This does look a fascinating project.
Has anyone come across any photos of the board?
I'm also kind of guessing that it's not going to be open source; I do hope I'm wrong on this.
cbmeeks wrote:
Yeah, I think re-implementing the signal vs. up-sampling is a much better way (at least with my limited knowledge). In my experiences, up-samplers always have tearing, lag and other graphic glitches. Even the high-end stuff.
That may well be true on the Apple II.
On the Acorn machines it's not quite so clear cut, as many games make use of various nefarious 6845 programming tricks to achieve some quite amazing effects. So as well as re-implementing the video memory, you would need a very accurate emulation of 6845, the SAA5050, the video processor, and even down to how an analogue monitor reacts to non-standard sync pulse widths. After several years of development, these things continue to occasionally catch out even the mature Beeb emulators (B-Em, BeebEm, ans JSBeeb).
That's really why with
RGBtoHDMI that Ed mentioned earlier I've taken the "Up Sampling" approach. It's also been great fun seeing how far you can push a $10 Raspberry Pi. And as of today, we have the HDMI output frequency locked to the incoming video rate, with error free pixel sampling.
The Raspberry Pi, used bare metal, is actually a very nice platform for doing this kind of stuff on.
Dave