The color purple

[pwyll]

I read something about the color purple not being a part of the natural light spectrum...

Anyway shooting (HDV) a girl in a purple dress inside - the dress looks fine but outside the dress showed up on tape as black. Loading it into my NLE that is what is is no color in it whatsoever.

Any know what's going on, and is there a way to fix it?

 

[j1clark@ucsd.edu]

I read something about the color purple not being a part of the natural light spectrum...

Anyway shooting (HDV) a girl in a purple dress inside - the dress looks fine but outside the dress showed up on tape as black. Loading it into my NLE that is what is is no color in it whatsoever.

Any know what's going on, and is there a way to fix it?

Don't know what the compression may do to specific 'colors'. Other than 'throwing away' or 'remapping' colors that the human eye would not 'detect' or 'mostly wouldn't detect'. Given that, given that under different lighting, the dress is presented with the camera's compression method...


...the thought occurs to me what was the 'value' of the dress, as in 'dark purple', or the like, and then in the 'sunlight' shots, was there something in the shot that, in order to have it the right value, pushed the 'dark purple' into 'black' (that is even 'darker' purple...).

(I have this 'dark... dark... green' Izusu Trooper... that at a distance in bright sunlight 'looks' black to the human eye... get up close, and 'the green' comes out...)... may not be quite an exact analogy... but that's the idea I have...

 

[Ryan Patrick O'Hara]

First of all, dark purple can be easily underexposed. That is one *simple* possibility. Other then that my best guess is that the camera is being blinded from these wavelengths for whatever technical reasons. Violet is the shortest wavelength in the human vision spectrum... hence the soon described 'ultraviolet' which means beyond violet.

Perhaps the camera sensor is not technologically built to see those wavelengths on the fringe... or perhaps the sunlight (which is filled with UV radiation and UV light) did an odd counterbalance to that particular dye or material. Kind of like how black clothes sometimes look like tints of green or red under near UV light. Much like the Izusu example above... i think. Those are all my guesses. Hard since I'm not a physics major nor do I know what the fabric/color of the dress was and how it exactly was reproduced on screen.

I'm just going to assume it is like the opposite of the IR effect which happens to film and video when the camera is heavily ND'd down, but the IR radiation is allowed to pass and hit the sensor. My best guess.


Interesting article on colors: http://www.theasc.com/magazine/jan05/conundrum/index.html

 

[pwyll]

I thought underexposure was the problem, maybe it was. It was a bright day - but analysing the image with FCP's scopes it's not black it's dark grey showing color values around

R 34
G 30
B 32

Comparing that to the same dress shot indoors (lit by indirect light from the windows)

R 50
G 15
B 35

I was on daylight WB the whole time. Skin tones, whites, blacks and all other colors display accurately.

I have noticed the black shift Patrick mentions with black lapels on tuxedos coming up reddish-brown on tape. Both the lapels and the purple dress were of a shiny fabric.