Experiment: msx1 1bit video player

Wow!!! l looks great!! Congratulations for your work!!

Nice! Not sure if it would help, but taking a quick peek at the code, I see you encode each frame separately. One optimization could be to only encode the differences with respect to the previous frame (as is done in mpg, and other video encoding formats). Not sure how much would it help in this case, given that the video is very "pixelated", and thus almost every pixel changes at every frame. But worth a try

What abput so e optimization or unrolling to keep the frame rate as high possible,.?

It's only a test, it has no real utility. I could scratch some bytes transfering only the changing pixels from the previous frame but I prefer to maintain the code simple. Anyway, the videos takes too much space, with or without encoding differences.

I think WebMSX it's detecting wrongly the memory mapper of the cartridge.

There are three bottlenecks. The first is the uncompression routine, the second is the cpu->vdp transfer rate and the third is the video size (compression is mandatory). There is no much code to unroll and the uncompression routine it's small and well optimized (or so I think). I discarded using colors for those three reasons.

I've made a turbor version with double buffer and a new video. It works a lot better.
Download turbor version

Because MSX2 and up has so much VRAM, I suggest to use triple buffer because then you don't have to think or wait anything ever... Idea is this...

Main routine:
- Fill 1st VRAM buffer
- Ready page = 1
- Fill 2nd VRAM buffer
- Ready page = 2
- Fill 3rd VRAM buffer
- Ready page = 3
- goto first step

Interrupt handler:
- Visible page = Ready page
- RET