home *** CD-ROM | disk | FTP | other *** search
- Path: informatik.tu-muenchen.de!fischerj
- From: fischerj@informatik.tu-muenchen.de (Juergen "Rally" Fischer)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: CODER NEEDED !
- Date: 18 Jan 1996 19:38:58 GMT
- Organization: Technische Universitaet Muenchen, Germany
- Distribution: world
- Message-ID: <4dm7ki$quu@sunsystem5.informatik.tu-muenchen.de>
- References: <w7lPPMD0aez4@p09.hotline.pfalz.de>
- NNTP-Posting-Host: hphalle5.informatik.tu-muenchen.de
- Originator: fischerj@hphalle5.informatik.tu-muenchen.de
-
-
- In article <w7lPPMD0aez4@p09.hotline.pfalz.de>, ROI@hotline.pfalz.de (Roi@hotline.pfalz.de) writes:
- |> Hello !
- |>
- |> I am desperatly looking for a programmer who is able to
- |> develop fast routines for zooming/scaling and rotating
- |> sprites !
- |> The sprites are in 32 colours and the routine is needed
- |> for a Space-Combat-Simulator such as WingCommander.
-
- Oki, I read this as rotzoomer on A500. Not easy.
- seems to be better if not c2ping whole disply.
- you wanna go 1x1 ? wow :)
-
- I'd suggest to
-
- 1) rotzoom the sprite to a planar bitmap with cpu
- 2) copy & stamp with blitter (so cpu rotzoomer doesn't need to handle
- transparent pixels)
-
- Fastest method would be a ship only having 3 out of 32 colors (only 3 :\)
-
- reason: we puzzle together the data of planes in a word-register
-
- add.w offset0(A0),d0 -------1-------0 <- this is how a pixel in source looks
- ^plane1 ^plane0
-
- add.w d0,d0
- add.w offset1(A0),d0 ------11------00
-
- do this 8 times and you get: 1111111100000000 2 planar bytes.
-
- you can now let cpu store the bytes to the planes separately or
- just store .w and let the blitter sort it (needs 1 pass) out before doing 2)
-
- the longword method gives you 16 colors out of 32:
-
- -------3-------2-------1-------0
-
- but source data needs lot of mem (4byte/pixel!) and .l operations
- are evil to 68k...
-
- another one: word again, but nibbles, best imho for 68000 doing 16 colors.
-
- ---3---2---1---0 let the blitter sort the nibbles to plane (needs 2 passes)
-
- on A1200 I would suggest to do a chunky to planar routine.
- I don't know, maybe blitter4pass is not that bad on A500 ?
-
- don't know...
-
- |> Any suggestions about the method or the maths of doing
- |> such routines are welcome !
- |>
- |> Thank You,
- |>
- |> Roi
- |>
- ------------------------------------------------------------------------
- fischerj@Informatik.TU-Muenchen.DE (Juergen "Rally" Fischer) =:)
-
-