[asterisk-bugs] [JIRA] (ASTERISK-25599) [patch] SLIN Resampling Codec only 80 msec
Alexander Traud (JIRA)
noreply at issues.asterisk.org
Tue Dec 8 08:38:33 CST 2015
[ https://issues.asterisk.org/jira/browse/ASTERISK-25599?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Alexander Traud reopened ASTERISK-25599:
----------------------------------------
When the Opus Codec is used together with forward-error correction ([FEC|https://en.wikipedia.org/wiki/Forward_error_correction]), the amount of samples doubles to 11520, because up to two 120 msec packets are decoded to up to 48.000 kHz. Therefore, I have to reopen this issue because the previous patch was incomplete.
> [patch] SLIN Resampling Codec only 80 msec
> ------------------------------------------
>
> Key: ASTERISK-25599
> URL: https://issues.asterisk.org/jira/browse/ASTERISK-25599
> Project: Asterisk
> Issue Type: Bug
> Security Level: None
> Components: Codecs/codec_resample
> Affects Versions: 11.20.0, 13.6.0
> Reporter: Alexander Traud
> Severity: Minor
> Attachments: opus_resample.patch
>
>
> {{WARNING: translate.c:402 Out of buffer space}}
> Currently, {{codecs/codec_resample.c}} uses a buffer with a fixed size of 8096 for each signed-linear resolution. Furthermore, these 8096 depend on the size of {{int16_t}} of the platform. On my machine, this results in a maximum of 4048 samples. However for example, the Opus Codec is able to create 5760 samples (6 frames a 20 milliseconds @ 48000 samples per second). Consequently, the current buffer is too small for the Opus Codec when used with a packetization time of more than 80 milliseconds, for example in chan_sip with sip.conf: {{allow=opus:120}} or {{autoframing=yes}}.
> The attached patch uses the maximum of the Opus Codec (5760) as the new minimum sample amount. Therefore, the actual amount of bytes must depend on {{int16_t}} and not the other way around. This guarantees we are able to convert at least 5760 samples, regardless of the platform.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
More information about the asterisk-bugs
mailing list