Hi ,
I am trying to generate mandarin using hts demo.My database contains 1000 clear 48KHz wav files.When I use SPTK(1 mix) synthesize a new sentence ,there is background noise.And I use STRAIGHT40 as vocoder ,the noise is still exist. Attachment wav is the sythesized voice.
I write questions my own ,is there anything wrong with my question files?
Or could you give me some advice to make synthesized voice better?
thanks and regards