Hi!
Task; A FHD project limited to only one physical SDI port for delivering multiple FHD CG channels with transparency.
Since the SDI port is capable of UHD, I thought I’d just make a quad-split within the UHD signal and then unpack (map & key) it on the receiver end (separate HW/SW).
The idea works up until the point of including transparency where I need a “key-only” version of a CG channel as part of the quad-split.
Apparently, this wasn’t an easy setup as I had in mind it would be.
The only two things that support “key-only”, AFAIK, within the Caspar configuration are DeckLink and screen consumers, neither which will work for this case as I need the “key-only” as a “virtual” channel within the quad-split, before sending it out on a single DeckLink SDI port.
Some research on the topic mentions FFmpeg filter extraction but as far as I understand, that only works for image/video files and not for dynamic content (HTML), as that does not go through the FFmpeg producer.
Other topics mentions the use of the MIXER KEYER operation in CCG, but after some testing, I understand that it only would work with non-dynamic content as it requires the alpha information to be within the red color channel of the content.
So, to the question at hand, is there a way to get a “virtual key-only” channel representation within CCG that can be freely routed and manipulated before, in this case, being sent out as part of a UHD quad-split?
As the screen consumer can produce a “key-only” representation (only visible within the window, not in the CG channel), I assume the combined key “data” is “created” in CCG (not relying on other HW/SW (DL, NDI etc.)) and should then in theory be possible to extract to a dedicated channel, even if that’s not accessible in a public version of CCG today. Thoughts?
Thanks in advance!
// Victor