/vr/ - Retro Games

>>4283156
You can load said bios through an everdrive too.

That's how this PALfag plays all regions.

>>4242425
Diamondtron is better

>>4194624
At least 40x40 I would say. Consider the dot (or stripe) pitch simulation aspect alone - take a look at the close-up Keio image I posted above and count the subdivisions, you'll need double this number for starters. Then there's the line height variation - notice how light pixels appear taller than dark ones? - this again needs extra resolution to appear convincing. Bloom and internal reflection within the glass? Yet more resolution.

Then there's the question of colour depth. CRTs generate many more shades than they are fed by virtue of these blooming and internal reflection artefacts. I'd estimate we would need at the very least 16 bits PER CHANNEL (48 total) to at least get in the same ball park, probably 32 to make it reasonably convincing.

All in all not only do you need massive display resolution and colour depth but by extension a literal supercomputer (using today's or the near future's technology) to generate the image. This is of course why even the most sophisticated shaders invariably look nothing like the displays they seek to mimic. Not necessarily "bad" as that is a subjective metric but objectively inaccurate. This will continue to be the case for some time unless there is a quantum leap in the supporting technologies.

>>4085753
Here's a wider shot (pretty much the entire screen) to judge scale.

Your shot shows the same thing except the scanlines are thinner. It definitely is the aperture grille.

>>4084752
>dcaling
scaling*

Wider shot of above