quick exposure question

Started Feb 14, 2013 | Discussions thread
jojo81 Regular Member • Posts: 120
Re: quick exposure question

BillHinge wrote:

steephill wrote:

Ah, so the lights are the subject. Use your camera's meter and histogram. I expect that these LED's will show a narrow spike on the histogram so all you have to do for each one is make sure you adjust exposure so that the spike appears in the same place on the histogram. I suggest aiming for a central spike for ease of use. That would in effect normalise the differing exposures.

I take it you intend to use the height of the spike to create the response curve? How will you prevent the sources from saturating the sensor?

-- hide signature --


Good question, my thinking is I have done this before but I wasn't sure if my normalisation calculation was correct. What I did was to measure the exposure on program/RAW and put them results in a spreadsheet vs diode wavelength, then saying a 40mW was 8 times brighter than a 5mW, therefore there was an 8 EV diff. If one bulb was very bright eg 656nm or 390nm then I adjust the iso. Having done this for each diode then I would divide each result by the correction factor and plot the chart

Trouble is I'm not sure if 8 times brighter measures 8 EV higher or sqrt(8) times

At the end of the day I'm only looking for an approximate result

By the way, another way of doing this (other than using a spectrometer) is to image certain stars with a diffraction grating on a telescope and then putting the recorded spectrum into a spectrum analysis program. There are certain functions in the software that will produce a camera response curve by comparing known star spectra with the recorded spectra, (this is useful to profile against other stars etc with unknown spectra). Unfortunately the software either costs or is complex to use so I was doing it the easy way as a first step. My reason for wanting to know the frequency response is for astro photography e.g. sensitivity to certain emission lines eg Ha, OIII, SII (which coincidentally correspond to the diodes

sounds exactly like something i had to do many years ago as a part of my astronomy course for my physics degree.  Just i wasn't calibrating my own camera, but the telescope/camera rig i was using at the time.

<physics lecture>

I think there ought to be a simpler way of doing it, but for some general tips, if you can calibrate the light out put of the diodes by using a photo diode/ammeter set up you can know the relative luminance of the diodes, this will give you a pretty good baseline for your calculations.

depending on the set up/quality of the photo diode you may need to normalise the results from the photo diode based on _its_ frequency response chart but that ought to be readily available from the manufacturer.

Alternatively you can just try contacting the manufacturer, and they will probably tell you that any unevenness in frequency response is automatically corrected by the camera, that might just be wishful thinking on my part, but i'd like to think they'd thought of that before i did.

as for distance modifications i believe you're on an inverse square law assuming a point source of light and no significant scattering.

area of a sphere = 4*pi*r-squared  so double the distance is a quarter of the brightness.

</physics lecture>

 jojo81's gear list:jojo81's gear list
Pentax K-30 Sigma 120-400mm F4.5-5.6 DG OS HSM Pentax smc DA 18-55mm F3.5-5.6 AL WR +2 more
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow