Re: [Audacity-nyquist] Reducing the volume on part of a sound.

[Dominic M. <do...@au...>]

Hi,

Here's how Nyquist works within Audacity: when you execute a plug-in function, 
the current selected audio is stored in the variable "s".  All of your Nyquist 
code is executed line-by-line, and the result of the final expression is what is 
returned to Audacity - if it's audio, that audio will replace your previous 
selection.

A defun doesn't actually _do_ anything, it just defines a function.  So what you 
need to do is add a line at the bottom that actually applies the function to s, 
for example:

(defun changeVol (s x y z)
   ...
)

(changeVol s 0.1 0.3 0.5)


In place of 0.1, 0.3, 0.5, you could stick control parameters, of course.

I hope that makes sense.

- Dominic


RACHAEL MAREE,HUNT wrote:
> Hi,
> 
> thanks for your suggestions. I have put them in a simple function, but I am unable to get the function to work... How does Audacity call a nyquist function in a plugin? I am currently using Audacity 1.2.3. 
> 
> Here's the psuedo-code version:
> 
> ; audacity header
> ; three controls with variables volR (int for percentage), volLen (real for seconds), startTime (real for seconds)
> 
> (defun changeVol (s volR volLen startTime) 
>   ( ; setting duration, start and end time as percentages, and volume to a "real" percent
>      (mult (pwl 0 1 start 1 start volume end volume end 1 1) s)
>   )
> )
> 
> When I run the code version of this it doesn't return any audio. So I thought that I had to call the function. I know that audacity treats any line in the plugin as a command. So I tried adding the following after defining the function:
> (changeVol (s volR volLen startTime)
> but still no audio was returned...
> 
> Going back to grass-roots, I tried creating a simple scale plugin.
> 
> ;audacity headers
> ;a control with variable scaleVol (int for percentage)
> 
> (defun scaleVol (s scaleVol)
>   ((scale-db (linear-to-db (/ scaleVol 100)) s))
> )
> 
> Which still returned no audio. What am I missing?
> 
> Thanks again!
> 
> Rach'
> 
> -----Original Message-----
> From: aud...@li... on behalf of David R. Sky
> Sent: Fri 2/3/2006 6:55 PM
> To: aud...@li...
> Subject: Re: [Audacity-nyquist] Reducing the volume on part of a sound.
>  
> Hi Rachael,
> 
> For this, you'd first need to determine the length (duration) of
> your sample in seconds, which you can only do in Audacity 1.2.3 and
> later. The global variable len gives you the number of samples in
> your selection, then you need to calculate time duration by
> dividing len by the sample rate:
> 
> (setf dur (/ len *sound-srate*))
> 
> Rachael:   * I can use the (loud) function to reduce the volume,
> and (scale-db), but neither seem to work in percentages. Is there
> a way to find how loud a sound is in decibels?
> 
> David: For what you want to do, finding the volume isn't necessary.
> To reduce your full selection a certain percent (eg. 20%), all
> you'd need to do is
> 
> (mult 0.8 s)
> 
> or
> 
> (scale 0.8 s)
> 
> or
> 
> (scale-db (linear-to-db 0.8) s)
> 
> The above works for mono and stereo audio in Audacity (not all
> Nyquist functions do but there are ways around this).
> 
> Rachael:  * How do I reduce the volume of a part of a sound?
> 
> David: I'd use a pwl (piece wise linear) function, which is a
> series of time/level pairs:
> 
> (mult (pwl 0 1 start 1 start volume end volume end 1 1) s)
> 
> You need to be very specific about what level at what time. If for
> example you left out
> 
> start 1
> 
> then instead of staying at constant volume until your start time,
> your audio would gradually reduce in volume from time 0 to start
> time.
> 
> When you've made a selection in Audacity, pwl works in percentages
> of the selection duration not time in seconds, so you'd need to
> calculate the start and end times as percentages of duration dur.
> 
> Even though it's imprecise, it doesn't appear to matter whether the
> duration of your volume reduction is greater than the length of
> your selection: the volume stays at the reduced level until the end
> of the selection.
> 
> If you have abrupt jumps in volume as I've done with pwl above,
> you'll notice a click at the start and end times, you can use a
> short fade time.
> 
> Hope this helps
> 
> David
> 
> On Fri, 3 Feb 2006, RACHAEL MAREE,HUNT wrote:
> 
>> Hi,
>>
>> Before subscribing to the list I sent this email to the list and was "awaiting moderation"... I'm a bit impatient, and joined a couple of days later. Here's hoping that I don't get a double post.
>>
>> I am a beginner to Nyquist / Lisp. That said, I am attemping to write an Audacity "filter" / plugin. (Hehe!)
>>
>> Ideally I would like to select 1-2 seconds of stereo audio in Audacity, tell the plugin a start time (within the sample i.e. 0.5 seconds in), a volume reduction in %, and a reduction length. The plugin would check whether the start time and reduction length fall within the bounds of the sample then reduce only that section of sound.
>>
>> Is this even possible?
>>
>> Now, I know that Audacity stores the selected sound in the s variable, and that I'd have to write a function to do this. However, there are a few things I don't know...
>>  * I can use the (loud) function to reduce the volume, and (scale-db), but neither seem to work in percentages. Is there a way to find how loud a sound is in decibels?
>>  * How do I reduce the volume of a part of a sound? Do I have to change it to an array / vector? If so, what is the relationship between array members and time?
>>
>> Thanks in advance for your help!
>>
>> Rachael.
>>
>>
>