25 Dec
2010
25 Dec
'10
2:57 p.m.
there doesn't seam to exist any compressor/plugin for Linux that does
it?
25 Dec
25 Dec
3:41 p.m.
Fabio wrote:
there doesn't seam to exist any compressor/plugin
for Linux that does
it?
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
Well it's easy enough to set up in Ardour using a bus and a regular
compressor. I expect it should be easy enough in other software.
Q
3:57 p.m.
Em Sat, 25 Dec 2010 14:41:19 +0000
Q <lists(a)quirq.ukfsn.org> escreveu:
Fabio wrote:
how you do that? none of the compressors i checked has ability to use
negative ratio.
upwards compression means it elevates volumes below threshold instead
of the "opposite"(lack of word) in normal compression
there doesn't seam to exist any
compressor/plugin for Linux that
does it?
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
Well it's easy enough to set up in Ardour using a bus and a regular
compressor. I expect it should be easy enough in other software.
Q 
4:36 p.m.
On Sat, Dec 25, 2010 at 6:57 AM, Fabio <capoeirista(a)arcor.de> wrote:
Em Sat, 25 Dec 2010 14:41:19 +0000
Q <lists(a)quirq.ukfsn.org> escreveu:
A bit wonky but:
1) Find some compressor code and do some mods, or
2) Maybe do it mathematically. Original wave file - compressed wave
file = new interesting wave file?
I do #2 all the time with stock market data. Seems like you should be
able to do it with sound data in sox or something.
Cheers,
Mark
Fabio wrote:
how you do that? none of the compressors i checked has ability to use
negative ratio.
upwards compression means it elevates volumes below threshold instead
of the "opposite"(lack of word) in normal compression there doesn't seam to exist any
compressor/plugin for Linux that
does it?
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
Well it's easy enough to set up in Ardour using a bus and a regular
compressor. I expect it should be easy enough in other software.
Q 
4:47 p.m.
On 12/25/2010 03:57 PM, Fabio wrote:
Em Sat, 25 Dec 2010 14:41:19 +0000
Q <lists(a)quirq.ukfsn.org> escreveu:
you can set up two busses, both fed from the signal you want to
up-compress. one is left as-is, and the other is set for compression at
the threshold below which you want the up-compression to set in.
now you mix both buses to the same master. the effect is that at low
levels, both signal paths will be uncompressed and add up. at higher
levels, the compressed signal path is attenuated and adds less gain,
which is the effect you're after.
there's a million variations of this scheme, you can do the same with
expanders.
at some point, sampo savolainen was working on a really nice compressor
that includes a parallel signal path - i don't know if it was ever
released, but it might be somewhere in his svn...
Well it's easy enough to set up in Ardour
using a bus and a regular
compressor. I expect it should be easy enough in other software.
Q
how you do that? none of the compressors i checked has ability to use
negative ratio.
upwards compression means it elevates volumes below threshold instead
of the "opposite"(lack of word) in normal compression
5:13 p.m.
Em Sat, 25 Dec 2010 16:47:04 +0100
Jörn Nettingsmeier <nettings(a)folkwang-hochschule.de> escreveu:
On 12/25/2010 03:57 PM, Fabio wrote:
wow, audio-production is math, too...lol
but man, this will result in a normal compression, only that the ratio
will be lowered.
I mean, lets say we have 4 signals with values: 1 2 4 8
threshold between 2 and 4 give for ex. this: 1 2 3 4
now i add both an divide by 2: 1 2 3.5 6
Em Sat, 25 Dec 2010 14:41:19 +0000
Q <lists(a)quirq.ukfsn.org> escreveu:
you can set up two busses, both fed from the signal you want to
up-compress. one is left as-is, and the other is set for compression
at the threshold below which you want the up-compression to set in.
now you mix both buses to the same master. the effect is that at low
levels, both signal paths will be uncompressed and add up. at higher
levels, the compressed signal path is attenuated and adds less gain,
which is the effect you're after.
Well it's easy enough to set up in Ardour
using a bus and a
regular compressor. I expect it should be easy enough in other
software.
Q
how you do that? none of the compressors i checked has ability to
use negative ratio.
upwards compression means it elevates volumes below threshold
instead of the "opposite"(lack of word) in normal compression
6:10 p.m.
Fabio wrote:
Em Sat, 25 Dec 2010 16:47:04 +0100
Jörn Nettingsmeier <nettings(a)folkwang-hochschule.de> escreveu:
wow, audio-production is math, too...lol
but man, this will result in a normal compression, only that the ratio
will be lowered.
I mean, lets say we have 4 signals with values: 1 2 4 8
threshold between 2 and 4 give for ex. this: 1 2 3 4
now i add both an divide by 2: 1 2 3.5 6
Compression reduces the dynamic range. Ordinary or "normal" compression
does it by affecting the loudest signals, reducing them in level,
upwards compression does it by boosting the volume of the quietest
signals -- compressing from the bottom up. This parallel processing
produces upwards compression as Joern expounded on my original statement.
You mention a negative ratio: are you perhaps thinking of upward expansion?
Q
how you
do that? none of the compressors i checked has ability to
use negative ratio.
upwards compression means it elevates volumes below threshold
instead of the "opposite"(lack of word) in normal compression
you can
set up two busses, both fed from the signal you want to
up-compress. one is left as-is, and the other is set for compression
at the threshold below which you want the up-compression to set in.
now you mix both buses to the same master. the effect is that at low
levels, both signal paths will be uncompressed and add up. at higher
levels, the compressed signal path is attenuated and adds less gain,
which is the effect you're after.
6:40 p.m.
Em Sat, 25 Dec 2010 17:09:54 +0000
Q <lists(a)quirq.ukfsn.org> escreveu:
upwards compression does it by boosting the volume of
the quietest
signals -- compressing from the bottom up. This parallel processing
produces upwards compression as Joern expounded on my original
statement
ok, thats what i want
perhaps my math is to weak...lol
I simply will try it out.
just giving more info: amateur-singer, half an hour of time, lives long
from here. we recorded her voice, but in the lower parts she sang very
quiet, so with normal compression i would compress 95% of what is
correct only to add more gain to those parts. so i thought about this
option instead. probably i should simply elevate volume in those part
instead.
7:48 p.m.
It seems like you would want to do normal compression with a very low
threshold (about the noise level) and a negative ratio.
then you end up with a curve like
\
/ \
/ \
Then you can add that into your original signal, to get something
approximating "upward compression".
Also, perhaps the expander from tap-plugins might do what you want:
http://tap-plugins.sourceforge.net/ladspa/dynamics.html
<http://tap-plugins.sourceforge.net/ladspa/dynamics.html>Jeremy
On Sat, Dec 25, 2010 at 12:40 PM, Fabio <capoeirista(a)arcor.de> wrote:
Em Sat, 25 Dec 2010 17:09:54 +0000
Q <lists(a)quirq.ukfsn.org> escreveu:
upwards compression does it by boosting the
volume of the quietest
signals -- compressing from the bottom up. This parallel processing
produces upwards compression as Joern expounded on my original
statement
ok, thats what i want
perhaps my math is to weak...lol
I simply will try it out.
just giving more info: amateur-singer, half an hour of time, lives long
from here. we recorded her voice, but in the lower parts she sang very
quiet, so with normal compression i would compress 95% of what is
correct only to add more gain to those parts. so i thought about this
option instead. probably i should simply elevate volume in those part
instead.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
9:01 p.m.
Em Sat, 25 Dec 2010 13:48:07 -0500
Jeremy <jeremybubs(a)gmail.com> escreveu:
It seems like you would want to do normal compression
with a very low
threshold (about the noise level) and a negative ratio.
then you end up with a curve like
\
/ \
/ \
Then you can add that into your original signal, to get something
approximating "upward compression".
Also, perhaps the expander from tap-plugins might do what you want:
http://tap-plugins.sourceforge.net/ladspa/dynamics.html
<http://tap-plugins.sourceforge.net/ladspa/dynamics.html>Jeremy
On Sat, Dec 25, 2010 at 12:40 PM, Fabio <capoeirista(a)arcor.de> wrote:
> Em Sat, 25 Dec 2010 17:09:54 +0000
> Q <lists(a)quirq.ukfsn.org> escreveu:
>
> > upwards compression does it by boosting the volume of the quietest
> > signals -- compressing from the bottom up. This parallel
> > processing produces upwards compression as Joern expounded on my
> > original statement
>
> ok, thats what i want
>
> perhaps my math is to weak...lol
>
> I simply will try it out.
>
> just giving more info: amateur-singer, half an hour of time, lives
> long from here. we recorded her voice, but in the lower parts she
> sang very quiet, so with normal compression i would compress 95% of
> what is correct only to add more gain to those parts. so i thought
> about this option instead. probably i should simply elevate volume
> in those part instead.
> _______________________________________________
> Linux-audio-user mailing list
> Linux-audio-user(a)lists.linuxaudio.org
> http://lists.linuxaudio.org/listinfo/linux-audio-user
>
that's not possible
curve has to be like this: (hope this will be look the way it looks like
here)
.
.
.
.
.
.
.
but like i said, none of the compressors i have installed can do this
(and I have a lot)
11:33 p.m.
On Saturday 25 December 2010 21:01:28 Fabio wrote:
curve has to be like this: (hope this will be look the
way it looks like
here)
.
.
.
.
.
.
.
but like i said, none of the compressors i have installed can do this
(and I have a lot)
Are you looking for a noise-gate without a hard gate? I think that would give
a level response like the one above...
Have fun,
Arnold
26 Dec
26 Dec
1:13 a.m.
On Sat, Dec 25, 2010 at 11:33:10PM +0100, Arnold Krille wrote:
Are you looking for a noise-gate without a hard gate?
I think that would give
a level response like the one above...
No, a soft gate would be more like
*
*
*
*
*
*
with the upper half being 1:1
The scheme Joern described will do the trick, producing
something like (idealised)
*
*
*
*
*
*
*
with the middle region being the compressed one.
The ratio has to revert to 1:1 for very low levels,
otherwise the gain would go to infinity.
Ciao,
--
FA
There are three of them, and Alleline.
5134
days inactive
5135
days old
linux-audio-user@lists.linuxaudio.org
11 comments
7 participants
participants (7)
-
Arnold Krille -
Fabio -
fons＠kokkinizita.net -
Jeremy -
Jörn Nettingsmeier -
Mark Knecht -
Q