Taking it all apart with todays tools is breezy easy but I can imagine with period tools it would of been far harder. Thank you for taking the time to explore this iconic and technical game. The fact they used DMA timing to achieve the horizontal scroll was a bold play. I wonder what the percentage of machines in the wild at the time were effected by bad memory timings?
Of course I understand VSP today, but back in 1990 and even with a few years of Action Replay hacking under my belt, I couldn't figure out the 'weirdness' of the scroll! Still, back then it was nice so see so many other effects being used to create a smooth, colourful game.
Thanks for making such a detailed video. The Creatures games were some of my favourites back in the day. I always wondered what that scrolling trick was about. Is there much of an advantage of doing scrolling this way? Not having to update so many tile indices maybe. If I remember correctly there was more of a hubbub about Mayhem's scrolling not working on all machines. Maybe they did it in a less compatible way, or maybe there was just more discussion about it.
Basically, the VIC compares the Y-Scroll setting with the lower 3 bits of the current Y raster position to determine when it should start a new character line. But the Y-Scroll is software-controllable, so when you set it to match the Y raster in the middle of a raster line, the VIC goes "Oh, I guess I should start displaying a new line of characters now!" and the horizontal position gets skewed. Has it ever been figured out exactly why this technique crashes on some C64 models? It cannot be the DRAM refresh because that's an entirely different function of the VIC that's independent of the character matrix DMA.
"VSP is achieved by triggering a badline condition during idle mode in the visible part of a rasterline. When this happens, the VIC chip gets confused about what memory address to access during the half-cycle following the write to $d011. It sets the internal bus lines to 11111111 in preparation for an idle fetch, but suddenly changes its mind and tries to read from an address with an LSB of 00000111. Now, since electrical lines can't change voltage instantaneously, there is a brief moment of time when each of the changing bits (bit 3 through 7) is neither a valid one nor a valid zero. But because the VIC chip changes the address at an abnormal time, there is now a risk that the RAS signal, which is generated independently by another part of the VIC chip, is sent while one or more bus lines is within the undefined voltage range." www.linusakesson.net/scene/safevsp/index.php
Taking it all apart with todays tools is breezy easy but I can imagine with period tools it would of been far harder.
Thank you for taking the time to explore this iconic and technical game. The fact they used DMA timing to achieve the horizontal scroll was a bold play.
I wonder what the percentage of machines in the wild at the time were effected by bad memory timings?
Modern tools definitely makes things a lot quicker. :) I remember researching stable rasters 35 years ago and it taking ages!
Of course I understand VSP today, but back in 1990 and even with a few years of Action Replay hacking under my belt, I couldn't figure out the 'weirdness' of the scroll! Still, back then it was nice so see so many other effects being used to create a smooth, colourful game.
It was one of those "it works, don't question it" type of routines. :)
Thanks for making such a detailed video. The Creatures games were some of my favourites back in the day.
I always wondered what that scrolling trick was about. Is there much of an advantage of doing scrolling this way? Not having to update so many tile indices maybe.
If I remember correctly there was more of a hubbub about Mayhem's scrolling not working on all machines. Maybe they did it in a less compatible way, or maybe there was just more discussion about it.
I'll have a look at Mayhem at some point
Thanks for the great video.
Thanks!
Thank you very much
Basically, the VIC compares the Y-Scroll setting with the lower 3 bits of the current Y raster position to determine when it should start a new character line. But the Y-Scroll is software-controllable, so when you set it to match the Y raster in the middle of a raster line, the VIC goes "Oh, I guess I should start displaying a new line of characters now!" and the horizontal position gets skewed.
Has it ever been figured out exactly why this technique crashes on some C64 models? It cannot be the DRAM refresh because that's an entirely different function of the VIC that's independent of the character matrix DMA.
"VSP is achieved by triggering a badline condition during idle mode in the visible part of a rasterline. When this happens, the VIC chip gets confused about what memory address to access during the half-cycle following the write to $d011. It sets the internal bus lines to 11111111 in preparation for an idle fetch, but suddenly changes its mind and tries to read from an address with an LSB of 00000111.
Now, since electrical lines can't change voltage instantaneously, there is a brief moment of time when each of the changing bits (bit 3 through 7) is neither a valid one nor a valid zero. But because the VIC chip changes the address at an abnormal time, there is now a risk that the RAS signal, which is generated independently by another part of the VIC chip, is sent while one or more bus lines is within the undefined voltage range."
www.linusakesson.net/scene/safevsp/index.php