Blur Busters Forums

I was messing around with CRT Beam simulator via shader glass on my XL2720. My monitor can only handle a 1092 vertical total at 180hz. I have been utilizing the phosphor fade BFI to great effect, enjoying panning shots at 1920 pixels per second. This has been massive for enjoying my monitor.

Anyway, I was fiddling with settings and saw that "Scan direction" on the shader glass alpha either at a setting of 1 and then at a setting of 4, gave me either horizontal beam scanning, or a vertical beam scanning. This allowed me to eye track at 3000 pixels per second on the display, vertically or horizontally (depending on the setting) but I would get double image trailing ghosts with rapid eye movement in the direction opposite the eye tracked object.

When I noticed this, it gave me a thought.

Could you do beam scanning, via the shader in a checkerboard pattern, going from top left to bottom right corner scanning horizontal and vertical simultaneously?

The effect I was seeing visually in terms of panning speed was great either vertically, or horizontally, but not both, and I figure since this is a shader, shouldn't arbitrary scan characteristics be possible?

Interesting, but I'm not sure that's what you're really going to want.
I mean, there's nothing preventing additive scan beams. The main issue is that, let's assume a 4x scan to frame ratio, you will only show 1/4th of the screen per scanout. That scanout pattern will be four quarter boxes in whatever diagonal you choose.

Obviously that's not impossible to mitigate or at least partially circumvent. One approach would be to cycle one direction relative to the other. e.g. For a 4x ratio in one direction, have a sum 16x total for the other. (4x * 4x) This would allow you to scanout each part of the screen at that ratio. It would lower persistence by the square of that ratio. That means far lower brightness. I would also say lower motion blur across the board but the issue is temporal misalignment based on which part of the screen you're looking at.

Because of that lowered persistence you could reduce the ratio to compensate but I really don't think the result would be any less artifacty than a single direction. Again, that boils down to temporal misalignment. CRT scanout adds that factor in order to mitigate the full screen flicker of normal BFI/strobing. There's no free lunch.


Ultimately, there is simply no replacement for framerate. Framegen is the closest we have but even that can't be perfect.