NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/sys/amiga/programmer/2363

Internet Message Format | 1996-08-05 | 4.4 KB

Path: informatik.tu-muenchen.de!fischerj
From: [e-mail] (Juergen "Rally" Fischer)
Newsgroups: comp.sys.amiga.programmer
Subject: Re: WritePixelArray vs. ClipBlit vs. ...
Date: 30 Jan 1996 15:30:16 GMT
Organization: Technische Universitaet Muenchen, Germany
Distribution: world
Message-ID: <4eldi8$[e-mail]>
References: <[e-mail]> <sqoWx*[e-mail]>  <4e9dju$[e-mail]> <[e-mail]>
NNTP-Posting-Host: hphalle7g.informatik.tu-muenchen.de
Originator: [e-mail]


In article <[e-mail]>, [e-mail] (Steve Koren) writes:
|>
|> [e-mail] (Juergen "Rally" Fischer) wrote:
|>
|> > : > This is annoying.  Is there any way to do case D) easier than what I am
|> > : > thinking?
|> >
|> > I do not quite get what you want to do. Looks like your engine renders
|> > planar on AGA and chunky on gfx-card. This should give 10 times more work
|> > than adapting the copy routines. I'm really wondering...
|>
|> Right now my engine never renders planar (it will run without a graphics
|> card, but I don't consider systems without graphics cards useful, so I
|> don't care about optimizing for that case).  It renders 24 bit truecolor
|> for 16 and 24 bit gfx card modes, and it renders 8 bit LUT for 8 bit
|> register modes.  WritePixelArray(cybergraphics.library) and
|> WritePixelArray8(graphics.library) are used, accordingly.  I would
|> prefer to *always* render 24 bit, but there doesn't seem to be an easy
|> way to blit a 24 bit bitmap to an 8 bit screen and have the remappinng
|> work.  So, failing that, I do more work in my code to render either 8
|> bit LUT or 24 bit truecolor graphics depending on the screen I'm going
|> to be talking to.
|>
|> > Also why 24bit mode. you want to play the 8bit chunky game also when
|> > 24bit is present I guess. Or do you even have a 24bit version of your
|> > rendering engine.
|>
|> Well, it is not a game, it is an application.  It would prefer to always

So if you don't do animations where converting to planar cares,
writepixelarray8(gfx.lib) is sufficient for any 8bit screen on any
gfx-system (AGA and cards).

|> deal with 24 bit color space.  I did briefly consider always rendering
|> to 24 bit internally, and using ObtainBestPen() on each pixel before
|> displaying to achieve the truecolor->LUT conversion.  However, this is
|> *terribly* slow on a 68040/25.  ObtainBestPen() is no practical for use

mhm I don't know, is obtainpen() one of that workbench-colorsharing
routines ? Or is the conversion 24bit-> 8 bit done by you ?
Ever though about using HAM8 ? There is a way on AGA to get Workbench
into HAM8 mode with just clicking on a (selfmade) prefs-icon BTW.
Rendering to HAM also needs its time, but at least the result is better.

|> on every pixel in a large image.  So instead I obtain some pens up front
|> and render using them, if I'm on a LUT screen.
|>
|> > anyway it's no wonder that updating a 256 color window on a 24bitscreen
|> > needs some time (colortable lookup...)
|>
|> Yes, but it is far slower than I can explain by doing lookups.  If it
|> was just doing one color table lookup per pixel, it should be much
|> faster (compared to WritePixelArray)) than it is.  So I don't understand
|> why it is so slow now.

? I thought you explained 24->8. 8->24 is slow seen from games, but
fast seen from a rendering program.

|>
|> > : I'm afraid not, except waiting for true RTG. As far as I understood
|> > : it, RTG would provide automatic format conversion when blitting
|> > : between any bitmap formats.
|> >
|> > huh ? Current functions do this, too. writeparr8, blitbitmap...
|> > if they don't, imho the gfxcard driver is not ok.
|>
|> They do in some cases, but not all.  I can blit an 8 bit array of pen
|> values to a 16 or 24 bit screen (although quite slowly).  But I cannot
|> blit a 24 or 16 bit truecolor array of data to a LUT screen mode.  I
|> believe that's what the poster above means.  Some systems do this -
|> applications can always speak 24 bit, and any necessary remapping is
|> done at display time.

if you got multiple windows needing different pallete it becomes
more difficult..

|>
|>    - steve
------------------------------------------------------------------------
   [e-mail] (Juergen "Rally" Fischer)   =:)