This is the nature of the music, I think, not an algorithmic problem. Increasing the rate with time stretching means the lower frequencies get less and less periods, and lower freqs make most of the amplitude.
Articles in this section
- Is there a compiled executable to test Superpowered time stretching?
- Does the time stretcher work in real-time too?
- How many channels are supported for time-stretch?
- The time stretcher is good for live, or only for offline processing?
- After I increase the rate of the whole song, increasing by 50% (1.5), there is an attenuation of about 6 db. This seems to be the same at a new rate of 1.75. This doesn't seem to occur when slowing the song down. Any ideas what's going on?
- I noticed that the time stretcher cannot stretch beyond 4x, as in if I pass in a parameter of 4.5, it will just get truncated to 4. I assume there is also a lower limit of 1/4. Is it possible to remove such limit, or extend it to 6 and 1/6?
- How much CPU the real-time stretching of a stereo stream (let’s say 44.1k) will take using your library? Do you think 5 is possible?
- How do I timestretch audio to fit a certain BPM? If I have a drum loop at 90BPM, how do I play it at 120 BPM without pitch changes?
- The positions in the player might be affected by the tempo change?
- I am using your SDK for iOS and I would like to change pitch of a sound seamlessly. At the moment it seems to start over or in some way reset it because the transition is not smooth.