(https://www.audiotool.com/sample/quantization_needed_example) I imagine the detection could work similar to the "trim silence" function, where it finds points where it hits a certain threshold, and removes audio/adds silence to make it fit to the nearest step of the selected subdivision, then moves to the next point where it passes the threshold. (more rough ideas)
I understand this request (and the explanations below) to mean to move audio events to specific times (fit events to specific beats and subdivisions), so as to correct the -timing- of a sloppy performance, not to mean to correct pitches that are off (too flat/ too sharp).
Comments (7)
You mean it should determine the pitch? Where and how shall this information be used then?
Or do you mean something different?
(very rough idea here) Like it finds the start of a sound in an input, and lines it up with the closest step of a chosen time/beat step length
Sounds non-trivial. Do you have an example for a sample you want to quantize?
Would it be enough to have an adjustable metronome (bpm, signature, offset) to check if your metadata is correct?
(maybe there could be function to "auto-tune/guess" that metronome)
(https://www.audiotool.com/sample/quantization_needed_example) I imagine the detection could work similar to the "trim silence" function, where it finds points where it hits a certain threshold, and removes audio/adds silence to make it fit to the nearest step of the selected subdivision, then moves to the next point where it passes the threshold. (more rough ideas)
So basically something like autotune but for beats/time?
Our editor addresses basic editing and I guess your request would be too ambitious.
Precisely, and thats fair enough
I understand this request (and the explanations below) to mean to move audio events to specific times (fit events to specific beats and subdivisions), so as to correct the -timing- of a sloppy performance, not to mean to correct pitches that are off (too flat/ too sharp).