[chuck-users] How can I filter silence?

[Jim Hinds]

Kas, Graham!

Ya gotta love the way you think out of the box.  Break the rules to  
make better ones!  Thanks for the comments, code and inspiration.

Jim
---"I dream of cheese, It's toasted, mostly"  Ben Gunn, Treasure Island.

On Sep 24, 2006, at 8:55 09PM, Graham Coleman <gc at gehennom.net> wrote:

>
> From: Graham Coleman <gc at gehennom.net>
> Date: September 24, 2006 9:50:44 AM HST
> To: ChucK Users Mailing List <chuck-users at lists.cs.princeton.edu>
> Cc: jahbin at romatictrances.com
> Subject: Re: [chuck-users] How can I filter silence?
> Reply-To: ChucK Users Mailing List <chuck- 
> users at lists.cs.princeton.edu>
>
>
> I came up with some cheesy code to implement the idea.
> It speeds up the tape when it hits silence. You'll need to
> tune the time constants and the threshold for best results:
>
> sndbuf buf => dac;
> buf.read("speak/wave.wav");
> buf.pos( 0 );
>
> 0.05 => float tresh;
> 0.01 => float rt; //release const
> 0.4 => float at; //attack const
> float xd;
>
> while(true) {
>
>   math.fabs( buf.last() )-xd => float a; //amp diff
>   if ( a < 0 ) 0=>a; //non-negative
>   xd*(1-rt)+at*a => xd; //the attack/release exponential filter
>
>   if ( xd<tresh ) {
>     buf.pos( buf.pos()+6 ); //skip ahead
>   }
>
>   1::samp=>now;
> }
>
> To compute the time constants, use this formula:
> //AT = 1 - e ^ (-2.2T/t<AT)
>
> or this if t is in samples:
>
> time const = 1.0 - exp( -2.2 / t );
>
> best,
>
> Graham
>
> On Sun, 24 Sep 2006, Graham Coleman wrote:
>
>> I'm not sure if there's anything in chuck that will currently do  
>> this,
>> but I think you could write something like a noise gate, that  
>> tracks the
>> amplitude of samples coming from a SndBuf and skips ahead until  
>> the level
>> rises above a threshold.
>>
>> Not sure how well this would run in real time.
>>
>> Better, would be something that preprocesses the file with in/out  
>> points
>> so the skips are quick.
>>
>> Graham
>>
>> On Sun, 24 Sep 2006, Jim Hinds wrote:
>>
>>> I want to scan some audio for dropped out sections.  It's long and
>>> talky with 'significant' pauses between question and answers.  These
>>> sections are filled with background noise, and often several seconds
>>> long.  I'd like to shorten these empty sections (there are a lot of
>>> them).
>>>
>>> I'd like to keep the files more or less intact (not sliced up into a
>>> zillion intermediate files)
>>>
>>> Is there a unit in chuck that will be able to delete samples as they
>>> pass through?
>>>
>>> Jim
>>> _______________________________________________
>>> chuck-users mailing list
>>> chuck-users at lists.cs.princeton.edu
>>> https://lists.cs.princeton.edu/mailman/listinfo/chuck-users
>>>
>> _______________________________________________
>> chuck-users mailing list
>> chuck-users at lists.cs.princeton.edu
>> https://lists.cs.princeton.edu/mailman/listinfo/chuck-users
>>
>
>
>
>
> From: Kassen <signal.automatique at gmail.com>
> Date: September 24, 2006 10:23:58 AM HST
> To: "ChucK Users Mailing List" <chuck-users at lists.cs.princeton.edu>
> Subject: Re: [chuck-users] How can I filter silence?
> Reply-To: ChucK Users Mailing List <chuck- 
> users at lists.cs.princeton.edu>
>
>
>
>
> On 9/24/06, Graham Coleman <gc at gehennom.net> wrote:
> I'm not sure if there's anything in chuck that will currently do this,
> but I think you could write something like a noise gate, that  
> tracks the
> amplitude of samples coming from a SndBuf and skips ahead until the  
> level
> rises above a threshold.
>
> Not sure how well this would run in real time.
>
> Better, would be something that preprocesses the file with in/out  
> points
> so the skips are quick.
>
>
> But.... It doesn't look like it needs to be realtime.
>
> I'd think about playing the file once and storing the sample  
> locations where the volume drops below a treshold (as well as the  
> ones where it comes back) in a array, then playing it a second time  
> and skipping those sections while writing the results to a new  
> file. in "silent" mode this should be doable faster then realtime.
>
> You could also make a array of the length of the file (in samples)  
> and only write buf.last() => array[sample++] as long as it's over  
> the treshold. Once the buffer reaches it's end you'd playback the  
> array again and record that, taking care to stop after the highest  
> value that "sample" got to in the first pass. This would be CPU  
> intesive but it's quite clean as a method, I think.
>
> It all stands or falls with the quality of the envelope follower.
>
> I've been thinking about doing something similar to the first  
> option and pre-scanning wavefiles for positive zero-crossings and  
> storing those in a array. This would take extra time when loading a  
> file but from then on you could do very clean loops, for example to  
> loop just a few cycles. I wonder how samplers that have a "snap to  
> zero crossings" setting do this. It might still be quite tricky  
> because not all samplers that do this do it all that well....
>
> Hope that's of some use.
> Kas.
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.princeton.edu/pipermail/chuck-users/attachments/20060925/bbc43732/attachment.htm