The thing that really impressed me was the chessboard scroll... Although you explained it, it took me quite a while to realize that I have to think colors "in SCREEN 3", graphics "in SCREEN 2" and final implementation in "SCREEN 1"... combined with accurate line sync that makes it THE "impossible on that platform"-demo for MSX1.
@Manuel: nothing in the pipeline for MSX, but: likely early next year, a major release on Amstrad CPC (which my heart belongs to).
@Ren: I will ask Factor6 if he agrees to share the PT3 tune here.
@Nyyrikki: thanks for your comments, I appreciate. I mean: there are very very few guys who could understand what was done on IO demo (Words in private from Dvik were also nice by the way). Within MSX community: I can see many gamers or HW-freaks which even don't see the impossibilities (1). Then, within whole demo-scene world: well, who knows about MSX1? As a conclusion: I am surprised (and happy!) about the nomination.
(1) Let's list them, from my side:
(unlimited sprites by page-flipping or large rotozoom or big-scroll-on-gfx are definitely not impossible)
accurate sync at 1st screenline/scanline (which then allows vdp orders at given screenline to achieve: rasters, split-screen, sprites mutliplexing, split-screen at each scanline for wobbler or so -> the plasma scene)
splitrasters border+screen, achieved by constant-time coding all over one frame (+ constant-time ym player)
painful patches for MSX1 vs MSX2 vs [MSX2+,TurboR]
Source code is available (warning: quite dirty, very few comments mixing french and English).
Wow, although i don't practically know anything about the CPC, it looks amazingly impressive to me! Well done!
I wonder what kind of tricks you used and whether they could also be used on MSX of MSX2...
) 
