Is measurement of relative photon fluxes possible via RAW?

wim001

« on: 29 / January / 2009, 16:18:33 »

Advertisements

I'm trying to measure how homogeneous a certain light source illuminates a flat area. I'm interested in relative differences in incident photon flux density in that flat area. I have almost no experience with RAW and only 'googled' knowledge, but got the idea (maybe it's an old one) that it should be possible to measure these relative differences in photon flux density via the raw data from a CCD. I installed CHDK on my P&S Canon (590IS) and took some raw pictures to test.
I would like to avoid to use a RAW converter that manipulates the RAW data in any sense to be sure that I get the pure signal from all pixels which (al least per individual color channnel). Essential is that the output per pixel is still linear with the photon flux.

How can I access this data in the raw file? I have no experience in programming and would like to use a simple converters that does not do any data manipulation (e.g. gamma correction). Reduction from 10 to 8 bit is probably not a problem and it would be nice if I could get the output in a simple matrix with values or output readable by ImageJ ...

I need some help: Do I make a mistakes in my reasoning of tackling the problem and what software can I use to convert the CHDK CRW files from my camera to this type of 'pure' output?

Wim

Logged

whoever

280
IXUS950

Re: Is measurement of relative photon fluxes possible via RAW?

« Reply #1 on: 30 / January / 2009, 03:12:14 »

One way is to convert RAW to 16 bit linear RGB with dcraw (http://cybercom.net/~dcoffin/dcraw/), so find a pre-compiled binary (or compile yourself, if possible) and give it a try:

dcraw -v -4 -o 0 -q 0 <your_raw_file>

Explanation of options:
   -v   verbose messages
   -4   linear 16 bit RGB output
   -o 0   raw output colorspace
   -q 0   minimal interpolation (for faster processing)
You could also use -T to write TIFF instead of PPM.

TIFF images should be readable by pretty much any decent software.

Logged

barret

125
S3 & a550

Re: Is measurement of relative photon fluxes possible via RAW?

« Reply #2 on: 30 / January / 2009, 09:45:45 »

and what about vignetting? would that affect results from pixels at corners of the image?
btw: may i ask, what are you doing this for?

Logged

my timelapses on Vimeo my startrails gallery

wim001

Re: Is measurement of relative photon fluxes possible via RAW?

« Reply #3 on: 30 / January / 2009, 10:24:58 »

Thanks for all the suggestions.
In our research, we measure leaf photosynthesis, which is a photon driven process. We would like to know how homogeneous the incident photon flux is distributed over the leaf area we illuminate.

Can you explain me what you mean with vignetting

Logged

barret

125
S3 & a550

Re: Is measurement of relative photon fluxes possible via RAW?

« Reply #4 on: 30 / January / 2009, 10:31:54 »

long, rude answer

http://en.wikipedia.org/wiki/Vignetting

short, polite answer:
vignetting causes decrease of brightness in corners of photograph. center is well lit, but corners are darker.
amount of vignetting depends on aperture of the lens.
you may have not noticed this if you used high apertures such as f/8 or f/16. on some lenses it is very visible for wide apertures such as f/2.7.

« Last Edit: 30 / January / 2009, 10:33:57 by barret »

Logged

my timelapses on Vimeo my startrails gallery

wim001

Re: Is measurement of relative photon fluxes possible via RAW?

« Reply #5 on: 01 / February / 2009, 16:50:01 »

Thanks, you helped me a lot with dcraw + commandline options, and also with the short polite answer about vignetting (:. We did not use the borders/corners in our image for analysis purposes to avoid these problems. It seems that we managed to measure what we intended to measure.

Logged