Page 1 of 1

OSI C4P Color question

Posted: Tue Jan 14, 2020 9:32 pm
by falter
Just posting this here in case it reaches anyone vcfed doesn't.

With the help of daver2 over there, I've managed to tune the color to be more or less what it should be, but one sticking point remains - both the manual and Dave's expertise suggest the background around a coloured character should be black when an 'inverted' color is used. This is not the case - if I command the machine to produce a blue tank, it produces a blue tank, but the background around the square the tank occupies is dark blue, not black. Same with every other colour - purple gives you purple tank, dark purple background in square, etc.

I have checked this in the few photos that color C4P users have posted, and in WinOSI it does this as well. Is that just how it is? Does anyone else out there have a working color C4P (or other machine with color enabled 540B video board) that could confirm this behaviour? Much appreciated!

Re: OSI C4P Color question

Posted: Tue Jan 14, 2020 11:50 pm
by Mark
When I was making the color palette for WinOSI, I connected my C4P to a video digitizer to sample the colors.
This is the image I captured:
Video capture
Video capture
osi-color-cap1.jpg (84.42 KiB) Viewed 5765 times
The color phase adjust would change the video noise beat frequency. The colors did not change appreciably. So that is what I see on real hardware.

When going over video interface notes for the 440 board and other monitor hacks, I also included and RGB based mode in WinOSI where inverted colors are black as you expect. I haven't actually wired up my OSI that way though.

Hope this helps,
-Mark

Re: OSI C4P Color question

Posted: Wed Jan 15, 2020 12:08 am
by falter
Many thanks! Okay so that's four original machines so far with color 540B boards that produce output like that. So I guess that particular board doesn't do black as suggested.

Re: OSI C4P Color question

Posted: Wed Jan 15, 2020 12:09 am
by falter
One other thing I was curious about Mark -- I see when you go into color mode it produces a solid dark yellow background. On mine it puts a whole checkerboard rainbow of coloured background blocks all over. When you were making the emulator did you actually have to figure out how the machine did that? Or did you just sample the color and tell the emulator to make it so?

Re: OSI C4P Color question

Posted: Wed Jan 15, 2020 3:08 am
by Mark
Earlier versions of WinOSI initialized the memory with 0's on start. This differs from the real OSI which did not initialize color memory on reset, so you got the color of whatever the 2114 static memory happened to power up with. I've found some static RAM maintain a good percentage of bits, even when powered off for a while. Turning on an OSI that did not power-on reset could give you a screen you could half read with stuff left over from it's last use. Other systems seem to initialize with almost the same bit pattern (mostly 0's) every time.

Later versions of OS65D or Plot Basic do initialize color video RAM on bootup. My current WinOSI version does initialize RAM with random values, and it's damn annoying as the screen is almost impossible to read when color is enabled. :-)

I think the 540 board could do black as suggested when using a custom RGB interface, but the composite color interface is as you have seen.

Re: OSI C4P Color question

Posted: Wed Jan 15, 2020 4:49 pm
by falter
Interesting. I think I got confused because the version of WinOSI I have clears the entire background to that 'orange' colour, and then when I was reading on someone's blog, they were doing the 'blue X' program, and it looked like they were operating with a solid orange background also. I hadn't been able to get a screen like his. However, what I hadn't noticed is that the manual he was using was a newer version of the one I had. The listing I had for the blue X was basically four lines, and it did nothing to clear the screen to a single color first. I see now that his did.

Okay so then I think I've got this thing fully working.. apart from tape and such.