| Summary: | ASTERISK-01333: [patch] g723 codec seems to calc ast_frame.samples incorrectly | ||
| Reporter: | fossil (fossil) | Labels: | |
| Date Opened: | 2004-04-02 16:55:40.000-0600 | Date Closed: | 2004-09-25 02:51:28 |
| Priority: | Minor | Regression? | No |
| Status: | Closed/Complete | Components: | Core/General |
| Versions: | Frequency of Occurrence | ||
| Related Issues: | |||
| Environment: | Attachments: | ( 0) codec_g723_1.c.patch | |
| Description: | The g723 code currently seems to calculate ast_frame.samples in milliseconds instead of in 8KHz samples. The current math would also fit if g723 were a 64KHz codec, so which it is exactly (msecs or 64KHz) I have no idea. The only thing affected right now is the cost computations though. ****** ADDITIONAL INFORMATION ****** The patch to correct this is attached. Tested to work fine with transcoding to and from GS HT-286. | ||
| Comments: | By: Mark Spencer (markster) 2004-04-04 17:28:03 Fixed in CVS already By: fossil (fossil) 2004-04-07 18:28:17 The same should be applied to 1.0 stable, shouldn't it? At least all codecs in 1.0 stable seem to report the actual count of 8KHz samples in ast_frame.samples -- all except g723. By: Mark Spencer (markster) 2004-04-12 10:56:57 ok | ||