I see that the best video seems to be coming from GOPRO devices mounted to the top of the ROV. I assume that there are some video losses required to obtain a sufficiently fast refresh rate despite bandwidth limitations for the surface operator to see a live feed. I assume these losses are either in down sampling the resolution or in the compression codec.
Has anyone thought about the logistics of splitting the video stream?
1. Onboard capture to memory card with optimal resolution etc..
2. Lossy compression etc.. sent to surface for operator