home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!well!moon!rds!doug
- From: doug@rds.com (Doug Kaye)
- Newsgroups: comp.speech
- Subject: VSELP vs. CELP 3.2
- Message-ID: <5mq2TB1w164w@rds.com>
- Date: Tue, 10 Nov 92 15:19:27 PST
- Distribution: world
- Organization: Rational Data Systems, San Rafael, CA
- Lines: 29
-
- We've had a very positive reaction to our voice annotation software
- during the beta test period, but our customers have unanimously
- requested three enhancements: compression, compression and compression.
-
- We currently offer Intel's DVI 4-bit ADPCM, but people want more. Joe
- Campbell has been most helpful in getting us started on CELP 3.2 (4800
- bps) which sounds pretty good. Although the basic algorithms are
- public, it seems that the implementations tend to be proprietary.
-
- Motorola's VSELP sounds a bit better than 3.2 CELP, but you'd expect
- that given its 7950bps rate. Unlike DoD's CELP, Motorola has kept
- VSELP proprietary and won't be making a decision on how to market the
- algorithm until January '93 at the earliest.
-
- In addition to DSP-assisted compression, we'd like to offer Intel-based
- (ie, PC) decompression, so my question is, does anyone have any
- experience with decompressing CELP or VSELP on a '386 or '486 PC?
- Real-time would be helpful, but not absolutely required, given that
- annotation is not typically a real-time application.
-
- Finally, I'm looking for opinions on the suitability of CELP and VSELP
- (or other algorithms) for use in voice annotation applications.
-
- ...doug
-
- ==================================================
- Doug Kaye Rational Data Systems San Rafael, CA
- doug@rds.com {moon|uunet}!rds!doug MHS:doug@rds
- ==================================================
-