[Coco] CoCoMax (was) Re: Introducing myself - Carlos Bragatto - Now: OT Amiga / CoCo musings
jcewy at swbell.net
Sat Apr 7 16:20:21 EDT 2007
> I have both an Amiga 2000 and a 1200 - and I didn't know there was a
> connection between them and my CoCo. It seems the connection is a
> little tenuous, but it is there, nonetheless.
> For the following, realize that I am talking out my rear, but I am
> pretty sure I am close with the information - correct me where I am
> wrong, and don't flame me too hard.
> I don't think the CoCo Max 3 palette swap trick and the Amiga HAM mode
> have anything to do with each other, personally. For one, the HAM mode
> was merely using the hardware (mainly the Copper, IIRC) to do
> something it was designed for (but not documented well?), but in a
> unique way. When an image was displayed in 4096-color HAM mode - it
> was a solid image - no artifacts, no noise, nothing. Your only
> limitation was that you couldn't do any animation (and I would be
> willing to bet, though I never saw an example, that this limit could
> be overcome by someone willing to do it - maybe a democoder). The
> "super HAM" mode on the 1200 and 4000 worked in the same way.
The Dynamic HAM mode I'm referring to (DHAM) is actually a combination
of the Amiga's HAM mode with per-line palette stuffing. Even in HAM,
the Amiga's palette is limited to 16 colors, it's just that you can
specify that certain pixels deviate from the color specified in the
palette register by a specific amount. DHAM allows you to select the
best palette for each scan line to represent the colors in that portion
of the image. Then you can HAM it up from there. I think it even used
the Copper to update the palette registers without assistance from the
CPU. But this extension of the HAM mode does employ the same technique
of updating the palette registers. There was a similar technique called
Sliced HAM (SHAM), but I don't remember the exact details of how it
differs from DHAM. But I think that the DHAM technique was 'invented'
at NewTek, which was founded by Tim Jennison, who had done the original
CoCoMax. So that was the possible connection I was wondering about.
> The palette swap trick on the CoCo 3, though, used a technique with
> the hardware to do something it wasn't really designed for, let alone
> documented. Somebody saw something in the hardware, a bulb lit up, and
> they tried it - lo and behold it worked. But it had noise (that junk
> on the left edge) - the video hardware wasn't designed to do this,
> unlike the custom chips in the Amiga.
I'm not sure whether updating the palette registers during screen redraw
was actually envisioned or intended by the designers of the Amiga's
chipset, though the inclusion of the Copper certainly helped make that
kind of thing feasible.
> As far as whether it had been done earlier? Not sure here - when was
> CoCo Max 3 first advertised? I have in my hands here the January 1987
> issue of the Rainbow. On page 20 (continued on page 24) is an article
> by Rick Adams and Dale Lear (are they on the list?) entitled "Color
> Chart for the CoCo 3" - which is also a 64 color chart. Interestingly,
> on pages 22-23 (in the "middle" if you will of the article) is a
> Colorware CoCo Max II advertisement - with no mention of CoCo Max 3.
> So - did these guys do this trick before CoCo Max 3 - and it was
> incorporated into it. Or, did development (and idea) paths cross (ie,
> CoCo Max 3 was in development at the time the article was released,
> but because it was a proprietary product, Colorware wasn't going to
> release that special code, and so someone else beat them to it)? I
> don't know. Does anyone?
Good question. I do know there was a little while between the
introduction of the CoCo 3 and the availability of CoCo Max III, which
is why I patched my CoCo Max II joystick interface to work on the CoCo 3.
> Finally, at the end of the article, the authors write: "You may have
> noticed this effect on some video games on the Atari VCS, which uses
> this same technique to expand the number of available colors."
> The "effect" they speak of is the "noise" on the left hand edge of the
> screen. They are right - this was done on the Atari 2600/VCS, for
> certain games. The only one I can think of off-hand that did this was
> the Battlezone cartridge (which, IIRC, was released when the tiny
> updated Atari 2600 system was released in the 1980s? - BTW, the
> cartridge works great in my vintage 2600) - but there are probably a
> few other examples.
> So, obviously, this kind of a trick is pretty well known, and if you
> have few enough colors and enough processing power, you can even get
> motion on the screen. Hmm - makes me wonder if that is how Sock got
> some of his amazing demos to run? I have always thought that there
> should be a way to use the "64-color mode" (or some variant of it,
> maybe with less colors) in a graphic demo or game, or something. Sock,
> can you shed any light on this? I know your hi-color programs use a
> variant on the trick (it, combined with some form of pixel
> offset/rotation trick like NTSC uses to cut flicker and increase
> colors displayed)...
> -- Andrew L. Ayers
> Glendale, Arizona
> Message: 8
> Date: Fri, 06 Apr 2007 17:23:58 -0500
> From: Joel Ewy <jcewy at swbell.net>
> Subject: [Coco] CoCoMax (was) Re: Introducing myself - Carlos Bragatto
> - Now: OT Amiga / CoCo musings
> To: CoCoList for Color Computer Enthusiasts <coco at maltedmedia.com>
> Message-ID: <4616C87E.1020407 at swbell.net>
> Content-Type: text/plain; charset=ISO-8859-1
> Gene Heskett wrote:
> >> On Friday 06 April 2007, Joel Ewy wrote:
> >> ...
> >>>> Don't forget that the Tim Jennison of NewTek fame is the original
> >>>> of CoCoMax. That, if nothing else, is some direct
> >> I hadn't been aware of that! Talk about cross-pollination, that's
> >> carrying the flowers to each other with introductions. To heck
> with the
> >> bees. Now I have to give credit for the history lesson, thanks Joel.
> >> Now I'm sorry I never bought a cocomax kit.
> I'm not sure what the time relationship was between CoCoMax III's
> palette-swapping routine and NewTek's Dynamic HAM. (Of course, by this
> time Jennison was doing Amiga stuff, so any connection would be
> indirect.) Probably the idea of re-loading the palette registers
> multiple times per screen isn't original with either the CoCo or the
> Amiga. Anyone have any idea where it was done first?
> Coco mailing list
> Coco at maltedmedia.com
More information about the Coco