Tricolour additive exposures : LUSENET : Large format photography : One Thread


I am building a 5 x 7 colour enlarger and i plan to use a technique that phillips used in their PCS 130 enlarger, which had RGB lamps which varied in intensity to control colour. The filters were dichroic. Only one exposure was required.

From what i can understand, the reason that primary colours were not used by the public for printing when using one white light and filter packs is that by using a set of RBG additive filters, the first filter would block the other two colours from reaching the next filter in the pack, so therefore the subtractive process was used. However you could do three seperate exposures, red,green, and Blue when using the additive system, too time consuming for the general population.

My question is would there be a specific bandwidth of filter for red green and blue for my three colour enlarger so that all colours would be availiable, or not?. Could i just use coloured gels that they use in theatre spot lights?? also, is the only benifit of dichroic filters be freedom from fading (ie increased repeatability and life?)

Any thoughts would be appreciated



-- Phil Brammer (, July 08, 2001


When Kodak owned Wratten filters they sold the exact filters you want as CP filters, they were called color seperation filters and they are made precisely for this purpose. Try contacting Tiffen as they now own Kodak's former filter division.

-- Ellis Vener Photography (, July 08, 2001.

I think I saw/heard/read about someone using/building an LED array for this purpose. The intensity of the light could be varied, achieving an effect similar to the filters.


-- Dave Willis (, July 08, 2001.

I'm not quite sure what the advantage of the additive system is, considering the narrow bandwidth of dichroic filters and the uniformity of color of voltage regulated Halogen lamps. It seems to me an unnesessary complication, unless you are printing from or to 3- color separated materials.

-- Bill (, July 08, 2001.

Most big commercial printing machines used additive exposure (probably they still do), by either having solenoid controlled filters, or giving different times for red, green, and blue filtered lamps. The additive exposure method can give better colour saturation, for reasons too complicated to go into here.

I don't think the choice of centre wavelength and bandwidth of the filters is too critical, but if you want to get scientific about it, then film and paper sensitivity curves are readily available on the web.
A quick look at the dye density curves of Kodak and Fuji films shows that they're pretty similar. The green and red peak at 550nm and 700nm respectively for both Fuji NPS160 and Kodak Proimage 100. The blue is slightly different, peaking at 435nm for Proimage and 460 for NPS160 (but strangely; Kodak portra paper has its maximum sensitivity at ~470nm, while Fuji paper peaks at 410nm in the blue, and at 660nm in the red. Get it together guys!). A compromise of 450nm would mean less filter changing between Kodak and Fuji emulsions.

If you're going for the dichroic option, then a bandwidth of plus/minus 50nm would be a good choice. Otherwise, it means looking for some narrow cut dye filters, which won't be quite as well defined and won't last as long, but will definitely be cheaper. Lee Filters can supply almost any spectral response possible with a dyed 'gel' filter, and they'll supply free sample swatches with the spectral response curves attached.

-- Pete Andrews (, July 09, 2001.

Sorry. I meant to say "less filtration changing between Kodak and Fuji films". I wasn't suggesting you change the additive filters for each type of film.;^)

-- Pete Andrews (, July 09, 2001.

If you want to use an RGB system which doesn't waste light you should investigate dichroic mirrors, not filters. With these you divert, say, the blue light into a path where its intensity can be modified, and the green and red pas through the mirror instead of being absorbed. A second, red mirror extracts the red light in the same way and, viola, you have all the light your lamp can put out, split into three beams. Adjust the intensity of each of them seperately and then recombine with ordinary mirrors or a mixing chamber.

Edmund Optics, among others, sells mirrors with the correct bandwidths, along with 'correction filters' to clean up the individual colour beams after seperation: d=1908

-- Struan Gray (, July 09, 2001.

Moderation questions? read the FAQ