- Meta Spark Studio
- Meta Spark Hub
- Features
- Learn
- Programs
- Blog
- Community
- Create AR effects for Instagram and Meta.Get Started
Vocoder Patch
Use the Vocoder patch to create robotic, synthesized-style sound effects. The vocoder patch works by filtering a carrier signal, usually musical content (an audio asset or Oscillator output) through a modulator signal. The modulator signal is usually a voice, i.e. the microphone.
Inputs
| Name | Description |
|---|---|
| Modulator | The audio source that you want to alter. Attach audio clip playback, microphone or other audio source. |
| Carrier | The audio source used as a carrier. |
| Mix | Effect dry/wet mix. Adding back some original audio can create some interesting effects. A mix of 0% will bypass the effect. 100% will output only the vocoded signal. |
| Flatten | If set true, the carrier gets "flattened", i.e. energy and the spectral envelope will be normalized to remove coloring caused by the carrier. The result will be closer to the modulator signal, in case of speech/singing that means better intelligibility and naturalness. Set this to true if you don't want the carrier to color the result. |
| VT Ratio | The modulator timbre (formants, not musical pitch) can be shifted up or down, resulting in monster/chipmunk sound in case of speech. |
| Bypass | Removes the sound effect from the audio source. |
Output
| Name | Description |
|---|---|
| Audio | The altered audio output. |
Example
In the example below we applied a robotic sound to the audio detected by the microphone, using an audio clip for the carrier signal.
This is what the audio effect sounds like.
We achieved this with the patch graph below:
In the Patch Editor, the Chord Loop audio clip represented by the orange patch is set up to play on a loop as soon as the effect is opened. Familiarize yourself with this setup in this article.
We then split the audio streams from the Microphone and Audio Player between two Vocoder patches. To do this, we connected:
- The patch representing the Microphone scene object to the Modulator input in each Vocoder patch.
- The Audio Player patch to the Carrier input in each Vocoder patch.
Splitting the audio between two vocoders lets you apply a slightly different vocoder effect to each stream of the modulator and carrier signal, before mixing the signals together again. For this effect, we selected the Flatten input in one Vocoder patch and left the input in the other patch unselected. This prevents the vocoder effect from overpowering the voice but it doesn’t obscure it entirely.
Finally we connected the:
- Vocoder patches to a separate Gain patch. We edited the Amount in each patch to 0.8, to lower the volume before mixing the signals back together.
- Gain patches to the Audio 1 and 2 inputs in the Audio Mixer patch.
- The Audio output of the Audio Mixer patch to a patch representing the Audio property of a speaker in the scene.
