A Moogle  🏳️‍🌈 💻 🎮 🎶 🇸🇪

IMHO, we'll never know what the intended look was supposed to be, unless you talk to the original developers. No two monitors are exactly alike, calibrated exactly the same, and as you said how the pictures were captured could have made a difference.

The number of times I've seen an artist look at a game on a properly calibrated reference monitor I can count on one hand. In VFX it's a lot more common, especially in the final color correction stages of a movie where they look at the whole thing on...

a calibrated projector, known screen material, same lighting conditions and at the same distance relative to screen size as you would expect for the central seats of a movie theater. Everything they can control, they will. We're not there in games yet.

This is a very good and underappreciated point. It's also why in MAME we err on the side of giving the user too many knobs to twiddle with our HLSL post-processing pipeline. Do perfectly-calibrated arcade monitors exist? Sure. Were you ever likely to encounter one? Less so.

But there are knowns: The dithering found in Genesis/Megadrive games was specifically because composite NTSC didn't have the bandwidth to accurately represent it, so carefully-selected neighboring luma or chroma values could be used to force more than the normal 9bpp palette.

Totally agree. Not all differences between the un-post-processed result and the real picture are affected by external factors, so those that aren't can be reproduced (close to) exactly. I really respect your attention to detail, it must be a nightmare to make sense of it all.

Oh, it's absolutely been a group effort. I wrote the initial pipe way back in 2011, but had the assistance of someone more versed in filter theory to explain the NTSC math to me. Then the phosphor chromaticity adjustments were from yet another contributor.