Science Buddies: "Ask an Expert"

Need help. Working on "using a digital camera to measure skyglow". Completed all aspects of experiement but cannot find/understand correlation with control data and final results/picture assessment. Can anybody help?

Completed all aspects of experiement but cannot find/understand correlation with control data and final results/picture assessment. Can anybody help?

Sounds like you got a bit ahead of yourself. You really shouldn't "do" an experiment much less complete it until you have figured out how to analyze the results.

Without knowing exactly what you are having trouble with understanding, it is hard to give you specific help. Post back with a detailed question about your data and we can be more specific in our help.

The goal of the calibration steps was to determine the "noise floor" of your digital camera. The photo-electronic sensors used in cameras only have a limited range of light sensitivity. If there is less light than the "noise floor", then the image you get is a mixture of mostly the noise floor and some from the real image. If there is more light than the "noise floor", then the image you get is a mixture of mostly the real image plus some of the noise.

With most digital cameras the use CCD type sensors, the noise floor is typically a function of exposure time. The longer the exposure, the more noise that will be integrated into the pixel gray scale count. The average pixel density you obtained for a region of sky can be compared to the baseline you obtained to determine how bright or dark that section of sky was. It probably wasn't clear from the proceedure writeup that you should pick regions of the sky without the moon, bright stars, and any house or street lights so that their light contribution doesn't affect your dark sky measurements.

Another aspect that isn't covered in the short writeup is how shutter speed, ISO, and aperture correspond to absolute light levels. On a bright day near noon, an 18% reflectance neutral grey surface such as an asphalt parking lot (or a calibrated gray scale piece of paper) will be rendered near 18% when the aperture is f/16 and the shutter speed is 1/ISO (in the case of ISO 200, 1/200 sec). Since the clear blue sky is about 18% reflective in the blues, what this little piece of information gives you is a way to reference how dark the night sky is to the clear daytime sky. You can try out your digital camera by taking a picture of an asphalt parking lot or street in the middle of the day. If you are at ISO 200 1/200 sec f/16, I would expect your pixels to be approximately 46,46,46 (RGB) or 46 (gray scale) on a digital camera with 8 bit channels.

Hope this helps in general, but be sure to post back with any specific questions you have about the original proceedure or your data or what it means.

Thanks for the information. It was very insightful. From my first reading of the experiment, I felt it would be pretty easy to make a correlation/connection to the gray level/shutter speed graph. My first question ...is about the graph-gray level vs shutter speed; the graph plots exactly what (i.e. ft/sec)? The process calls for a number of pictures of the night sky. Each one of those pictures with the help of ImageJ give us an average pixel level (mean or mode). We compare those results to the graph? I am assuming that is what we are trying to determine. Am I comparing the correct data?

My first question ...is about the graph-gray level vs shutter speed; the graph plots exactly what (i.e. ft/sec)?

Exposure (film or sensor) is a function of relative aperture, shutter speed, and sensitivity. The relative aperture is given relative to the focal length of the lens as in f/2.8. This correlates to how much of the light possible is transmitted to the film or sensor. It is extremely expensive to make lenses large enough in diameter to get near to f/1.0 and if you could, they wouldn't be very good at rendering images with enough depth of field to be useable for general photography. The shutter speed is simply how long the light is allowed to hit the film/sensor. The sensitivity of the film with a given development process or the sensor on a gray scale is how much light it takes to expose the film so that when it is developed it is 50% or half way between transparent and opaque or in the case of an 8 bit digital gray scale, a 128 reading. Converting and calibrating this to absolute scientific units is a difficult process as there are many non-linear behaviors involved.

ft/sec would be a velocity and I don't think of anything in this process that involves a velocity. The typical shutter is two fast moving gates that are timed with some separation in time so that the total open time matches the desired shutter open time (called shutter speed but it really isn't a velocity). If the separation of the shutters is 1/8 of an inch and the shutters travel at one inch per second, each part of the film will be exposed for 1/8 of a second. If the separation of the shutters is 1/32 of an inch and the shutters still travel at one inch per second, each part of the film will be exposed for 1/32 of a second. Note: The velocity of the shutter does not have to change to get different shutter open times. Typical shutter velocity is faster and can be predicted by knowing the "flash synch speed". When the shutter is set to the flash speed synch, the separation in the shutters will have the entire film/sensor frame exposed at one instant. Faster "shutter speeds" are achieved by not having the entire frame exposed at the same time like the slit I talked about earlier. Slower "shutter speeds" are achieved by having the entire frame exposed for longer than an instant so that the two gates don't start traveling at the same time.

The best way to "calibrate" your setup is to photograph known objects. For example, with a full moon. At f/11 and the shutter speed = 1/ISO, you should be able to take night pictures that are perfectly exposed. As with the f/8 and shutter speed = 1/ISO giving you perfectly exposed clear day images. In other words, the difference in light from the sun and the moon one f-stop or 1/4 in terms of light intensity.

Photo.net learning area on aperture, shutter speed, and exposure is a good place to read about this http://photo.net/learn/making-photographs/