Path: utzoo!utgpu!news-server.csri.toronto.edu!torsqnt!lsuc!becker!bdb
From: bdb@becker.UUCP (Bruce D. Becker)
Newsgroups: comp.graphics
Subject: Re: 64K colors?
Message-ID: <98724@becker.UUCP>
Date: 8 May 91 19:23:33 GMT
References: <1991May2.001518.30298@kuhub.cc.ukans.edu> <10497@labtam.labtam.oz>
Organization: G. T. S., Toronto, Ontario, Canada
Lines: 54

In article <10497@labtam.labtam.oz> graeme@labtam.labtam.oz (Graeme Gill) writes:
>In article <1991May2.001518.30298@kuhub.cc.ukans.edu>, bbb@kuhub.cc.ukans.edu writes:
>> Dear Net-ers:
>> 
>> ... and just today got my  
>> Tseng Lab Inc. (TLI)  ET-4000 controller 
>> data book.
>> 
>> While checking out the specs on the 
>> 4000-chip, my eyes bugged as I saw that 
>> it is (in theory anyway)  capable of 
>> producing 65535 colors on screen using 2 
>> bytes to produce a  pixel.
>
>	As I understand it, this is some sort of HAM (Hold And Modify) scheme
>where the chip is put in a special mode, and certain colour values are reserved
>as "instructions" to interpolation hardware to generate colours between two
>of the available palette colours. This basically means that it is great for
>specially written CAD packages that know how to use it for ant-aliased lines
>etc., but it is not much good for general purpose use. (ie. If you want to
>be able to set any pixel to any one of 65535 colours). It also has a dynamic
>palette feature, that allows the palette to be re-programmed on the fly. One
>only has to look at the difficulty people have had in using the Amiga HAM mode
>to wonder how usefull all this stuff is. 

	BBBRRRRZZZZZAAAAKKKK!!! ...Wrong answer...

	2 bytes per pixel == 65536 colors, therefore
	no interpolation required.

	The Amiga HAM mode works with effectively
	6 bits/pixel (in bitplane layout), but achieves
	4096 (12 bits) colors via special coding of the
	last 2 bits to indicate whether the first 4
	bits are 1 of 16 colors from the palette, or
	else a modification of either the red, green,
	or blue output values only. The implications
	of this is that it may take up to 2 intermediate
	changes to get from one color to the next
	(since only one of R, G, B is changed at a
	time), so the order of changes can have an
	effect on the visual appearance. Because the
	human visual system is relatively insensitive
	to hue detail (as opposed to luminance detail),
	adroit programming can produce a very satisfactory
	result; less skilled implementations seem to
	show significant color fringing, however...


-- 
  ,u,	 Bruce Becker	Toronto, Ontario
a /i/	 Internet: bdb@becker.UUCP, bruce@gpu.utcs.toronto.edu
 `\o\-e	 UUCP: ...!utai!mnetor!becker!bdb
 _< /_	 "It's the death of the net as we know it (and I feel fine)" - R.A.M.