gRainbow
size 3 MB / 11 MB
gRainbow is a synthesizer that uses pitch detection to choose candidates for granular synthesis or sampling.
It was created to overcome a few shortcomings of traditional granular synths.
Pitch variations in the input clip can produce inharmonic tones, which isn’t always wanted. This leads users to often use short single-pitch clips, restricting the synth to a single timbre. Manual pitch matching to the input clip is often required to produce the correct notes with MIDI input, which can be difficult and repetitive. Pitch shifting is commonly done with timestretching, which can create unwanted artifacts when shifting multiple octaves in either direction.
Instead, gRainbow prefers longer, pitch-diverse audio clips, automatically produces harmonics matched for MIDI input and avoids too much timestretching by generating harmonics that are already near their target pitch.
- Notes: one of 12 notes in the chromatic scale. gRainbow repeats its single octave over all MIDI notes, though that may change in the future using a more advanced pitch detection algorithm.
- Generators: create grains for a particular note at a particular position in the audio clip.
- Candidates: a collection of viable positions in the audio clip where the pitch is harmonic with a particular note.
- Parameters: just like usual ones, but organized a bit differently. Parameters use their global value by default, but can be overridden at both the note and generator level. The lowest-level parameter that is different from its default value is used, in the order of global->note->generator. In this way, you’ll only have to change the lower-level parameters when you need to without losing any flexibility.
When an audio file is loaded into gRainbow, a pitch detection analysis is ran to segment the clip into discrete pitches over time. Next, gRainbow finds up to 6 pitch-matched position candidates for each note, guaranteeing harmonically agreeable grains. Finally, the user customizes each note and its generators (i.e. changing parameters and candidates) to their liking, and can save everything into a preset to avoid processing things again next time.
https://bboettcher3.github.io/grainbow
gRainbow-Windows ( 3 MB )
gRainbow-macOS ( 11 MB )
gRainbow-Linux ( 16 MB )
