It’s a legit remark, reference color for the two bright-boost values can be adjusted, so the dark-pixel bright-boost influence is reduced over the brighter spectrum.
Slotmask had one, but is removed in the new versions. ‘Trinitron’ masks 5,6,7 had some odd ways to function with lottes parameters, including brightness corrections.
New calculation for brightness correction are quite accurate, but at humble levels. If i explain, once you switch to a ‘trinitron’ mask, very dark lines that include dark pixels get their overall brightness reduced by 2x. Max compensation is 1.7x in linear space, which means 1.247x in gamma space at 2.4 gamma conversion.
To illustrate, it’s like you would multiply very dark image parts in a standard shader by 1.247. Since the brightness compensation goes away if you decrease the ‘dark’ mask value, there should be no situations to reduce bright-boost etc. in this case.
@Nesguy was fond of them quite a time ago. In general they allow you to switch between scanline and mask types without much circumstances. But they are performing at like 50-70% of the correction requirement and once you drop in the slotmask, the image gets too dark. There are rare cases when i actually needed to lower the bright-boosts.
I agree, that scanline type 2 correcture could be disturbing though. Maybe it will get less invasive or removed.
I quickly checked with a screenie and GDV keeps 0,0,0 black 0,0,0 and it also has no mechanisms to raise black levels. You could get issues though to turn very dark colors into black by multiplication. GDV has no general color substraction also.
I think it’s RA related, since the glow parameter gets a pretty direct and short appearance. The parameters can be transformed into doing smaller steps though in the shader code. If i change the code into division by 5, you can use 0.005 and it’s a 0.001 functional step for glow increment.