Public Lab Research note


This is an attempt to replicate an activity.

Another one about over-exposure...

by viechdokter | April 09, 2016 16:15 09 Apr 16:15 | #12950 | #12950

Warren told me it would be interesting for some of you to know what happens when we try to avoid over-exposure. Well, here are my first results.

Direct sunlight is too strong. One way to reduce intensity is to use only a reflected part of it. I used plain white paper for the next two spectra. In the first I aimed at a shadowed part of the paper:

sunlight_13.jpg

Reds and greens are low (so that yellow and turquois are missing). No second (middle) red peak anymore. Almost no light at 600 nm.

Now I show you a spectrum of sunlightthat was reflected off the brighter part of the paper (without shadow).

sunlight_14.jpg

Some more light at 600 nm but still no yellows at all, no turquois. Some flickering in the "waterfall" picture and some darker small lines in the green and blue part. Have those lines any meaning? Between 420 and 480 nm the blue curve looks "too flat" somehow. I am not sure if that is just co-incidence.

Here you can see the combined (overlayed) intensity curves:

sunlight_13_and_14_combined_intensity_curves.jpg

Blue part of the spectrums is almost the same, red is only a bit reduced. The main difference is in the green range between 480 and 590 nm. Again, blue seems to be the colour of the shadows.

The worst over-exposure happens in the green range. It seems that using light reflected off white paper helps against over-exposure but it affects the whole spectrum by swallowing parts more or less completely (and more than other parts! The whole curve changes.). Perhaps there are other ways to reduce light intensity without changing the spectrum too much. Perhaps we should try polarisation filters???


3 Comments

Hi viechdokter,

These are great observations. They reveal important features of spectra created using consumer cameras as the detector. It's really interesting that your camera decides to output gray when all three channels are saturated. I never noticed cameras doing that. It would be good to know what camera you are using and what automatic or manual camera settings you are using (white balance, exposure). Other details about the spectrometer (grating, slit width, dimensions) would also be helpful.

Thanks, Chris

Reply to this comment...


Hi Chris,

you are right, the settings are important too. I absolutely forgot to mention them. Here is a (translated) snapshot of the settings:

camera_settings.jpg

I used the PLab spectrometry kit 3.0 and worked with the same settings for all of today's spectra. I don't know the brand of the USB webcam that belongs to the kit but for the settings it uses a Dell webcam Central program that was delivered with my notebook. The slit width is 0.4 mm, the grating is a normal DVD.

I was astonished, too, about the gray overexposure output but perhaps I might get different outputs with different settings. I will publish any further results as I stumble upon them.

Reply to this comment...


Hi, thanks for posting! One thing I've wondered about is whether different exposures -- that are not clipping/overexposed but are different overall heights -- would height-adjust to match again. That is, if you take 4 spectra, at (for example) 20%, 40%, 60%, and 80% max height, and you add them to a set and height adjust them to equalize their max heights, would they have the same curve shape? I'd love to see a test like that.

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Login to comment.