Remix.run Logo
PaulHoule a day ago

Sometimes I find interesting is how (1) screens and printed matter can’t represent greens as saturated as… a green laser pointer but (2) nobody complains because real-life greens (plants) are nowhere near near as saturated as the usual (0,255,0).

adrian_b 12 hours ago | parent | next [-]

While you are right, I find much more annoying the displays that can still be found and that are limited to the sRGB color space, or the movies or images recorded using the sRGB or BT.709 color spaces.

These color spaces cannot reproduce a green laser pointer, but what annoys me is that they are far more limited in their ability to reproduce the colors in the red-orange corner. Unlike with very saturated greens, very saturated red, orange or purple colors, which are outside the sRGB or BT.709 color spaces, are frequently encountered in flowers, fruits, gems, clothes etc.

Many good displays have a color gamut close to DCI-P3, with a much more saturated red than in sRGB. If the images or movies are also encoded using a wider color space, e.g. BT.2020, than more natural colors can be reproduced.

For reproducing the color of a laser pointer, the only solution is to use a laser projector, which also uses laser diodes for the primary colors.

For reproducing an even wider color space than possible with 3 lasers, one would have to use 4 or more lasers.

However, because the band-pass filters corresponding to the human photo-receptors are imperfect, any external source of light will excite multiple receptors, so no external display can produce the same sensation as described in the parent article, where they have used intraocular aiming of the laser spot, on individual receptors, simulating thus receptors with perfect color selectivity.

It should be pointed though, that perfect band-pass filters for photo-receptors do not have rectangular non-overlapping graphs of the frequency characteristics, but triangular graphs, where the frequency response becomes null only at the maximums of the adjacent in frequency photo-receptors, so that most colors excite two kinds of photo-receptors and the ratio between their responses allows the computation of the color hue (i.e. equivalent wavelength/frequency) of the received light.

Therefore, even for perfect eyes, unlike human eyes, pointing a laser spot inside the eyes will cause different sensations in comparison with normal vision, but in that case the sensation would not be new, but it would just show a fake color, identical to the color corresponding to a maximum in the frequency response of the photo-receptors, instead of being the actual color of the laser.

PaulHoule 10 hours ago | parent [-]

I’ve had trouble printing pictures of yellow flowers like Rudbeckia, which my camera records as 100% saturated so they are out of gamut for the camera to begin with and the yellow I do get is out of gamut for print so if I don’t bring the colors into gamut in post the printer does it for me which often looks awful.

I learned to turn on the Gamut warning in photoshop and have a choice of a few ways to bring colors into the CMYK gamut so I can make something that looks the same on the screen and print.

Blue sky in photographs are usually out of the print gamut but again, people don’t complain.

What’s bothered me the most is that when I med red-green anaglyph stereograms an sRGB red gets turned into (186,16,15) on a Display P3 monitor which includes Apple mobile devices. The excess green goes through the wrong lens to the wrong eye. This can be dealt with by attaching a color profile, but you can’t always pick one to match the device so you might attach a too-saturated profile and figure a stereogram will look weird anyway.

My bit about natural greens not being very saturated comes from the fact that grass and trees do OK in anaglyphs because they tend to the yellow side and show up enough in the red channel that it works, although now that I LUT grade my non-sports photos I am thinking of making a cube map that desaturates colors in the right way to make the left-right value balance closer.

perilunar 20 hours ago | parent | prev [-]

I don’t know about ‘nobody’ — I’ve definitely heard people (myself included) complain that screens are noticeably limited in the green part of the spectrum, and aren’t as rich as real life (or even photographic transparency film).

One thing I’ve also noticed is that screen greens seem much richer when using f.lux or Night Shift. They can’t change the screen gamut, but somehow seem better.