I need to know the spectral response curve of a compact digital camera which has a CCD image sensor. However, I don't have the facilities for the same at the moment and I need to address this issue quickly. So I was wondering whether how variable these curves are for the compact cameras which are easily available in the market and cost around 75-100$ ? Will it be fine to use the response curve for some other camera?

  • 4
    \$\begingroup\$ I'm voting to close this question as off-topic because based on comments in the answers, this question is about using digital cameras for scientific measurement in a context that is not likely to be relevant to photography. \$\endgroup\$
    – scottbb
    Commented Jun 24, 2016 at 14:35
  • \$\begingroup\$ If you really want RAW and low cost you may find that the free CHDK software (Canon Hack Development Kit) that woks with many older Canon camera and I think some newer ones may give you RAW. \$\endgroup\$ Commented Jun 25, 2016 at 0:37

2 Answers 2


There isn't any such thing unless you have a camera which outputs RAW images. For all cameras the internal preprocessing (which generates the JPGs) applies responsivity corrections before you ever see the data. I doubt you'll find a camera in that cost range which has RAW output.

Now, as a Systems Engineer and Analyst, I always ask: What is the problem you are trying to solve? THere may be a completely different approach which will work better for you. There's also another important design rule: Good; Fast; Cheap: pick any two. Which is to say, you simply cannot achieve success under your proposed approach (since you basically required Fast and Cheap).

  • \$\begingroup\$ I am using leds of a specific wavelength. I need to calculate the amount of energy of a particular wavelength of light that is being captured by the camera lens. Now if I know the spectral response I can trace back to the energy(I do not need an exact value, just some parameter whose value can hold proportionality to the exact thing) \$\endgroup\$ Commented Jun 24, 2016 at 12:25
  • \$\begingroup\$ there is no practical difference between RAW data having unknown response curve and preprocessed data. of course RAW is still superior because of higher precision, which might or might not be significant for the problem you are trying to solve. \$\endgroup\$
    – szulat
    Commented Jun 24, 2016 at 14:12
  • \$\begingroup\$ @szulat the difference is that the JPG has unknown spectral correction curves applied. For RAW, "all" you need is knowledge of which pixel is behind which Bayer filter. \$\endgroup\$ Commented Jun 24, 2016 at 14:22
  • \$\begingroup\$ @CarlWitthoft if you need exact values then you will have to calibrate your measurements anyway, and then the color correction does not matter because it will be zeroed out by your calibration. \$\endgroup\$
    – szulat
    Commented Jun 24, 2016 at 14:31

Just take your own measurements

You don't really need complete response curves, just single data points for the leds you use.

  • Compare photos of your leds set to multiple power levels.

  • Shoot in manual mode, in case your camera does some tricks and effects to "improve" the result.

  • If using JPEG, apply reverse gamma function to obtain linear intensity values.

  • Defocusing your led image will allow you to apply high brighness (more signal and less noise) and sample (average out) many image pixels without overexposing.


Not the answer you're looking for? Browse other questions tagged or ask your own question.