4 ms·
> ‘The goal was to get the sprites to 8×8,’ explains Dailly, ‘but they actually ended up about 8×10 – hardly anyone has ever noticed this figure was wrong.’ I
by cpeterso 10y ago
> ‘The goal was to get the sprites to 8×8,’ explains Dailly, ‘but they actually ended up about 8×10 – hardly anyone has ever noticed this figure was wrong.’
I understand how 8x8 sprites would be easier to tile than 8x10, but what is so "wrong" about 8x10 that people might notice?
- erik 10y agoOld games usually kept their sprite sizes to multiples of 8. 8x8, 8x16, 16x16, etc. So 8x10 would have been weird for the time, and anyone who thought about it would probably assume they were 8x8. But there was likely nothing actually "wrong" about 8x10. I am just speculating here though.
- AimHere 10y agoNothing 'wrong', but the reason for the multiples of 8 was likely because sprites back then needed masks so that they could handle 'holes' properly. This was in the days before alpha channels, so there'd be spritemask instead - essentially a 2-colour sprite that acted like a single-bit alpha channel. And if your sprite is a multiple of 8-pixels wide, then your spritemask takes up a whole-number of bytes, you can do funky low-level bit-twiddling to handle occlusions and it's more efficient in terms of memory space.
- justinhj 10y agoThe Amiga had a powerful chip called a blitter for moving arbitrary rectangles of image data around (it could even draw lines) but most developers were used to nes, c64, snes, megadrive and these all used character based hw so 8 bit alignment made things easier and fast.