[hts-users:00674] LPC coefficient for syntheis
- Subject: [hts-users:00674] LPC coefficient for syntheis
- From: Fang Li <lifang3cis@xxxxxxxxxxxx>
- Date: Thu, 17 May 2007 15:59:37 +0800 (CST)
- Delivered-to: hts-users@xxxxxxxxxxxxxxx
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com.cn; h=X-YMail-OSG:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding:Message-ID; b=BW//OGk6dzZQzOTu0XtHX7sy9IVHx+CAGp6DhfThCh9rEnz4+0s4eQ7+uZBXDExxST8pSECkEhn6P8SCH3swno7taB4Tr4EB4cliB6JUAnxB3AARFgT9ZIFn1pT+lnl3Kpst6UMjD62vCsRSj5CGCchDAk6lz8/RHn6O3wKut+E=;
Hi Heiga
Thanks for your reply.
I just leaned that LSP synthesis showed better performance than MCEP synthesis.
But I found it sounds a little noisy.So now I want to use LPC coefficient for both training and synthesis in HTS-2.0.
I made LPC coefficient from raw, merged it with lf0 to generate cmp file, trained them and it generated .pit and .lf0 file.
When generate wav file, I didn't use Postfiltering( ) function in scripts/Training.pl .
Finally the .wav file was generated without voice.
I know little about the speech signal processing.
Are there other modifications before I use LPC coefficient for synthesis?
Best regards
Lifang
抢注雅虎免费邮箱3.5G容量，20M附件！
- Follow-Ups
-
- [hts-users:00675] Re: LPC coefficient for syntheis, Heiga ZEN (Byung Ha CHUN)