msx vaporware - true or false ?

Pagina 5/24
1 | 2 | 3 | 4 | | 6 | 7 | 8 | 9 | 10

Van ARTRAG

Enlighted (6567)

afbeelding van ARTRAG

09-06-2012, 15:24

The noise is due to the AVI compression
Look at the frames of the scrolling in this file
https://sites.google.com/site/testmsx/Home/double-buffer-in-...

Van ARTRAG

Enlighted (6567)

afbeelding van ARTRAG

11-06-2012, 01:26

The x2 scrolling is starting to work.
This rom comes from a demo picture by Lord S.M. Zett by Darkstone (thanks!)
https://sites.google.com/site/testmsx/Home/double-buffer-in-...

There are a couple of glitches I'm investigating (my data probably).

PS just a note: the average number of tiles updated per frame is 2,3 (including in the average the initial loading of the initial tile sets ... @hit9918 it is about 40 bytes per frame + the PNT! )

PS
For the sake of comparison, this is the same level by Lord S.M. Zett with x4 scrolling
https://sites.google.com/site/testmsx/Home/double-buffer-in-...

Van ARTRAG

Enlighted (6567)

afbeelding van ARTRAG

11-06-2012, 01:28

This is the x2 scrolling of a picture done reusing the Salamander graphic:
https://sites.google.com/site/testmsx/Home/double-buffer-in-...
No glitches this time (dunno why, but I like it in this way)
The average number of updated tiles per frame is 1.2366 , i.e. about 20 bytes per frame....!!
@hit9918, now you can compare it with the avi of the simulated decoder

Van hit9918

Prophet (2911)

afbeelding van hit9918

11-06-2012, 11:31

When PingPong told about diffs, from day one I never was worried about the update time.
The critical number I look at is the scroll granularity Tongue i.e. the cricial thing is vram cache size.
Because that is about things onscreen, the more you bang at this size, the more it is IMPOSSIBLE that large updates happen. With this engine in practice you will never have update time worries and blast 60fps hands down.

The engine got two elements that are actually independant:
1. vram cache doubling to 512 chars, the hybrid mode.
2. Update the cache. It is no dynamic alloc every frame, still through the level tile numbers much change their meaning.

Van hit9918

Prophet (2911)

afbeelding van hit9918

11-06-2012, 11:44

Another thing: actually the non-hybrid fullscreen got 3*256 = 768 chars vram cache, the full 12k!
It is less flexible but custom puzzeling can have more tiles onscreen.

If I got that right, when a tile is same pattern over multiple scrollsteps, it needs to be only one time in the cache?
Very interesting. This would be number 3. in previous postings engine feature list.
This needs somehow managing an individual nametable for every scroll step.

Would be interesting to know what this can do fullscreen 1 pixel scroll.
Is Malaika 2 still planned? Big smile

This reminds me to collision detection:
For every tile you need know a path to the collision type, two bits "left corner, right corner".
The corners are moving 0..7 with the scrollstep.
Both bits set means the char is solid in all scroll steps.

Van Lord_Zett

Paladin (807)

afbeelding van Lord_Zett

11-06-2012, 11:50

I really like this demoBig smile

artrag gona make a game of it?

Big smile

Zett

Van hit9918

Prophet (2911)

afbeelding van hit9918

11-06-2012, 12:40

Oh my god, the Zett demo looks like made for this engine!
Please make demo with 2 pixels in 2 frames!

It will be like this
http://www.youtube.com/watch?v=A5rlpJY7gXw
and this
http://www.youtube.com/watch?v=NUFAmJPqxK8
:D
(for framerate, watch it on youtube and not in this browser fade applet thingy)

Van ARTRAG

Enlighted (6567)

afbeelding van ARTRAG

11-06-2012, 13:50

2 pixels per frame (I remved the extra halts with an hex editor)
https://sites.google.com/site/testmsx/Home/double-buffer-in-...

Van Lord_Zett

Paladin (807)

afbeelding van Lord_Zett

11-06-2012, 13:47

Damn it looks nice on that speed. only strang the base of that flowers.

Van ARTRAG

Enlighted (6567)

afbeelding van ARTRAG

11-06-2012, 15:38

It is due to the dithering

Pagina 5/24
1 | 2 | 3 | 4 | | 6 | 7 | 8 | 9 | 10