Hi all, I'm trying the SMSMorphing example in NetworkEditor to morph some sound samples. The output sounds turn out quite bad in quality (i.e. could hear two distinguish sounds from output, some artfacts was introduced, the intensity of the sound was reduced, etc). I'm wondering if I have something wrong with setting up the whole system or using the tools? Or it was made to be like that because it's just an example? How can I use the SMS tools provided by CLAM to produce high quality morphing sounds, or do I need to implement these features by my own? I've tried to read the source code and the API documents, but still not quite understand how the morphing algorithm works? Can someone tell me where can I find detailed explanation on the SMS Morphing algorithm or implementation (i.e. mathematical expression, pseudo code of the core algo, etc)?
b.t.w. I don't know if it is a bug, but my NetworkEditor occasionally crashed when I was playing the SMSMorphing examples. For the same network, it would play normally for the first time, but the second time when I press the play button, it crashed with a segmentation fault, and I had to restart the NetworkEditor again. It was not like that right after I just installed CLAM, but happened these two days. :-( Thanks very much for helping. If these doubts about morphing were cleared , I would like to use CLAM in one of my school project. ;-p Regards, Zhijia _________________________________________________________________ 手机也能上 MSN 聊天了，快来试试吧！ http://mobile.msn.com.cn/ _______________________________________________ CLAM mailing list [email protected] http://clam.iua.upf.edu
