Artifacting "Color"

[+mytek]

1 hour ago, DrVenkman said:

Here you go, Michael.

Looks like that did the trick! So I guess that artifact adjustment really works.

[Heaven/TQA]

hmm.... sad that this is only happeining on NTSC quite good. as the orange/blue artifacts colors are nice combo.

[Krenath]

On 3/20/2021 at 2:15 PM, mytek said:

The UAVs have a trim pot for adjusting artifacting. Might be nice to try turning that with the kaleidoscope program running to see if you can match the 800 results.

 

I've tried on my 800 with both the SCCC and a UAV I had ordered before the SCCC arrived.

With the stock CPU card and stock video hardware, all of my 800s produce blue/green artifacts.  I can get other artifact colors, but only if I completely mess up the normal colors.   If the normal memo pad/BASIC screen is blue, the artifacts are always blue/green on any of my machines' stock video hardware.

I even have an extra CPU card I ordered off eBay and it does the same blue/green colors as above.

With the SCCC or UAV, the artifacts change to purple/green and no amount of fiddling with the delay/color pots can get it back to the original 800 colors.

I've even tried modifying the SCCC to get back to 800 artifact colors, but it seems that the 800-style artifact delay is outside of the range of the SCCC/UAV color adjustment range.

[Faicuai]

On 3/21/2021 at 3:15 AM, Heaven/TQA said:

hmm.... sad that this is only happening on NTSC quite good. as the orange/blue artifacts colors are nice combo.

 

Correct.

 

Anyone playing close attention, will also notice that adjusting Color Phase or "Hue" on Video Processor on screen (LCD/CRT) should also affect rendition of artifacts, themselves. It all depends on the starting phase of LCD or CRT for hue-band $10 on Atari.

 

Modern NTSC color decode, usually renders $10h hue-band as parrot-green. By retarding Color Phase / Hue to render $10 band as "yellow GOLD" ( as it should) and adjusting color-POT to re-center $A0 hue-band (for correct blue-grayish shields on Star Raiders), I get nice Blue/Orange artifacts, on my CTIA and GTIA 800's.

 

All of the above via COMPOSITE (not RF) output.

[ClausB]

Apple II colors come from artifacts like this. FS2 was ported from A2, as were several games. Jawbreaker comes to mind...

Edited March 23, 2021 by ClausB

[Kr0tki]

On 3/22/2021 at 9:59 PM, Faicuai said:

Modern NTSC color decode, usually renders $10h hue-band as parrot-green. By retarding Color Phase / Hue to render $10 band as "yellow GOLD" ( as it should) and adjusting color-POT to re-center $A0 hue-band (for correct blue-grayish shields on Star Raiders), I get nice Blue/Orange artifacts, on my CTIA and GTIA 800's.

Faicuai, if you only adjust the colour pot in the computer, does it change the hue of the artifacted colours, for example on a black-and-white GRAPHICS 0 screen? I was always curious about it, and the GTIA specs sheet suggests that it should, but I was not able to confirm it either way.

Edited March 24, 2021 by Kr0tki

[Faicuai]

8 hours ago, Kr0tki said:

Faicuai, if you only adjust the colour pot in the computer, does it change the hue of the artifacted colours, for example on a black-and-white GRAPHICS 0 screen? I was always curious about it, and the GTIA specs sheet suggests that it should, but I was not able to confirm it either way.

To specifically address your question, I would have to fire up my CTIA "reference" 800, as my production unit now runs Sophia-II (DVI) and color-pot has no effect.

 

However, the following I can vouch for, based on my past (and multiple) efforts to run a proper calibration routine on the 800's (and my XL's too):

 

1. Screen (LCD/CRT) "hue" or Video Processor's "Color Phase" DOES AFFECT right-hand artifact-column of Altirra's COLORMAP utility, as well as hue-bands $10 to $A0, as shown in same utility (will be the same in ACP, for instance).
2. Host computer's Color-Pot mostly affects hue-ramp between $A0 and $F0 color-bands, as shown on same utility.
3. The fastest and best overall color-calibration routine I have found so far for modern NTSC color-decode, is:
   • Step 1: launch ACP or COLORMAP, and RETARD (dial negative values) of CRT/LCD/Video Processor's color "HUE/PHASE", until hue-band $10 appears YELLOW-GOLD (instead of PARROT-GREEN, as modern NTSC displays). If current video-path already decodes $10 hue-band as yellow-gold, then no color-phase (and artifacts) may be possible.
   • Step 2: adjust host computer's COLOR-POT so hue-band $A0 becomes a dark, grayish blue, with a very, very minor hint of green. It will still look bluish. This will ensure the Color-map becomes "centered", properly, and Star Raiders shields come out as originally intended.
   • Step 3: just verify, after #2, that hue-band $F0 falls right "behind" (in hue terms) or hue-band $10. This is very important, to ensure that color-ramp is "cyclical" AND final color-map does indeed show 256 color luma/hues, instead of 240 (where the majority of adjustments out there totally fail).

 

It is in STEP #1 above, where artifact-colors WILL change, when displayed via computer's Composite output (immediately visible in ColorMap utility). I have not seen major artifacts change with with just the color-pot, though.

 

 

 

Edited March 24, 2021 by Faicuai

[ClausB]

I think you're both correct. Artifact colors are caused by the ANTIC mode F pixels themselves, not by the color delay circuit. When COLPF2 is $0F or less (grays only) then the color pot will not affect artifact colors. However when COLPF2 is above $0F then the color circuit is active and will alter the hues of artifacted pixels.

[leech]

41 minutes ago, ClausB said:

I think you're both correct. Artifact colors are caused by the ANTIC mode F pixels themselves, not by the color delay circuit. When COLPF2 is $0F or less (grays only) then the color pot will not affect artifact colors. However when COLPF2 is above $0F then the color circuit is active and will alter the hues of artifacted pixels.

Does this mean in any way that we could control the artifact colors with software?

[_The Doctor__]

1 hour ago, leech said:

Does this mean in any way that we could control the artifact colors with software?

there are experiments on that combining artifact with other colors and are posted on the forums someplace. the problem still would remain that the software needs to ask what you see to pick colors based on machine 800 / XL / XE and if it's 180 degree reversed on the 800.

[leech]

21 hours ago, _The Doctor__ said:

there are experiments on that combining artifact with other colors and are posted on the forums someplace. the problem still would remain that the software needs to ask what you see to pick colors based on machine 800 / XL / XE and if it's 180 degree reversed on the 800.

Yeah, I have only seen Ultima IV do that.  Was wondering if it would be possible to create a memory resident program similar to Translator.

[Mark loves Stella]

On 3/24/2021 at 7:29 PM, _The Doctor__ said:

there are experiments on that combining artifact with other colors and are posted on the forums someplace. the problem still would remain that the software needs to ask what you see to pick colors based on machine 800 / XL / XE and if it's 180 degree reversed on the 800.

I came across this recently. ICEIRG These modes are pretty impressive displaying 256/192 with 20 colors using NTSC artifacting.  http://atarionline.pl/forum/comments.php?DiscussionID=1099&page=1

Edited March 27, 2021 by Mark loves Stella

[leech]

So here's a question!  Anyone try one of these with a VBXE to see if we can tweak the Artifacting and get the colors to work again?

 

https://github.com/hoglet67/RGBtoHDMI/releases

 

Edited May 21, 2021 by leech