which colormeter to use ?

[56 DIN]

i need to measure the temperature of LED light, composed by the use of RGB diodes. I want to create daylight, warm white and candlelight effects as well as variable colors themselves with the diodes. The colors are sufficient to compare to test charts, but the initial installation of light ( intensity of the 3 diodes ) and "white" temperatures should be done more precise. Has anybody of you used a Minolta, Kenko or Gossen colormeter and can give me some adice ? The 3 diodes are free to manipulate in 255 steps each and current is also adjustabel for the channel if necessary, as i expect to run red with more amps than blue and green to get the same cd values
thanks in advance

[poilu]

[56 DIN]

thanks poilu, but it doesn´t help me here - i need to compose the light out of the 3 sources and i wanna know, if somebody in the forum has experience with these photography related colormeters to tell me if they measure the LED spectrum accurate enough or not
otherwise i have to invest much money or stop the project. LED does create other spectral peaks than fluorescent lamps and a lot of elder colormeter are not calibrated to go with LED

[Orio]

I have the Gossen colormeter, it works, but I doubt it gives the exact values that you are looking for.
It's a pre-digital instrument, works with needles not with digits, and it's purpose was to give a guidance
for choosing the colour filters for film, not really a high precision task.

[poilu]

my x-rite give precise temperature but I don't understand why you need more precision than a Nex

[Sevo]

More affordable photo colorimeters only did a blue/orange colour balance evaluation. The last Minolta/Kenko or two can additionally do a three colour check - this will usually be good enough to determine appropriate compensation filters for three layer colour film, but it is not good enough to evaluate the tint on a discontinuous light source. You really need a spectrophotometer - these aren't rare (and often not that expensive used) as they are quite common print shop and lab equipment.

[56 DIN]

thanks for your answers - i think Sevo pointed that out, what i was afraid of. There will probably be a step necessary into a 1k€.- measurement equipment plus software for my notebook - hm - no decision made so far, cause i do not print my own money.
To get a further idea i will borrow a colormeter adapter for my Gossen, like Orio has described it and check, if it helps me for the beginning. The color checkers like spider etc are for low emmission screens or printing, i am not shure if they do well "against" a LED to illuminate a room

[Arkku]

Take a raw photo of a photographic grey card lit by the LEDs, use raw processing software to set the color temp from the grey card. This will give the perfect result where photographic accuracy (with the same camera, lens, and workflow) is concerned. Even if some other meter is more accurate by some technical measure, I would argue that for photographic use this is the best way.

That said, due to spectral peaks it may not be possible to fully simulate arbitrary color temperatures throughout the visible spectrum with the described set-up. Perhaps if you had many RGB LEDs with different RGB balances in each you could even out the spectrum.

(A possible link of interest: http://www.deadbread.com/crumbs/spectro.html )

[56 DIN]

Arkku,

thanks a lot for this link - it will really help me to define the start point of my tests. After doing that i probably will end with a more specific and cheaper equipment than i was afraid i have to buy