View Full Version : Autochrome RGB Spectrums via Modern dichroic filters

Mustafa Umut Sarac
07-04-2010, 01:44 PM

Autochrome RGB Spectrums above

Thorlabs Dichroic Filters Spectrums below

How do you campare them , What do you think ?

Thank you Charles , Thank you very much.







07-14-2010, 12:57 PM
Any conclusions yet?

I'm wondering how one could use this information to deduce whether my TV/monitor experiment is approaching the same colors as an autochrome (or similar). I suppose I could take a macro shot of the monitor and perhaps the patches would then be large enough to analyze?? Of course, I have absolutely no means by which to do so (the analysis, that is).

07-14-2010, 01:09 PM
The blue band is a bit wider (towards the red end) in the Autochrome.
The red dichroic filter comes closest at 0 degree, else is a bit wide towards the blue, compared to Autochrome.
The green dichroic is a bit narrow, too little on the red end of the band, compared with the Autochrome.

Why the cyan, yellow and magenta filter graphs?

Mustafa Umut Sarac
07-14-2010, 01:21 PM
I think answer is simple. Do you remember my Getty Museum post ? If you want the exact autochrome match with the color quality of your pictures , you must use same chemicals , inkjet printer and a transparent medium you can put your emulsion or film one side , ink jetted dots to the other side.
If you are not purist , you can continue to your research. The basic other answer is to spectrum analysis your monitor generated RGB dots on your film.You can predict the results and compare with top french image .Of course its depends the gear of your analysis lab. May be switching between different films could help you to close to something. I learned you have no photoshop. Visit

www.gimp.org and download your photoshop similar program.

And there are some film recording machines which used to expose digital files on to the color film. You need large format films and I could not find one doing business for them. May be someone helps. You can dublicate this one film with contact printing new ones also

Mustafa Umut Sarac
07-14-2010, 01:22 PM
QG , I sent them as reference at the hand

07-14-2010, 01:27 PM
But could I expect to get a decent analysis w/ a scan?? I think not, as the color would undoubtedly shift/alter signficantly from the real-world film to a virtual scan. However, indeed it is worth trying. Perhaps there's some standard material with a known spectral-signature that you could scan, and then apply the change in your analysis.

Frankly though, this is beyond my abilities & understanding at present.

John Shriver
07-23-2010, 02:29 PM
Remember that on an autochrome, the same filter that filtered the light hitting the emulsion is the one filtering the light you view the emulsion with. So there's no mismatch between the spectral sensitivity curve and the dye curve.

However, Dmin must be pretty darned high on Autochromes. No real whites, just grays. Probably a substantial part of the look.