A triumvirate of British computers

[martinw]

That’s really cool thank you for explaining it

Martin

[Lardo Boffin]

I ended up doing the Tim Worthington modification to get my 2600 to output RGB as I like crisp pixels.

Anyhow, enough off topic - back to Sinclair…

Sorry to drag it back off topic again - I have been thinking about doing that to my 6 switch woody. Is it worth it? Its not cheap!

[mikeh_nz]

Is it worth it?
I think so. I’m guessing my wife would think otherwise…

Summary
If I was doing it again, I would still do the upgrade. I would still feel conflicted about the expense and I’d hee & haw over it. But it’s sufficiently worth it for me, & I’d pull the trigger again.

Probably More Details Than You Wanted…
I found doing the mod and also assembling the SCART cable a fun project. I do like sharp edges of a modern display - some will prefer the nostalgia of blurred edges & think the sharp edges make it look like it’s fake & you’re running on an emulator (fyi there’s an emulator for the Atari 2600 called Stella - you could probably run that and get a feel for the output you’d get)

Since I couldn’t find anything with RGB input here in the states (no SCART on the back of the TV/monitor), I also ended up going with a Retrotink 2xSCART (upscales SCART Rgb to hdmi with minimal latency - but also friggin expensive)

If you were going straight from RGB to a monitor/tv (Eg possibly through the SCART plug), the display might have trouble with the low number of scan lines / refresh rate - but probably not.

Also the Atari 2600 is known for its output of an inconsistent number of scan lines in some games (Eg Empire Strikes Back (when you kill an AT-AT) and Battlezone). This can cause vertical sync to be lost at points in the game. However, I’ve only found this in a few games. The solution is to use a version of the game that has been tweaked to remove the timing problem.

Really in the weeds
The inconsistent scan lines / v sync timing happens because the game code is literally in charge of feeding the video chip what to paint as it draws the scan line across and down the screen - there’s no screen buffer in memory that the video chip pulls from. (Search for videos with “Racing the beam”). This means that most of the time the game is spending its cycles just constantly feeding the video chip (with perfect timing). Only during the scanlines off the top/bottom of screen and h sync & v sync fly back does the poor game get any time to compute moving enemies, processing your input, etc. While the video chip handles triggering the Horizontal Sync, it’s up to the game code to trigger the Vertical sync. Some games don’t do this with consistent timing, resulting in a loss of V sync on the display.

Given that the code is so tightly coupled to painting the screen, it results in distinct pal & ntsc versions of each game.

NTSC: 262 scanlines (3 vsync, 37 vblank, 192 picture, 30 overscan)
PAL50: 312 scanlines (3 vsync, 37 vblank, 242 picture, 30 overscan)

Im very impressed by someone who writes an Atari game from scratch…

[mikeh_nz]

I never checked out the clean comp on the Atari 2600 as it wasn’t available at that stage. That might be a good compromise.

[Lardo Boffin]

I never checked out the clean comp on the Atari 2600 as it wasn’t available at that stage. That might be a good compromise.

Thanks for all the info! Currently I run 2600 games on an s-video modded 7800 which produces a great picture but really like the idea of getting a good picture out of my old 2600.

I have the book Racing the Beam. Superb read!