Alvin@cup.portal.com (Alvin Henry White) (07/09/90)
Recent discussion has included books on massageing waveforms. I am interested in the ability to change time without changing pitch. This then allow one to set a locally made tape of a poetic reading of a commercial such as "Eat at Joe's diner, on the corner of nth and etc." to the prerecorded classical or other musical background. In the discussion of a book by F. Richard Moore called ELEMENTS OF COMPUTER MUSIC I think I heard mention of pitch change without time change. If anyone knows if time change without pitch change is also contimplated I would like to know. Also other source. It seems someone also mentioned that it might take sixty seconds of large scale number crunching to get one second of sound. Parallel processing may improve this and some of the supercomputer centers might need extra work. Imagin patching a duet of Michael Gorbechev and Ronald Regan singing what a friend we have in jesus or the "international."
jk87377@metso.tut.fi (Kouhia Juhana Krister) (07/10/90)
In article <31559@cup.portal.com> Alvin@cup.portal.com (Alvin Henry White) writes: >Recent discussion has included books on massageing waveforms. I am >interested in the ability to change time without changing pitch. This 1. Change pitch without changing time 2. Change samplerate or stretch the sound Example: 1. Double pitch with pitchshifter (2*) 2. Double lenght of sound (2*) Now Your sound is 2* longer and pitch is same than in original. Juhana Kouhia jk87377@tut.fi
jthornto@fs1.ee.ubc.ca (THORNTON JOHAN A) (07/10/90)
Changing pitch independent of time, and changing time independent of pitch
are essentially the same process. In the digital domain, they are the same
process.
------- _/__/ -----------------------------------------------------
_| ___| E l e c t r i c a l | Johan Thornton, Esq.
| | |_/ E n g i n E E r i n g |-------------------------
|/| __| U n i v e r s i t y | jthornto@fs1.ee.ubc.ca
|-| |/__ o f B r i t i s h |-------------------------
| |_____| C o l u m b i a | This space for rent
---- |__|/_| ------------------------------------------------------dt@yenta.alb.nm.us (David B. Thomas) (07/10/90)
Check your friendly radio shack catalog. There are a couple of tape recorders with the capability to play back a recording fast, but compensate the pitch. They call it VSC (Variable Speech Control) or something like that. The way it works is it samples the sound, chops hunks out, then plays back what's left slower. Since the tape is already going extra fast, what you can get is correct pitch but faster time. Naturally, the sound quality is poor, because of the chopping, and because it is a radio shack tape recorder :-) David
srmjr@cbnewsd.att.com (samuel.r.mullins) (07/10/90)
There is a wealth of literature concerning DSP algorithms for Time-Scale Modification of speech. I can't say for sure but I would guess that some of the algorithms would not work well on music. Anyway for a representative sample of articles see Part Four of: Speech Enhancement, Edited by Jae S. Lim, Prentice-Hall Signal Processing Series, 1983. ISBN 0-13-829705-3 In particular, the algorithm (based on modification of the Short-Time Fourier Transform) presented by M. R. Portnoff gives very high quality results (at least for speech.) If you're not comfortable with DSP, don't bother (mostly IEEE Journal articles) Sam Mullins -- "What a strange, strange boy. He sees the cars as sets of waves. Sequences of mass and space. He sees the damage in my face." - Joni Mitchell
smithj@hpsad.HP.COM (Jim Smith) (07/11/90)
The cleanest digital algorithm to do what you want is called a Phase Vocoder, but it is really computationally intensive, as you mentioned. Here are a couple of good references: The Use of the Phase Vocoder in Computer Music Applications James A. Moorer Journal of the Audio Engineering Society Jan/Feb 1978, V26 N1/2, pp42-45 Implementation of the Digital Phase Vocoder using the Fast Fourier Transform Michael R. Portnoff IEEE Transactions on Acoustics, Speech, and Signal Processing Vol. ASSP-24, No 3, June 1976 And probably the best first article on the topic; The Phase Vocoder: A Tutorial Mark Dolson Computer Music Journal, Vol.10, No.4, Winter 1986 pp14-27 Mark was with the computer music department at UCSD, and perhaps they would have reprints.