I managed to port recording function into WebRTC server using RecordRTC. Now I can save the video that is conveyed.

current structure:

But it strikes me that i might be encoding the video stream on the screen that is actually decoded from the bitstream. If so, I want to make the data flow structure different, because the decoding process is not necessary if I just signed up for containing the bitstream into a file only. I'm struggling with finding a way around decoding process.

the structure that I aimed for:

Is there a solution for this? If not, suggest me a possible alternative for this.