Re: [linux-audio-user] Filtering out sounds?

From: Robin Gareus <robin@email-addr-hidden>
Date: Sat Mar 10 2007 - 02:16:25 EET

terrence@email-addr-hidden wrote:
> I pretty much stick to CLI programs because my vision is limited
> and, though I can sort of see the screen if the print is big enough and/or
> I use a magnifier, speech output is much more efficient for my needs.
> I'll have to check out that anoi program and see how it works out.

maybe there are better suited algorithms in GUI software that one can
wrap into a CLI macro. and if you're into coding: Fons Adriaensen's
convolution code might come in handy: http://www.kokkinizita.net/linuxaudio/

You might be able to build an adaptive filter that keeps track of your
pitch with supercollider in text only. or use BIG pd boxes... :)

I dare say if you "know how to get there" the linux command line offers
plenty of opportunities to do the job! - personally I would need to
resort on graphical-sound analysis to tackle the problem.. I'll take off
my glasses and think again.

I have never tried to FFT-away a voice pattern. sometimes I adjust mixed
vocals or voices over a background track. and get pretty good results
with a series of small-band eq's in the 200-800 Hz range. and a wider
bandpass around 1.4k to trigger a compressor. You might want to add a
threshold/cut. - i usually hand edit the automation..

if it's a loooong tape. cut it.

I'd rather cut it anyway - you'll be much more satisfied if you keep the
unequalized, original voices if you want to remember their sound and
articulation. just trash the annoying parts of the conversation :)

robin

PS. off-topic.

I know a blind cabaret artist who does his own recordings incl. editing
and post-production with CoolEdit2000 under windows - according to him
the only graphical DAW that can be controlled by visually impaired with
a PC-keyboard and no extra (braille OR audio-mixer/MIDI) hardware!
he even automates some VST plugins, but I don't have any details. -
anyway he once asked me about the possibilities to build an open-source
DAW for blind users. eg. there are still issues remote controlling
ardour without visual feedback... while braille and speech seem to work
well under linux - it still seems not [yet] possible to set up a linux
Audio Workstation that just works without an eye on the screen. and even
worse: usually audio-hardware manufacturers have "flashy" websites and
the most tech-specs become invisible! the usual music-shops were
overburdened with his requests and high-end solutions are not
acceptable... he managed to research on and online-order a M-audio
Firewire and PCI Firewire Card for his purpose.
I'm not in regular contact with Vincent, but I said I'll drop him a line
when I stumble over something interesting..

forgive my ignorance. Is there a central good resource for headless
linux DAW setups? maybe we can start a category or wiki page on
http://apps.linuxaudio.org/ ?

There is a paper about using ecasound & braille from Julien Claassen,
but I have not followed the subject myself.

robin
Received on Sat Mar 10 04:15:02 2007

This archive was generated by hypermail 2.1.8 : Sat Mar 10 2007 - 04:15:02 EET