I’ve been trying to use the vocoder plugin to make a robot voice effect.
I’ve split the vocal stereo track, put the vocals on one left channel, put a generated sinewave tone (with a frequency of 440 Hz and amplitude of 0.8) into the right channel, made both into a stereo track.
When I click “OK” in the vocoder window, the progress bar loads halfway before a Nyquist dialog pops up telling me “Nyquist did not return audio” without affecting the track.
When I click “Debug,” I’m given the Nyquist dialog again, and when I click “OK” I get a text dialog with the following:
error: division by zero
Function: #<Subr-/: #566d8f8>
Arguments:
1
0
Function: #<FSubr-SETF: #56697dc>
Arguments:
S
(SCALE (/ 1 PEAKAMP) S)
Function: #<FSubr-COND: #566bca8>
Arguments:
((= MST 1) (VECTOR (AREF S 0) (SETF (AREF S 1) (SCALE (/ 1 PEAKAMP) (AREF S 1)))))
((= MST 0) (SETF S (SCALE (/ 1 PEAKAMP) S)))
Function: #<FSubr-LET: #5669944>
Arguments:
NIL
(COND ((= MST 1) (VECTOR (AREF S 0) (SETF (AREF S 1) (VOCODER)))) ((= MST 0) (SETF S (VOCODER))))
(COND ((= MST 1) (SETQ PEAKAMP (PEAK (AREF S 1) NY:ALL))) ((= MST 0) (SETQ PEAKAMP (PEAK S NY:ALL))))
(COND ((= MST 1) (VECTOR (AREF S 0) (SETF (AREF S 1) (SCALE (/ 1 PEAKAMP) (AREF S 1))))) ((= MST 0) (SETF S (SCALE (/ 1 PEAKAMP) S))))
S
Function: #<FSubr-IF: #566cfc8>
Arguments:
(ARRAYP S)
(LET NIL (COND ((= MST 1) (VECTOR (AREF S 0) (SETF (AREF S 1) (VOCODER)))) ((= MST 0) (SETF S (VOCODER)))) (COND ((= MST 1) (SETQ PEAKAMP (PEAK (AREF S 1) NY:ALL))) ((= MST 0) (SETQ PEAKAMP (PEAK S NY:ALL)))) (COND ((= MST 1) (VECTOR (AREF S 0) (SETF (AREF S 1) (SCALE (/ 1 PEAKAMP) (AREF S 1))))) ((= MST 0) (SETF S (SCALE (/ 1 PEAKAMP) S)))) S)
1>
I am using version 2.0.2 which I got with the .dmg installer running on OSX 10.8.4
The settings I used in the vocoder window are as follows:
Distance: 20.0
Output: both channels
Vocoder Bands: 68
Amplitude of original audio: 0%
Amplitude of white noise: 0%
Amplitude of Radar Needles: 0%
Frequency of Radar Needles: 77.0 Hz
The vocoder plugin was included in the build; I didn’t download any other plugins not included in the download package.
I would have solved the problem by downloading the 2.0.4 version, however I do not have admin privileges to install new software. If this is the way to fix it, so be it, but any solution that does not include reinstalling all of Audacity all over again would be preferred and much appreciated.
Thanks for replying so quickly.
Now when I set any value to a number greater than 0%, with output into both channels, the result is silence in both left and right channels. When I set the output to right channel only, the right channel is silenced with no change in the left channel.
The waveform doesn’t look like normal silence, though. Here’s what it looks like:
I need to leave for a while, but all the tool is doing is multiplying the two sound values. Is your voice normal level? If it’s really low, then the output performance will vanish. Use my MP3 voice as the test track and use Generate > Pink Noise 0.8 amplitude as the carrier track.
Were you able to find the instructions? There’s not much there because it usually just works.
That topic contains the before and after sound clips of the work I did for that poster. Use that voice (I know we tell people not to use MP3 for anything) and Generate > Pink Noise for the other side. Use the numbers I posted for you.
It should look like the illustration just before you apply the effect. It should sound close to RobotVoice.mp3 when you get done. Let us know how it works.
It’s possible you ran out of memory or hard drive space and everything else is working just fine.
Thanks a lot for your help! Everything worked as planned and the results were similar to what you had in the other thread.
I think I found out what I was doing wrong: the voice sample, although it’s volume was normal, had a rate of 22.05 kHz when the carrier channel and the project rates were both 44.1kHz. Simple fix.
If the sample rates do not match within a stereo track, the following error message is returned:
“Sorry, cannot apply effect on stereo tracks where the tracks don’t match.”.
The plug-in code won’t be executed and the debug screen will be empty.
Solution: mix and render the stereo track (track menu)
In contrast, your first error was produced because one stereo track was empty and the peak 0.
So, always use ‘Make Stereo Track’ with non-empty mono tracks.
You can even duplicate the original voice.
There’s a special carrier already integrated into the effect, namely a buzz sound (“Radar Needles”).
Try the following:
Distance 1
240 Bands
all amplitudes 0, except Radar Needles 100 %.
Frequency 80 Hz.
I trust you know this emotion-less robot voice from dozens of movies.
