[linux-audio-dev] idea: wavelets of variable frequencies

Hi! I just joined this list. I have an idea for a sound format which is similar to spectrum analysis (but better?): represent the sound as a synthesis of sinus waves where the frequencies are free and can change with time. Instead of doing a discrete fourier transform when reading a small frame of the sound, do a dense transform (every 0.1 Hz?) and pick out the peaks. Then assume that a similar enough frequency in the next frame comes from the same source, keep joining those and your sound will be represented by a lot of oscillators (wavelets) with amplitude and frequency curves. I guess this would overcome the weaknesses in having fixed frequencies. This idea shouldn't be new, where can I read about it? Has it been implemented anywhere? Thanks in advance, Tom Weber