MAMEWorld Forums - EmuChat

Honestly, I think that's a little unfair. I always appreciate reference photos from real CRTs.

I spent about 15-20 minutes after seeing the pics, trying to tweak the settings to be "just right", but I wasn't able to make much headway.

The thing that I find interesting is that there's a bluish tint to the large ring segments in the photos, whereas there's a yellowish tint to the large ring segments in MAME.

Completely turning HLSL off, the colors are accurate to what the system is generating, but it's an interesting challenge to try to figure out why the segments appear more white with a bluish tint.

There are a few potential explanations, but I'm not sure which is the most plausible. In no particular order:

- The camera itself may have had a CCD that's more sensitive towards the blue end of the spectrum.
- The blue phosphors on the CRT itself may have been more vibrant or had a different wavelength than the model of CRT that the current phosphor values in MAME are tuned towards.
- The f-number of the camera may have been such that phosphors of a certain wavelength were overexposed (see the first point).
- There may be unaccounted-for Rayleigh scattering through the CRT glass itself.

The last one is particularly interesting to me, as it's a factor that I'm not sure has ever been accounted for in a CRT shader. Current models assume that the CRT glass is a perfect transmission medium, and while it would be ideally, nothing is ever ideal in real life. In particular, on the Wikipedia page for Rayleigh scattering, the effect of (what I'm assuming is) sunlight shining into opalescent glass is that the glass takes on a blue hue. Now, obviously, CRT glass isn't opalescent, because that would be at odds with providing accurate transmission of the phosphor wavelengths, but it's an interesting thought to chew on nonetheless.

So yeah, I think you're being overly harsh. Those photos provoke some fascinating questions and thoughts, not all of which have immediately clear answers.