Replies: 2 comments 1 reply
-
i don't think this is needed, and would discourage the adoption and implementation of E-RTMP. code changes are already needed to support the new codecs and any codec-specific framing and special messages, so why not just do it the Enhanced way to begin with? extracting the codec ID the Enhanced way is literally just a few lines of code that should be factored to one place anyway, like: https://github.com/zenomt/rtmfp-cpp/blob/73fd63e6f3ae44de3554a1aa2c7b391fb7c560fe/src/TCMessage.cpp#L158-L176 also, allocating all of the remaining "classic" code points removes future flexibility for additional extension, should it ever be needed (like how audio was Enhanced by taking one of the unallocated code points). |
Beta Was this translation helpful? Give feedback.
-
Consider it pulled from ideas then; I understand the points made. |
Beta Was this translation helpful? Give feedback.
-
I'd like to propose, for the sake of compatibility and in the spirit of the original codec identification scheme that the following modifications and additions be made to the enhanced specification. In short, that the following byte identifiers be used:
Audio codec / sound format
0x0c
0x0d
0x0f
(reserved / device specific in spec)0x10
0x11
0x12
Video codec
0x08
0x09
0x0b
0x0c
0x0d
These changes should in no way preclude the implementation nor use of enhanced RTMP specified functionality.
Lastly, I am not aware of any devices or implementations on the market using any of these identifiers other than
0x0c
for HEVC, for which there are several available.Beta Was this translation helpful? Give feedback.
All reactions