pouët.net

After three years of delay, I finally released a VIC-20 demo that uses the bus trick I described in summer 2003 in my "VIC-20 frontiers" article. Basically, the trick makes the processor put specific bytes on the data bus in a specific order, and these bytes get fetched by the video chip as "bus noise". It's quite nice for the kind of effects where you pick individual lines from a bitmap picture to be displayed on specific scanlines. Y-stretchers, rotating bars, fake-3D twisters, you name it.

On the last day of working on the demo, I noticed that there was some difference in the bus implementation in different individual PAL VIC-20s, resulting certain bit leaks to occur differently. Although I declared in the demo that "the last secrets of the VIC-20 hardware have finally been discovered", it seems that there is still some work to do regarding the differences in different motherboard revisions. It is like doing fundamental science: every time you think you have grasped the ultimate explanation, the reality suddenly becomes more complex.