Hello
The "room" with the dual codecs and single camera - Will there be active participants in that room?
What will the camera be viewing? Basically what type of venue?
Will French and English speakers be in that same room?
How many rooms will need to see all participants? Will French far-end participants need to see English far-end participants?
How many sites will have both English and French Speakers?
I have done similar multi-language (always 3-5) many times but always with separate audio channels per language using devices purpose made for multilingual multisite conferences and audio encoders. VTC session is conducted as normal and the audio is translated (by translators in soundproof booths) and users would subscribe to the language of their choice using the IR headsets. Participants would speak whatever language they want into the mics and others would hear the speaker live on room speakers and in the subscribed language on their headsets. the multi channel (different languages) audio to and from external sites was handled by encoders and decoders -
For your solution based on what you describe you will need to send out only "one video" channel and really only one audio channel out per codec. Since each channel will use a dedicated codec.
Each CODEC dials into a dedicated conference on the bridge - one for English and one for French. The far end participants will dial into the respective language conference. The far end will hear the language of the specific bridge. NOW STOP
This is where complexity starts and is more of a A/V Engineering issue than a Telepresence issue.
The Audio and Video from the far on the separate CODECs need to be integrated into the room without feedback distortion etc.. The video is pretty easy but the audio will be the most complex to work through.
Required Components
- DSP with mic mixer
- Matrix Switch
I would use three or four displays then have one display for each near-end codec displaying the far-end(s) video (one per language). The third display will be for content and if 4th display is used it can be for content in second language or another site
The DSP with mic mixer to route all mics and the received far-end audio. teh DSP will mix the incoming audio and mix/minus the mics and program audio to prevent echo, feedback and other audio issues.
All the basic concepts and principles of this architecture is proven but I have never implemented it in exactly the way you have it. Working in the areas I was, with multiple languages it was easier to add audio channels than it was to add additional codecs
I'd love to hear how this comes out and help if needed
