Help to accurately reproduce colour of Wratten-filter

Discussion in 'Open Discussion' started by Neftun, Feb 14, 2014.

  1. Neftun

    Neftun Mu-43 Veteran

    Jul 15, 2012
    Patrick Kristiansen
    I was tasked with reproducing the colours of two wratten filters on a digitally controlled led-lamp with a 8 bit colour scale.

    What I did was the following:

    The manual for the lamp stated its colourtemperature, 6500k, so I set my em5 accordingly. Took my OM90mm macro, a lens as far as I can tell, is neutral in colourrendering, and placed one the filters in between it and the lamp set to produce white, defocusing to blend its three (rgb) diodes.

    The image was imported as raw, with wb untouched, and I used the colour-measuring tool in OSX. That gave me an rgb value. Plotted this value into the lamp's UI and voila! The same colour. Or is it?

    It seemed a bit lighter than the filter's, not as saturated maybe. And I'm not sure if the colour I came up with is the same wavelength.

    Does anyone know a way to check the colour, or see any flaws to my method?

    Any input appreciated.
  2. Petrochemist

    Petrochemist Mu-43 Top Veteran

    Mar 21, 2013
    N Essex, UK
    The accurate way of checking the colour would be with a spectrometer. But I doubt your budget would stretch to one, and many models would't be suitable for external light sources. A monitor calibration device may work for you if your lucky.
    I certainly would be rather cautious of relying on the listed colour temperature, it will probably change as the lamp warms up, and is probably only quoted to the nearest 500K
    Another potential problem is the edges of colour sensitivity. The border between UV & visible as well as the one between IR & visible is not hard and fast (nor is the red/blue/green transition). it will shift a little from one person to another & one sensor to another. Your light sources are probably not just monochromatic sources variations in the intensity of light that one system sees but the other doesn't will effect the colour experienced.
    • Like Like x 1
  3. Neftun

    Neftun Mu-43 Veteran

    Jul 15, 2012
    Patrick Kristiansen
    So what you're saying is that there are plenty sources of error here, and summed up my result may deviate significantly. Figures.

    As for spectrometers, I recall in my old school's science lab there where these simple spectrometers that worked kinda like binoculars, you just put them to your eyes and pointed it towards a lightsource. Showed an analogue readout. Didn't seem expensive to me, gonna check it out.

    Do you know of a lightsource that has a reliable given colourtemperature? Or if the custom wb / one shot wb on an oly camera can be reliable in an semi-scientific context?

    Thanks for your reply btw:)
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.