nametable & double buffering

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

By ARTRAG

Enlighted (6506)

ARTRAG's picture

17-05-2012, 00:09

A small scroll demo using double buffering (step x4)
Nemesis III lovers should try the file nemesisIIIdemo.dsk at

https://sites.google.com/site/testmsx/Home/double-buffer-in-...

use arrows to scroll, ESC to exit to dos

By ARTRAG

Enlighted (6506)

ARTRAG's picture

17-05-2012, 10:15

continuing tests...
is there anyone that could try the "cloningspritedemo.rar" on a real msx 1 and tell me what he/she sees ?

https://sites.google.com/site/testmsx/Home/double-buffer-in-...

in the sole odd frames of the scrolling, he/she should see 8 sprites from the pattern cloned from the middle of the screen to the lower part of the screen
All other sprites should stay not cloned.

A picture of the result would be great ;)

If this works, the next step would be to split the SAT in two segments: one for sprite 0-7, the other for 8-31

The game logic will spread the virtual sprites across the two segments according to their y coordinates
If the segment [0-7] becomes over crowed, one can resort to standard tricks for sprite multiplication (e.g. reversing the order of the virtual sprites with y<128 ).
This will introduce a small flickering but will allow up to 16 sprites with y<128 without cloning problems.

After all, on msx1 we live with flickering from the beginning of the time ...

By hit9918

Prophet (2901)

hit9918's picture

17-05-2012, 12:15

@ARTRAG, @PingPong, here we go again Wink

The mood has flipped to "bohoo, the VDP cant do".
Short before it was "that ZX coder should have done better".

My mood is constant.
The scroll needs graphics without wild neighbour tiles combinations. No news here.
With my own pixeling and puzzeling, the charset does not fill wild.
Dumping a Nemesis 3 level is a different story.

By PingPong

Prophet (3737)

PingPong's picture

17-05-2012, 12:48

ARTRAG wrote:

continuing tests...
is there anyone that could try the "cloningspritedemo.rar" on a real msx 1 and tell me what he/she sees ?

Can you provide a .cas file version or a bloadable format? most msx1 does not have disks.

By PingPong

Prophet (3737)

PingPong's picture

17-05-2012, 12:50

hit9918 wrote:

@ARTRAG, @PingPong, here we go again Wink
The mood has flipped to "bohoo, the VDP cant do".
Short before it was "that ZX coder should have done better".
My mood is constant.

Ok, we are no-costant people. So then? Question

By ARTRAG

Enlighted (6506)

ARTRAG's picture

17-05-2012, 13:19

@hit9918
the msx1 vdp has very nasty problems,
utopian could have ported Genesis on msx with some different choices (even using your smooth scroller) getting konami like results,
Nemesis 3 levels scrolling smoothly on msx1 look cool ( and will look even cooler with sorted sprites - teh subject should remind you something :-) )

No contradictions, no mood change. Wink
the rule is, If you want to see something coded on msx, code it yourself

PS
don't you see how nice is the Nemesis 3 level when scrolled halfway?
A key point of my solution is that the second tile set is not pixel accurate, just an approximated version of the shifted tiles
This allows to trade accuracy with number of tiles.
Anyway, the initial tile set for offset 0 is about 160 tiles, the shifted one is about 250 tiles. So both are perfectly fitting in vram. Moreover, as the tile set for offset 0 is used also on the upper bank of the screen, having only 160 tiles used leaves room for a lot of tiles for the score bar.

@Pinpong
I dunno how to get BIN or CAS versions. But I could try to do a ROM if none is able to run the current demo.

By hit9918

Prophet (2901)

hit9918's picture

17-05-2012, 14:00

@PingPong
"Ok, we are no-costant people. So then?"

The 9918 is a bitch.
First, you go thru all the pain again.
Then, the fat magenta Nemesis 3 flower scrolls.
Smile

By PingPong

Prophet (3737)

PingPong's picture

17-05-2012, 15:58

hit9918 wrote:

@PingPong
First, you go thru all the pain again.

If you refer to my previous sentence about how things were done on vdp chip, ("It's amazing how the vdp designer arranged things to prevent any useful usage of vdp. Really."). I was only a bit sarcastic.

By ARTRAG

Enlighted (6506)

ARTRAG's picture

18-05-2012, 00:49

I did new tests with ripped graphic and a first attempt of workaround for the sprite cloning problem (I hope Utipian will forgive me for the use of his scorebar).

https://sites.google.com/site/testmsx/Home/double-buffer-in-...

Could anyone tell me if, on areal msx1, the files in "demo4.rar" or in the "demo4.dsk" show sprite cloning ? (same files in different packages)

I expect to see the same result both on msx1 and on msx2, i.e. the upper 16 sprites flickering, the lower 16 sprites fixed and perfectly displayed.

The idea is that in hyrid mode the sprites in the lower bank of the screen should appear correctly.
Can anyone confirm with a test on the real thing?

By hit9918

Prophet (2901)

hit9918's picture

18-05-2012, 01:17

I start to wonder.
There always was focus on sprite cloning.
The sprite engine thinks the beam is more up.
Shouldn't there also be sprite disappearance.
If the low charset ever gets masked, I smell trouble.

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