Hi Thomas!
>Manuel wrote:
>> -> Attached is my current Starfield WIP. After moving
>> 160 pixels, the horizontal *reset* of is not as
smooth
>> as it should be, but apart from that I'm very pleased
>> for the start :-)
>You will know that already, but I think, there are far
too many stars and
>it's looking much to symmetric.
100% agreed. I'm working on it ;-)
>For making the thing looking less symmetric, you could
change the HMOVE
>values inside a frame or use two different values (e.g
on even frames +8
>and -5 on odd ones).
That won't work. +8 does the trick, nothing else.
Consider this:
Horizontal scrolling works with +-1 and vertical
scrolling with +-8 pixels.
If you shift only a single line with anything else than
+8, the vertical scrolling is out of sync immediately.
>Then you should try to display only every ~4th star
>(perferable randomly, starting with different vertical
offsets on odd/even
>frames)
Hm... that's ~ how I'll try to work.
Total randomness won't work, since you've to *repeat*
what you're doing every frame, but I'll experiment with
that. I certainly won't sacrifice any RAM for that, but
maybe one can just scan bitwise somehow through a
randomly chosen part of the ROM for example. Or maybe
writing a *sequence-generator*, though I fear that any
possible implementation of that'll cost way to much
execution time during the display kernel.
> and maybe try to give the stars different colors.
Depends on how many cycles remain in the end.
Non-mirrored sprites first, then fancy colors :-)
>I think, that is what Solaris does too.
I'm too lazy studying the Solaris source because of all
the renamed TIA registers, but I analyzed the general
idea behind this technique in 'Earth Dies Screaming'.
I'd assume Solaris does similar.
Greetings,
Manuel
----------------------------------------------------------------------------------------------
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/
