Product: Camera: Color
RGB Flash Camera   (+12, -4)  [vote for, against]
B&W Sensor + Colored Flashes

Take a digital camera with a black and white image sensor, and add a red flash, a green flash, and a blue flash.

Whenever a colored image is desired, take 4 pictures in quick succession: one with no flash, and one with each colored flash.

Subtract the "no flash" image out of each monochrome-flash image, and then combine the three monochrome-flash images together to make a single colored image.

Because the image sensor doesn't have built in colored filters, it will be cheaper to make, and more sensitive to light.

Naturally, the camera can still be used as a regular black and white camera (with or without a flash; if a flash is desired, all 3 colored flashes would be used together).
-- goldbb, Jun 01 2009

Cool, but would only work for static subjects. I'm going to bun despite the link that shows the idea isn't completely original. (+)
-- MisterQED, Jun 01 2009


Unless you photograph in the dark this would be impossible.

Why not just expose the film simultaneously onto three different areas each behind a filter? Then you don't need multiple exposures, processing is easier and you can shoot in the light. Your idea is ridiculous by comparison.
-- WcW, Jun 01 2009


I'm bunning this for sheer impracticality, and because [21Q] boned it.

//Unless you photograph in the dark this would be impossible//
Not at all, given a sufficiently insensitive film, and impractically bright LEDs.
-- coprocephalous, Jun 02 2009


QED, the subjects don't necessarily need to be static, just not moving very fast :) After all, since it's a digital camera, the four images would be created in very fast succession.

WcW, why is it impossible unless done in the dark? After all, background light (as identified in the no-flash image) will be subtracted out from each monochrome image.

Also, what film are you talking about? My idea is purely about a digital camera... Hmm, you must be talking about Ian Tindale's sandwich photo... he used B&W film, and three filters, one red, one green, one blue. Since the particular filters he used (Hoya "Pop Art" filters) are designed to go onto the camera, I would assume that that's where he placed them, and the flash he used (if any) was white.

I'm intentionally using *no* filters, to increase (instead of reduce) the amount of light reaching the image sensor. This allows a relatively dim flash to be used. The total amount of light from the three colored flashes would be about the same as you'd get from a conventional white flash.

Given that Ian was able to create a technicolor image using a film camera, I suppose that you could use film (instead of an image sensor) to implement my idea... but it would *definitely* have to be a still image, and you'd still need to take four images (not three).
-- goldbb, Jun 02 2009


it's so much more intuitive to place the filters in front of the film rather than in front of the light source. I don't think that your unfiltered subtraction is going to work. Also how is the final combination of channels achieved? Do you process it three times in three different photosensitive dyes?
-- WcW, Jun 02 2009


even in ambient light?
-- WcW, Jun 03 2009


who killed my anno? and why?
-- loonquawl, Jun 03 2009


//who killed my anno? and why?//
<Hercule Poirot mode> And so, we conclude, mon ami, that there can be only two possible suspects in this case of annocide. M. [goldbb], who for reasons best known to himself, took it upon himself to delete the anno of M. [WcW], who, being blackmailed by the illegitimate daughter of his housemaid Ruby,( and who was, incidentally, herself the sister of M. [goldbb]'s butler's valet), decided to commit the dreadful deed himself, in the library, with the delete key.

</hpm>
-- coprocephalous, Jun 03 2009


<Inspecteur Clouseau mode>An anneaux you say? [trips over annotation, falls down 4 flights of stairs]</icm>
-- zen_tom, Jun 03 2009


<Lt. Columbo mode> Oh before I go, there is just one more thing on my mind... <puffs cigar>
-- theleopard, Jun 03 2009


<Lt Frank Drebin mode> Nice Beaver! </lfdm>
-- 4whom, Jun 03 2009


WcW, if we're talking about digital, the subtraction is trivial. For that matter, combining the layers digitally is also trivial.

If we're talking film, then we would take the negative made by the unflashed shot, and use it to expose another piece of clear B&W film, making sure to keep the size and alignment exactly the same. Then develop it. This will be used as a filter, to subtract out the ambient light.

Having done that once, repeat two more times.

Take one of the anti-ambient-light filters, and carefully align it to the B&W negative made with the red flash. Glue it in place to ensure that the alignment stays correct. Repeat with the other two negatives.

Next, glue red, green, and blue filters to each of the assemblies you've just made.

Next, expose some color photographic paper, using each of the three color layers you've assembled, one at a time.

Then, develop the color photo paper.

(It could probably be done without glueing film layers together, but then alignment becomes insanely difficult).

Naturally, this is ridiculously impractical, earning coprocephalous's bun twice over :)
-- goldbb, Jun 03 2009


Canon's LiDE scanners do just this - the lighting element is an RGB LED array. Of course, they have the advantage of operating in darkness, on a static subject, the desirability of which has already been mentioned by my co-bakers.

When I used to design high resolution scientific CCD cameras, we made a colour camera by placing an electronically tunable filter in front of a monochrome CCD sensor. The colours weren't very good, I seem to remember, but I daresay they've been improved since then.
-- English Bob, Jun 04 2009


Interneg, hmm...

I'm not sure what that is, but I do know that "inter" derives from the Latin meaning "in between" or "among", while "neg" I presume is an abbreviated form of "negative", which of course comes from the Latin "negativus", a term that i 1853 was appropriated for the photographic sense.

In or among negatives. Ah, now I see!

Hurrah!

(1686, alteration of huzza, apparently infl. by similar shouts in Ger., Dan., Swed. May have been picked up during Thirty Years' War. According to Moriz Heyne, this was the battle-cry of Prussian soldiers during the War of Liberation (1812-13).)
-- theleopard, Jun 04 2009


<dead anno simulacrum resurrection mode> As [English_Bob] noted, the process works best in the dark; interestingly, the substraction process will ensure that the pictures will always look as if taken in the dark, as any light coming from ambient sources is filtered out by the substraction. A nifty way to make any photo look as if taken without ambient light.</dasrm>
-- loonquawl, Jun 04 2009


There's an advantage to making a photo which looks like it was taken in the dark, with the flash as the only light source: specifically, it completely eliminates variations cause by different colors of ambient light.

If you're photographing a person for identification purposes, and you use this technique, then all photos of that person will look about the same, regardless of sunlight, clouds, moonlight, incandescent lights, flourescent lights, etc..

For that matter, since once could easily get a wide variety of differently colored flashes, you could do something of a spectral analysis of a person... with flashes that are infrared, yellow, cyan, voilet, or ultravoilet. These might not be easily compositable to create a "color" picture, but they still might look interesting (especially if someone has a tatoo in not-normally-visible flourescent ink).
-- goldbb, Jun 04 2009


The 'beyond RGB' possibilities occurred to me too.

Monochrome LEDs are currently amongst the most efficient light sources available; there is surely no call for filters on the flash!

The icing on the cake would be if you could include, in addition, a lens filter assembly to allow ambient light colour pictures. It could even be after market, and driven by the same circuits as power your RGB flashes.

The fact that the spacial location of the colours is not offset within each pixel is yet another advantage. (+)
-- spidermother, Apr 07 2011



random, halfbakery