Re: [stella] Repositioning Proof Of Concept

Subject: Re: [stella] Repositioning Proof Of Concept
From: Manuel Polik <cybergoth@xxxxxxxx>
Date: Sun, 17 Mar 2002 20:38:52 +0100
Hi Piero!

>>It repositions _both_ sprites in the same scannline!

>How do you do that?

It works pretty straightforward. Basically, as one would 
expect, I just do


in one scannline.

With the second RESP, I loose one pixel, so using the 
same HMOV values for both sprites would produce a one 
pixel gap.

You lose 15 cycles every iteration in the timing loop 
before the RESPs, but you have 16 different HMOV values, 
i.e. a one pixel shift is just possible!

So I worked out two tables, that are shifted by one 

Here's a little example code:

    LDA movetab1,X
    sta HMP0
    LDA movetab2,X
    sta HMP1
    AND #$0F
    bpl RespLoop
    sta RESP0
    sta RESP1

The tables look like this:

	.byte $6X,$5X,$4X,$3X,$2X,$1X,$0X,$FX
	.byte $EX,$DX,$CX,$BX,$AX,$9X,$8X

	repeat with X+1;
	repeat with X+2;

	.byte $7X,$6X,$5X,$4X,$3X,$2X,$1X,$0X
	.byte $FX,$EX,$DX,$CX,$BX,$AX,$9X

	repeat with X+1;
	repeat with X+2;

>Are there any restrictions?

Well, there's some troubles at the left/right edges. I 
have yet to determine the maximum range where the big 
ships can be freely moved. And I've yet to find out the 
ranges where RESPX is working properly. 
(I think there's a table somewhere in the archive :-))

Another restriction is, that all you can do with this 
technique is gaps with 0-2 pixels between both sprites. 
This is working perfect for my game, but if your 
intention would be total free positioning of both 
sprites, the solution would be... uhm... complex. I 
think it is possible, but...


Archives (includes files) at
Unsub & more at

Current Thread