euhm... can you show an example (just a few tiles) ?

http://ccompilerformsx.googlepages.com/wolf_test
the difference is the screen 5 detail I mean

Hmm yea well, that's rather simple, and the emptiness spots can easily be RLE'ed away. I never make simple gfx like these tho..

My skill as screen 2 artist are quite low (it was an egg BTW)
The idea is that starting from good screen 2 tiles, you need
only sparse details to move them to screen 5.

Quote:

My skill as screen 2 artist are quite low (it was an egg BTW) The idea is that starting from good screen 2 tiles, you need only sparse details to move them to screen 5.

This looks a nice idea, but there'll be the need to work with two bitmap tile-sets to generate the diffs, then convert the tile set 1 to screen 2 mode... and the game has to convert that data to bitmap again, before applying the diffs when using screen 5. Is it what you had in mind?
It could produce good graphics for a dual game using little more space, at the cost of some pre-processing. I believe it's a neat idea.

Yes it is what I mean.
All the work you say has to be done off line by a PC.
The msx code is very simple instead.
Games that update the tiles only during a room change could have enough time to expand in ram the screen 2 tiles
expand the "difference" and produce (on fly) a screen 5 tile set by writing pixels in VRAM.

A color tileset, uncompressed, takes only 4Kbytes.
Building a screen 5 image from that is only a matter of few lines of code that interpret screen 2 data and do "pset(x,y,c)" with the foreground or the background color according to the right bit in the shapes.
The screen 5 "correction" using the "diff" data can be done after having built the image above directly in VRAM,
by interpreting the RLE data that store the differences.

It works well for games that do not change the patterns during the game, as you pointed the other thread about GnG.
For animated tiles all what can be done is store in VRAM all the precomputed frames.
Even in this case the storing system could be the same, but the problems are those you pointed in the other thread about GnG (e.g the need of a 16 bit name table in ram, the need to find updated tiles in it, etc etc...)

So we want MSX2 games just because they look better????

MSX1 screen2 is so much fun to work with as a pixeler (and noob coder). I really like pattermodes because they are fast, simple and it has an adorable color-clash as a bonus

And more important; A good gfx pixeler in screen2 is better than a bad screen5 pixeler (looking at noone in particular). That is the main problem with any project. People want the most fantastic features but its nearly impossible to find a gfx artist that is interested/capable to create such gfx.

Well Caos Begins could be updated to screen 5 in a second...
that is the kind of game "à la" kings valley that can have two
version sharing the very same engine...

Quote:

Well Caos Begins could be updated to screen 5 in a second... that is the kind of game "à la" kings valley that can have two version sharing the very same engine...

Maze of Galious would also be a good candidate.

I thought the two were the same

Quote:

and it has an adorable color-clash as a bonus

actually.. about the color-clash.. what's the definition of color-clash? Is it gfx optimized for a screen2-style system? Or is it a clear error when you're converting a bitmap image (such as a digi) to pattern mode? I'd say the latter.. and therefor I'd say that only very few MSX1 games have color-clash.. ^_^

Color clash is when visible background obviously disappears or changes color because a software (tiled) sprite jumps up and shoots the pooh out of the pixels. Rather obvious examples are all those shitty Speccy conversions...
(And also 'the Cure' had this effect very prominent)

Commands as LINE, CIRCLE and PSET in BASIC can also cause "color spill".

I'll never forget when I was drawing in screen 2 and it happened.

"Huuuuuh... WTF?"

Quote:

I thought the two were the same

I thought the two are very different.

Note the sarcasm, MrSpock, note the sarcasm