Vibrato

[steve]

Applies a variable vibrato to the selected audio.

Limitations:
Maximum length is 99,999 samples (per audio channel).
This limitation is hard coded into Audacity as the maximum size for *table* is 100000 samples.

Thanks to Edgar-rtf for the suggestion of using (fmosc), which in spite of the length limitation provides an elegant solution.
To overcome the length limitation, the function (snd-compose) could be used - this function uses a sound as a look-up table rather than using *table*.

Updated version: vibrato.ny

[Irish]

Interesting effect, Steve.

I just took a quick look at it, and tried to get a slow vibrato (approx 1 to 2 Hz), but entering 1.0 for Initial and Final speeds gave a result that sounded like about 10Hz, and entering anything less than 1.0 didn't seem to make any difference.

Sorry I can't be more constructive, but I haven't yet got my head around pwlv and fmosc. Maybe later ...

POL

[steve]

I just took a quick look at it, and tried to get a slow vibrato (approx 1 to 2 Hz), but entering 1.0 for Initial and Final speeds gave a result that sounded like about 10Hz, and entering anything less than 1.0 didn't seem to make any difference

Thanks for checking it out, and you're correct, the vibrato speed is too fast.
This update should fix it.

vibrato.ny

(1.7 KiB) Downloaded 231 times

Re. (pwlv), have a play with this code in the Nyquist Prompt:

Code: Select all

(setq L1 0.5)
(setq L2 1.0)
(setq L3 0.0)
(setq L4  0.5)

(setq T2 0.33)
(setq T3 0.67)
(setq T4 1.0)

(control-srate-abs *sound-srate*
   (pwlv L1 T2 L2 T3 L3 T4 L4)
)


The function (control-srate-abs) srate beh) forces beh to be calculated at a sample rate of srate.
In this case, the sample rate is set to *sound-srate*, which is the sample rate of the selected audio.
The reason for having that function here is that envelopes are normally calculated at the control rate, which is normally much lower than the sample rate of the track and so the output of pwlv would be shrunk down to a tiny length when it is returned to the audio track. By forcing pwlv to be calculated at the sound sample rate, the output from pwlv will be such that a time value of 1.0 is equal to the length of the selection. (the time values are relative time values).