Siren: Difference between revisions
No edit summary |
(fill missing details) |
||
| (One intermediate revision by one other user not shown) | |||
| Line 2: | Line 2: | ||
* Company: [[Microsoft]] | * Company: [[Microsoft]] | ||
* Decoder: https://github.com/kakaroto/libsiren | * Decoder: https://github.com/kakaroto/libsiren | ||
This audio codec is used by MSN Messenger for sending/receiving voice clips. It is also one of the available codecs for the 'Computer Call' feature (audio conference). | This audio codec is used by MSN Messenger for sending/receiving voice clips. It is also one of the available codecs for the 'Computer Call' feature (audio conference). | ||
| Line 10: | Line 9: | ||
(excluding the tcl_siren* files). | (excluding the tcl_siren* files). | ||
This codec is the same as [[Vivo Siren]] except for some slightly different bitstream format (e.g. no sample rate ID in the beginning or four-bit CRC at the end and frame data is padded) and some parameters related to sample scaling. | |||
[[Category:Audio Codecs]] | [[Category:Audio Codecs]] | ||
Latest revision as of 06:30, 22 July 2023
- Format: 0x28E
- Company: Microsoft
- Decoder: https://github.com/kakaroto/libsiren
This audio codec is used by MSN Messenger for sending/receiving voice clips. It is also one of the available codecs for the 'Computer Call' feature (audio conference). It is based on the G722.1 codec and only has a few differences to it. For example, it uses a 2 byte big-endian format for storing the encoded data instead of little-endian, and it also adds a few bits (mostly 2) as a header to each frame to specify the sampling rate, and a 4 bits footer which contains a checksum of the frame (but keeping the size of the block to 40 bytes per frame).
The codec has been reverse engineered and is available for download at : https://amsn.svn.sourceforge.net/svnroot/amsn/trunk/amsn/utils/tcl_siren/src/ (excluding the tcl_siren* files).
This codec is the same as Vivo Siren except for some slightly different bitstream format (e.g. no sample rate ID in the beginning or four-bit CRC at the end and frame data is padded) and some parameters related to sample scaling.