Objective

To develop a prototype strategy for (highly flexible) sound morphing using granular synthesis.

Strategy

My first attempt makes use of diskgrain. Ideally, there is a smoother way to transition between the sounds than I have chosen here (probably by altering the density of the grains). However, I spent most of the evening learning to make the opcode work and how to set the parameters properly.

Materials

I used trumpet and flute sounds from the University of Iowa's Music Instrument Samples site. I cut out the same note on each instrument (F4), normalized each, then amplified to peak full scale.

Result

The opcode works as advertised, and the sounds apparently morph quite well. However, they could be smoother. Part of the problem is, of course, the volume of a flute is nowhere near that of a trumpet -- the juxtaposition of the two sounds seems somewhat jarring.

To smooth out some of the 'bumps', I made a slight variation in the grain sizes between the two instruments (38 ms and 42 ms).

Conclusion

My next attempt will make use of k-rate density alteration to morph between the two sounds. I will also the more sophisticated granular synthesis opcodes. The experience I've gained with this step will be valuable.