Re: Proposal: A more full RTCRtpParameters (RTCRtpCodecParams, RTX, FEC, SSRCs, and the beginnings of simulcast)

Comparing this proposal with the latest editor's draft (http://ortc.org/wp-content/uploads/2014/02/ortc.html), I have some questions. 

"dictionary RTCRtpCodecParameters {
 // The value that goes in the RTP Payload Type field.

​ unsigned byte payloadType;
​ // ​The same value that goes in the RTCRtpCapabilities.

 RTCRtpCodec codec;

​ // Format parameters that are only used for controlling
 // what is sent, and shouldn't be signalled.
​​ // For example, with opus, stereo=1
 ​
 sequence<KeyValueParam>        formatParameters;

​ // Because they are so different,
 // RTCP feedback params are separated out.​

 sequence<RTCRtcpFeedbackParam> rtcpFeedbackParameters;
};"

[BA] Looking at Section 7.2 of the Editor's draft, RTCRtpCodec is defined as follows: 

dictionary RTCRtpCodec {
    DOMString                      name;
    unsigned byte?               payload-id;
    unsigned int?                  clockRate;
    unsigned int?                  numChannels;
    sequence<RTCRtpCodecParameter> formatParameters;
};

Given the definition of  RTCRtpCodecParameters in the proposal, I'm wondering whether things like payload-id and formatParameters should be removed from RTCRtpCodec since they are included in RTCRtpCodecParameters.   Note that in Section 7.1 of the Editor's draft, RTCRtpCapabilities also depends on RTCRtpCodec: 

dictionary RTCRtpCapabilities {
    sequence<RTCRtpCodec>    audioCodecs;
    sequence<RTCRtpCodec>    videoCodecs;
    sequence<DOMString>      headerExtensions;
    sequence<RTCRtpFeatures> features;
};

However here the payload-id wouldn't be relevant, so removing it would be fine.  I'm wondering about formatParameters though, since I'm wondering how else the applicatino
might figure out what formatParameters are supported for a given codec. 

Received on Saturday, 15 February 2014 19:44:26 UTC