v0.40.0-preview

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

JetSetIlly

I've written comprehensive release notes on the github page. But just to emphasise, this is a "preview" release of some major changes.

It's a "preview" because there are some changes that have not yet finalised - the default colour palette. However, I think the project would benefit if people can have a look at it now.

https://github.com/JetSetIlly/Gopher2600/releases/tag/v0.40.0-preview
https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

JetSetIlly

Has anyone had a chance to look at this? I still have a question about the precise colour of hue 1. In other words, what exactly is the "gold" of hue 1. Everything hinges on that really and I don't have an NTSC TV to check.

I'm fairly sure I've not got it exactly right but that's all I know for sure.

There's an ongoing conversation on AtariAge on this subject but it's going nowhere frankly. A lot of opinions but not a lot of facts.
https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

JetSetIlly

#2
The simplest maths produce what we see in the screenshot. The thing is, hue 1 doesn't look like "gold" to me.

So I've convinced myself that hue 2 is gold and the current version of Gopher2600 has a hue 1 more like the hue 2 we see in this image. In other words, the hue has been shifted by about 24 degrees.  But I'm pretty sure that's not right - I've deluded myself.

Can somebody advise me one way or the other which "gold" is correct? As I say, I simply don't have an NTSC system to check.

https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

Thomas Jentzsch

I had a greenish hue 1 too. So I tweaked the angle until the hue 1 did fit.

JetSetIlly

Quote from: Thomas Jentzsch on 16 Jan 2025, 08:14 AMI had a greenish hue 1 too. So I tweaked the angle until the hue 1 did fit.

Right. That's what I've done too in a way. But we really wanted to know what is produced by the console before the television controls are adjusted.

I want a HQ photo of the gold as it appears on an NTSC system, with the TV controls set to neutral.

Looking at the photo (not high-quality) in this post and hue 1 looks green to me. So I honestly don't know what to think anymore :-\

https://forums.atariage.com/topic/375698-how-are-ntsc-console-colors-really-set-up/page/5/#findComment-5599307


I have a PAL machine here and my calculations recreate what I see on the hardware, so I'm reasonably confident I'm getting the maths correct. It's just that damned colour burst on the NTSC that I can't decide about.

https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

Thomas Jentzsch

Yes, that one looks greenish. But the old documentation describes it as gold. So it cannot be green (unless the doc is wrong).

As you know, the discussion is still ongoing. I tend to follow the old doc, but that's subjective.

JetSetIlly

Quote from: Thomas Jentzsch on 16 Jan 2025, 09:43 AMYes, that one looks greenish. But the old documentation describes it as gold. So it cannot be green (unless the doc is wrong).
But it's the colour burst, it can't ever be different. That's what's confusing me.

The only way the colour burst can appear different is if the TV is mistuned (or a faulty console I suppose). I doubt Strum has his TV mistuned. I'll ask him.

QuoteAs you know, the discussion is still ongoing.
I think that discussion has fizzled out.

TBH, I'm beginning to think the only way I can get the answers I want is to acquire an NTSC machine myself.
https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

Thomas Jentzsch

Quote from: JetSetIlly on 16 Jan 2025, 10:29 AMBut it's the colour burst, it can't ever be different. That's what's confusing me.

The only way the colour burst can appear different is if the TV is mistuned (or a faulty console I suppose). I doubt Strum has his TV mistuned. I'll ask him.
Could be anything, I wonder how he will explain.

QuoteI think that discussion has fizzled out.
I hope not.

QuoteTBH, I'm beginning to think the only way I can get the answers I want is to acquire an NTSC machine myself.
Not sure if that will help. Probably you need more than one to even get an idea. Plus multiple CRT TVs.

JetSetIlly

#8
Quote from: Thomas Jentzsch on 16 Jan 2025, 07:17 PMNot sure if that will help. Probably you need more than one to even get an idea. Plus multiple CRT TVs.
The immediate problem I want to solve is the natural colour of hue 1. Based on the information we have, I don't believe hue 1 would ever change between consoles (the adjustment pot doesn't change the output of hue 1 -- according to everyone who has answered so far).

The purely mathematics answer is that hue 1 is that weird greeny-yellow. But as you say, that doesn't match the descriptions in the old manuals. The only thing I can think of is that when Atari said "gold" what they mean is, "tune your TV so that hue 1 is gold", rather than, "the natural output of hue 1 is gold".

But there are other possibilities perhaps. How does the console produce the colour burst? According to the Stella Programmer's Guide page 18.

   // "A hardware counter on this chip produces all horizontal timing (such as
   // sync, blank, burst) independent of the microprocessor, This counter is
   // driven from an external 3.58 Mhz oscillator and has a total count of 228.
   // Blank is decoded as 68 counts and sync and color burst as 16 counts."

Does that tell us anything useful? I'm not sure.


The issue of what phase the hues are I've more or less settled on. I think it's the "Field Service Manual". I don't think we'll get a better answer than that, despite it being a very vague description. 24degrees is right out. Despite how perfect an answer it might be, it's just wrong.
https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

JetSetIlly

Quote from: JetSetIlly on 16 Jan 2025, 09:33 PMBut there are other possibilities perhaps. How does the console produce the colour burst? According to the Stella Programmer's Guide page 18.

   // "A hardware counter on this chip produces all horizontal timing (such as
   // sync, blank, burst) independent of the microprocessor, This counter is
   // driven from an external 3.58 Mhz oscillator and has a total count of 228.
   // Blank is decoded as 68 counts and sync and color burst as 16 counts."

Does that tell us anything useful? I'm not sure.
I suppose I'm wondering why the Programmer's Guide is telling us that the colour burst happens on 16 counts. Unless I've missed something, that's not information that's useful to the programmer.
https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

Thomas Jentzsch

16 counts = 16 hues?

alex_79

#11
Quote from: JetSetIlly on 16 Jan 2025, 11:15 PMI suppose I'm wondering why the Programmer's Guide is telling us that the colour burst happens on 16 counts. Unless I've missed something, that's not information that's useful to the programmer.

Knowing the length of the color burst is of no interest for a programmer, but that part is from the document called "TIA 1A", which is a hardware manual and contains some more technical oriented info such as timing diagrams. This was a separate document from the actual TIA Programmer's Manual, but they are both included in the reconstructed PDF and HTML versions that's been floating around for many years.
The "Programmer's Guide" used in that case was the '79 version, but there are scans of others versions floating around.

In the "Jan" manual (the combo chip that includes 6507, RIOT and TIA in the same package, used in some "Junior" models), there's this paragraph about color phases:

QuoteThe color phase shifter produces a reference color output (color burst) during horizontal blank. Then during the unblanked portion of the line it produces a color output shifted in phase with respect to the color burst.
The amount of phase shift determines the color and is selected by the four color code lines from the color-lum multiplexer. Code 0 selects no color. Code 1 selects light orange (same phase as color burst). Codes 2 through 15 shift the phase from zero through almost 360 degrees allowing selection of 15 colors. Each phase shift is approximately 20 ns from the reference color output. Two phase shifts would have a delay of approximately 40 ns.

The TIA manual from COLECO (in which the TIA is actually referred to as "VIC", for Video Interface Circuit) defines color "$1x" as "YELLOW-ORANGE" (page 65)

Thomas Jentzsch

QuoteCodes 2 through 15 shift the phase from zero through almost 360 degrees allowing selection of 15 colors.
Almost 360 degrees? We are running in circles.

JetSetIlly

Quote from: Thomas Jentzsch on 17 Jan 2025, 04:48 AM
QuoteCodes 2 through 15 shift the phase from zero through almost 360 degrees allowing selection of 15 colors.
Almost 360 degrees? We are running in circles.

Heh

The attached table on Page 38 of the JAN programming guide, seems to support the VideoSoft test cartridge calibration instructions. "Light Orange" for Hue 1 and Hue 15

https://github.com/JetSetIlly/Gopher2600
@JetSetIlly@mastodon.gamedev.place
@jetsetilly.bsky.social

Thomas Jentzsch

#14
What a mess! Also color 14 is now "gold".

Did they change this for a reason? Or was there just total chaos?