home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!wupost!waikato.ac.nz!comp.vuw.ac.nz!actrix!templar!jbickers
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: Chunky Pixels vs. Bitplanes (was: Chunky Chip Set...)
- Message-ID: <jbickers.0mem@templar.actrix.gen.nz>
- From: jbickers@templar.actrix.gen.nz (John Bickers)
- Date: 2 Jan 93 11:33:59 PST
- References: <1992Dec30.115759.22097@mpifr-bonn.mpg.de> <doiron.0ka3@starpt.UUCP>
- <1992Dec31.011428.2926@mpifr-bonn.mpg.de> <C03uC5.15o@unix.portal.com>
- <1992Dec31.174137.10865@mpifr-bonn.mpg.de> <jbickers.0mdl@templar.actrix.gen.nz> <1993Jan1.142652.20451@mpifr-bonn.mpg.de>
- Organization: TAP
- Lines: 56
-
- Quoted from <1993Jan1.142652.20451@mpifr-bonn.mpg.de> by mlelstv@speckled.mpifr-bonn.mpg.de (Michael van Elst):
- > In <jbickers.0mdl@templar.actrix.gen.nz> jbickers@templar.actrix.gen.nz (John Bickers) writes:
-
- > > So do you concede that 8-bit chunky pixels are indeed faster than
- > > 8-bit planar pixels for CPU-driven drawing operations?
- >
- > *sigh*. They are slightly faster for area operations and significantly
- > faster for non-horizontal line drawing. But I was not talking about
-
- Well, that's the key thing. As for the rest (< 8 planes), the
- differences are not so clear-cut. On the one hand you have to
- shift and mask so much for bitplanes, on the other you waste bits
- for chunky pixels, or end up stuffing several pixels into each
- byte, which approaches having planes (exactly equivalent for 1-bit
- pixels, eh?).
-
- You also lose all the marvellous things you can do with the
- bitplane pointers, but if such pointers aren't documented in the
- future it doesn't matter, eh? :)
-
- > > Can you present your algorithm for bitmap scaling that performs as
- > > well as a chunky-pixel based algorithm?
- >
- > It uses a 256k table (64k longwords) or less when I have a processor that can
- > do several instructions during video memory accesses. Why would that slow down
- > the algorithm ?
-
- This is a nice approach, though it gets more complicated when you
- want to scale by some factor other than 2 or 4. For each end of
- the horizontal lines you are scaling, you also need to mask off
- the source and destination areas, if any, that shouldn't be
- affected. You may also have to shift the results if the source and
- destination are not aligned with each other.
-
- > > Me. A digital dissolve is basically filling an area by writing
- > > single pixels.
- >
- > Hmm. I thought it was a mapping operation.
-
- Well, call it what you will but it's still filling an area by
- writing single pixels.
-
- > > Horizontal wipes do indeed fill the screen with vertical lines.
- >
- > True.. Now, who talks about special cases :) BTW, you do not want to
- > make a horizontal wipe arbitrarily fast. It could become invisible :)
-
- Heh. I would like to be able to complete a vertical rectangle
- within one frame though. You can do it quite nicely by copying in
- 8-pixel wide strips, until you get a case where you need to mask
- and shift.
-
- > Michael van Elst
- --
- *** John Bickers, TAP. jbickers@templar.actrix.gen.nz ***
- *** "Radioactivity - It's in the air, for you and me" - Kraftwerk ***
-