Path: utzoo!utgpu!watserv1!watmath!att!dptg!ulysses!andante!princeton!njin!rutgers!usc!elroy.jpl.nasa.gov!ncar!gatech!purdue!haven!mimsy!mojo!russotto
From: russotto@eng.umd.edu (Matthew T. Russotto)
Newsgroups: comp.sys.apple2
Subject: Re: 3200 Gif converters
Keywords: DreamGrafix, DreamWorld, 3200
Message-ID: <1990Nov26.184751.681@eng.umd.edu>
Date: 26 Nov 90 18:47:51 GMT
References: <1990Nov26.054401.24321@cunixf.cc.columbia.edu> <1990Nov26.055040.14047@cunixf.cc.columbia.edu>
Sender: news@eng.umd.edu (The News System)
Organization: College of Engineering, Maryversity of Uniland, College Park
Lines: 38

In article <1990Nov26.055040.14047@cunixf.cc.columbia.edu> yk4@cunixb.cc.columbia.edu (Yong Su Kim) writes:
>I was playing around with Jonah Stich's 3200 gif converter, and it
>struck me that there had to be a better way of converting gifs into
>3200 colour pictures on the GS.
>
>Jonah's program is extremely slow in converting gifs into 3200 colour
>modes using the Median Cut algorithim. I don't really see why he uses
>the median cut algorithim since median cut algorithim involves areas.
>All he really has to do is use the popularity algorithim on each
>individual scan line. That way it's possible to cut down on a lot of
>processing time.

In fact, neither algorithm is very satisfactory-- the problem is that you need
consistency from one scan line to the next, and neither algorithm will give
that to you consistently-- popularity seems to do a bit better than median
cut, but some 16-color median cuts look as good as 3200 color popularity
pictures.  (I have some REAL UGLY 3200s to prove it)
I don't think median cut is inherently slower than popularity, BTW.

>Another solution might be to write a 3200 gif converter for the GS
>using C so that it runs under standard UNIX. The programmer could
>design it such that it accepts standard UNIX binary files as gifs,
>process them using various optimization algorithims for 3200 colour
>mode, and then output a binary file which can then be downloaded to my
>GS. All I would have to do is change the filetype and then I would be
>able to view the conversion. This would be much much quicker and it
>would probably allow for superior conversions. 
>
>Anyone out there have the time to write such a program? 
>
>I would do it if I had the time and the knowledge...

Forget writing one from scratch-- take a look at the pbmplus programs-- there
seems to be one for everything, and one ought to be easily modified to do
what you want.
--
Matthew T. Russotto	russotto@eng.umd.edu	russotto@wam.umd.edu
     .sig under construction, like the rest of this campus.