f ):
idtable = [ (('MPEG', 'layer III'), ('mp3')),
(('RIFF', 'PCM', 'Microsoft'), ('riff')),
(('Ogg', 'Vorbis'), ('ogg')) ]
# 'file' misidentifies flac files as MP3's, so we use metaflac
# to see if it IS a flac file first
#
is_flac = Popen( "metaflac --list %s" % f, shell=True, stdout=PIPE, stderr=PIPE
).wait() == 0
if is_flac:
ftype = 'flac'
else:
# now fall back to normal 'file' identification
#
p = Popen( "file %s" % f, shell=True, stdout=PIPE )
info = p.communicate()[0]
ftype = None
for rule, tag in idtable:
ftype = tag
for str in rule:
if str not in info:
ftype = None
break
if ftype:
break
return ftype
def Flac2Wav( src, dst ):
return Popen( "flac123 --wav=%s %s" % (dst, src), shell=True, stdout=PIPE
).wait()
def Wav2Flac( src, dst, meta ):
cmd = "flac"
args = ""
if meta:
if meta['song']:
args += ' -T "title=%s"' % meta['song'][0]
if meta['artist']:
args += ' -T "artist=%s"' % meta['artist'][0]
if meta['album']:
args += ' -T "album=%s"' % meta['album'][0]
if meta['date']:
args += ' -T "date=%s"' % meta['date'][0]
if meta['tracknum']:
args += ' -T "tracknumber=%d"' % meta['tracknum'][0]
if args:
cmd += " %s %s %s" % (args, src, dst)
else:
cmd += " %s %s" % (src, dst)
return Popen( cmd , shell=True, stdout=PIPE ).wait()
def Ogg2Wav( src, dst ):
return Popen( "oggdec -o %s %s" % (dst, src), shell=True, stdout=PIPE ).wait()
def Wav2Ogg( src, dst, meta ):
if best:
cmd = "oggenc -q 10"
else:
cmd = "oggenc -q 7"
args = ""
if meta:
if meta['song']:
args += ' -t "%s"' % meta['song'][0]
if meta['artist']:
args += ' -a "%s"' % meta['artist'][0]
if meta['album']:
args += ' -l "%s"' % meta['album'][0]
if meta['date']:
args += ' -d "%s"' % meta['date'][0]
if meta['tracknum']:
args += ' -N %d' % meta['tracknum'][0]
if args:
cmd += " %s -o %s %s" % (args, dst, src)
else:
cmd += " -o %s %s" % (dst, src)
print cmd
return Popen( cmd , shell=True, stdout=PIPE ).wait()
def Mp32Wav( src, dst ):
return Popen( "mpg123 -w %s %s" % (dst, src), shell=True, stdout=PIPE ).wait()
def Wav2Mp3( src, dst, meta ):
if best:
cmd = "lame -V 0 -h "
else:
cmd = "lame -V 4 -h"
args = ""
if meta:
if meta['song']:
args += ' --tt "%s"' % meta['song'][0]
if meta['artist']:
args += ' --ta "%s"' % meta['artist'][0]
if meta['album']:
args += ' --tl "%s"' % meta['album'][0]
if meta['date']:
args += ' --ty "%s"' % meta['date'][0]
if meta['tracknum']:
args += ' --tn %d' % meta['tracknum'][0]
if args:
cmd += " %s %s %s" % (args, src, dst)
else:
cmd += " %s %s" % (src, dst)
return Popen( cmd , shell=True, stdout=PIPE ).wait()
def NewName( f, meta, type ):
types = { 'riff':'wav', 'flac':'flac',
'ogg':'ogg', 'mp3':'mp3' }
if type not in types.keys():
return None
if meta and meta['song']:
if meta['tracknum']:
song = "".join( [ x for x in meta['song'][0].split(' ')] )
name = "%02d%s" % (meta['tracknum'][0], song)
trans = string.maketrans( '', '' )
return ".".join( [string.translate( name, trans,
':,\\/.!\"$%^&*()+=-\'_ ' ), types[type]] )
else:
name = os.path.splitext( os.path.basename( f ))[0]
trans = string.maketrans( '', '' )
return ".".join( [string.translate( name, trans,
':,\\/.!\"$%^&*()+=\' ' ), types[type]] )
def Transcode( src, dst, meta ):
srcFmt, srcFile = src
dstFmt, dstFile = dst
decoders = { 'flac': Flac2Wav,
'ogg' : Ogg2Wav,
'mp3' : Mp32Wav,
'riff': None }
encoders = { 'flac': Wav2Flac,
'ogg' : Wav2Ogg,
'mp3' : Wav2Mp3,
'riff': None }
if os.path.isfile( dstFile ):
raise "refusing to overwrite %s" % dstFile
if not os.path.isdir( os.path.dirname( dstFile )):
raise "directory %s does not exist" % os.path.dirname( dstFile )
if srcFmt == dstFmt:
# only a copy required
#
print "Copying %s to %s" % ( srcFile, dstFile )
shutil.copyfile( srcFile, dstFile )
else:
handle, tempname = tempfile.mkstemp( '.wav', 'pyxcode', '/tmp'
)
os.close(handle)
decode = decoders[srcFmt]
encode = encoders[dstFmt]
if not decode:
shutil.copyfile( srcFile, tempname )
else:
decode( srcFile, tempname )
if not encode:
shutil.copyfile( tempname, dstFile )
else:
encode( tempname, dstFile, meta )
os.remove(tempname)
print "Converted %s" % os.path.basename( srcFile )
def Usage():
print """
exec.py --odir=path --fmt=[flac|wav|riff|ogg|mp3] [--best] file [file...]
"""
sys.exit(1)
if __name__ == "__main__":
files = []
odir = None
fmt = None
args = sys.argv[1:]
for arg in args:
if arg[0:2] == '--':
opt = arg[2:]
sep = opt.find('=')
if sep > 0:
key = opt[0:sep]
value = opt[sep+1:]
if key == 'odir':
odir = value
elif key == 'fmt':
if value == 'wav':
value = 'riff'
fmt = value
else:
Usage()
else:
if opt == 'best':
best = True
else:
Usage()
else:
files.append( arg )
if not odir or not fmt:
Usage()
for f in files:
if os.path.isfile( f ):
ftype = DetermineFileType( f )
if ftype:
meta = GetMetadata( ftype, f )
src = ( ftype, f )
dst = ( fmt, os.path.join( odir, NewName( f, meta, fmt ) ) )
ret = Transcode( src, dst, meta )
if ret:
print ret
6311
days inactive
6314
days old
linux-audio-user@lists.linuxaudio.org
12 comments
8 participants
participants (8)
-
Adam Sampson -
Arnold Krille -
Brian Dunn -
Ken Restivo -
Matt Price -
Paul Davis -
paul wisehart -
rob