nesdev.com

Should not be using that color. (Search the wiki/forum.)

_________________
Download STREEMERZ for NES from fauxgame.com! — Some other stuff I've done: fo.aspekt.fi

Yeah, some emulator palettes will have an extra dark colour for $0D, or rather all the other blacks are raised. A few people like this because their TVs also do something that looks similar.

Unfortunately, on a lot of TVs (both new and old) it will fuck up the blanking signals and cause the picture to wobble, or completely disappear. Even if the TV can handle it, on many it won't actually be darker.

There's a bunch of possible different behaviours for $0D, really, and a lot of them will make the game unplayable for the user. It really isn't a good idea. (This is why my palette test ROM doesn't show $0D by default.)


There's also the two "extra" greys in the $XD column, if you need a few extra levels of grey. These do have slightly different values than the greys in the $X0 column, but some emulator palettes seem to make them identical, which is a little bit sad. Most games seem to just avoid that column anyway, maybe because it becomes an annoying special case for fades.

That's a bummer. I was under the impression that since it was used in a number of commercially released games that the compatability issues would have been a non-issue. Will keep that in mind.

_________________
www.mteegfx.com

Did you really mean $0D (the invalid black), though, and not $2D (the dark gray)? I don't really see how $0D would help your dark scene.

$2D is fine to use, although it will be black in the official Nintendo RGB PPUs (not widely used) and maybe some emulators(?). One more option is to set all emphasis bits to darken the screen further (this, also, will not work on Nintendo's RGB PPU).

_________________
Download STREEMERZ for NES from fauxgame.com! — Some other stuff I've done: fo.aspekt.fi

0D is actually darker than 0F (apparently not on all TVs, though?). I'm aware that it was suggested against due to some form of issue. But, with The Immortal and some other games using it, I figured it wouldn't be game-breaking or a big deal.

If it can be gamebreaking for anyone, I guess better to avoid it.

_________________
www.mteegfx.com

Good, because it's insanely cool.


Some TVs really don't like it.

Thanks for the link, the explanation, and for bringing it to my attention, all. That's a shame. I was really looking forward to it, too.

_________________
www.mteegfx.com

If I were to try and improve it, I'd make the side of the bridge thicker (it doesn't make sense how it's as thick as the back because of perspective and that you should be seeing the side of the road, not just the guardrail) and I'd make the skyline palette and the gradient different. Other than that, it looks great.

Do you have an image of it on your computer though, like a screenshot from an emulator? I want to mess with it to see what you think.

I see what you mean about the outer height of the guardrail. I do believe that I have some space to alter that without having to change one of the heights of the scroll breaks. I might go back and do that. As for the gradient, what's shown is somewhere in the middle of a series of palette & emphasis transitions for sunset, and the colors are heavily washed out on that video. I'm satisfied with that aspect as it is.

Edit: as for the screenshot, am on my phone and don't have one immediately available. But I get the gist of what you're saying. Thanks for the offer, though.

_________________
www.mteegfx.com

I'm aware that #2D is roughly all over the place between emulator palettes, but it's a bit darker than #00 through my NTSC CRT on a Famicom, so I decided to go with #2D for my darkest grey and use all three emphasis bits. I was worried that it wouldn't be dark enough for the intended effect, but I think it'll do fine:



If anyone else can chime in as to whether #2D is darker or lighter on their setups than #00, that'd be helpful.

_________________
www.mteegfx.com

It's darker. There's a video capture from my NES in my palette test thread, if you want an example.

On FCEUX's default palette it's the same, though, which sucks.

A while back I had lidnariq measure the output levels so I could fix the IRE brightness test in 240p test suite.

To oversimplify: 1D 2D 00 10 20 = 0.0 0.3 0.4 0.7 1.0

A color ($x1-$xC) would presumably be the same luma as the average between $x0 and $xD. The difference between $1D and $0F is that $1D is affected by emphasis, while $0F is not.

Thanks so much for the palette feedback. We're trying to decide whether it'll be better to black out everything but the cursor, or to use #1D with all the emphasis bits to get that darkened room effect.

In the meantime, I've put together a splash screen for the game.


It was really interesting to do this because the KHAN Games logo contains a lot of colors:



(I know splash screens are a divisive subject. Personally, I love them, and enjoy making them, so any "I don't think games should have splash screens." posts will be fruitless. )

_________________
www.mteegfx.com

It looks nice.
It's not splash screens I hate so much as a "minimum hold time" for them.

Sometimes a licensor demands a minimum hold time as a condition of allowing the game to be published at all. Nintendo's Tetris for NES, for example, shows the "Copyright Elorg" screen for a minimum 256 frames, then 256 more frames or when Start is pressed, whichever comes first. (GB version uses 250 + 250 according to NieDzejkob.)

And in games on disc-based platforms or games for cartridge platforms using CPU-bound decompression, the minimum hold time covers loading. I've read Super Mario World does a bunch of decompression while "Nintendo Presents" is showing. And behind "MARIO START!" in both SMW and the Super Mario Bros. 3 port in Super Mario All-Stars is a bunch of SPC loading.