Science Buddies: "Ask an Expert"

Okay, so I'm going to begin this experiment: https://www.sciencebuddies.org/science- ... p022.shtml

I need help because I was really busy one night and I decided to begin taking pictures even though I hadn't done the white paper part of the procedure. I thought that I didn't really need that part. So I took two pictures every two hours (starting at 9:45pm) until 5:45am, and the setting were 15 seconds at f/2.8. I set the focal length at infinity, and the sensitivity at ISO 200. The camera was set at black-and-white mode, and pointed straight up on a tripod. Then after I had all of the pictures, I followed the procedure to crop the photo and create a histogram of pixel values. Then I kept track of all of the pixel intensity information. But when I went to keep going, it said to create a graph, and I didn't take the pictures at different shutter speeds, only at 15 seconds. Does this mean that I have to do the experiment over? If so, how should I revise my process? And the example graph shown in the experimental process is completely incomprehensible to me. Could you explain it? Please note that I had to save it with eight million pixels. Does that affect it? I could save it slightly smaller if I had to do it again, at two million pixels.

Also, I do not understand how to get the amount of skyglow from the graph once I complete it. Please reply soon, it's urgent!

Thanks,
NinjaPerson

Okay, so I'm going to begin this experiment: https://www.sciencebuddies.org/science- ... p022.shtml

I need help because I was really busy one night and I decided to begin taking pictures even though I hadn't done the white paper part of the procedure. I thought that I didn't really need that part. So I took two pictures every two hours (starting at 9:45pm) until 5:45am, and the setting were 15 seconds at f/2.8. I set the focal length at infinity, and the sensitivity at ISO 200. The camera was set at black-and-white mode, and pointed straight up on a tripod. Then after I had all of the pictures, I followed the procedure to crop the photo and create a histogram of pixel values. Then I kept track of all of the pixel intensity information. But when I went to keep going, it said to create a graph, and I didn't take the pictures at different shutter speeds, only at 15 seconds. Does this mean that I have to do the experiment over? If so, how should I revise my process? And the example graph shown in the experimental process is completely incomprehensible to me. Could you explain it? Please note that I had to save it with eight million pixels. Does that affect it? I could save it slightly smaller if I had to do it again, at two million pixels.

Also, I do not understand how to get the amount of skyglow from the graph once I complete it. Please reply soon, it's urgent!

Thanks,
NinjaPerson

Hello ninjaperson221,

Please read the document, https://www.sciencebuddies.org/mentorin ... =26&t=8326 and links in it for help. Get back to us if you need more help.

WW

The discussions that WW pointed out should mske things a lot clearer I hope. But just to be extra clear, you asked:

> Does this mean that I have to do the experiment over?

Yes.

>If so, how should I revise my process?

Read the project description and the materials that WW pointed out. Follow the project directions this time.

>And the example graph shown in the experimental process is completely incomprehensible to me. Could you explain it?

This is covered in the references. The response (0-255) of each pixel is related to the amount of light that fell on it in a way that is rather complicated when considered in detail; in particular it is not just proportional to the number of photons hitting the pixel. The response versus shutter speed is, by contrast, reasonably simple: for your purpose you can assume that the amount of light picked up by each pixel is linearly proportional to the length of the exposure (within the spectral response curve of the camera optics and detectors). The graph then allows you to convert something complicated (the relation of the pixel readout value to the amount of light) to something that is very simple, the equivalent shutter speed. You can then directly compare shutter speed needed to reach a given pixel value for various light levels using your calibration.

>Please note that I had to save it with eight million pixels. Does that affect it? I could save it slightly smaller if I had to do it again, at two million pixels.

Fewer pixels should be OK

Hope this helps.

Hi,

I did the 'calibration' pictures with the white paper, and now I'm looking at the pixel gray level statistics using the application ImageJ. I'm on my 10th photo out of 14, and so far, all of the information has been the same. The mean is 255, the standard deviation is 0, the minimum is 255, the maximum is 255, and the mode is 255. Did I do it wrong again? I changed my shutter speed by a factor of two with each photo, and I did them all one after another one the same day and everything... I'm kind of worried, I have a lot of work to do on this project and I need to get it done right soon. Please reply A.S.A.P., I really need the advice!

Thanks,
NinjaPerson

I have something to add:

With a 15 second shutter speed at f/2.8, I have a mode of 65, which is, obviously, less than 128. So what I was wondering was how do I find the skyglow with the average intensity only being 65? I'm not going to get to 128. Although it is a little cloudy out...

Again, please get back to me soon!

Thanks,
NinjaPerson

NinjaPerson,

Your camera represents the brightness in an image as a number from 0 to 255. 0 is black. 255 is white. In your series of exposures that all yield a mode of 255, the 255 tells you that all of the images are white or "overexposed." You need to reduce the exposure further until you see numbers below 255. Ideally you will continue to reduce the exposure conditions until you get 0 or close to zero. With this data in hand, you can make a graph of image brightness versus exposure.
The total brightness you get during an exposure is proportional to several camera settings.

1. The Exposure Time.
   Light enters the camera for the exposure time. An exposure of 2 sec. lets in twice as much light as 1 sec. An exposure of 1 sec. lets in twice as much light as 1/2 sec. An exposure of 1 sec. lets in 1000 times as much light as 1/1000 sec.
   The Aperture
   The aperture also affects the exposure. Aperture refers to the diameter of the opening in the lens. The numbers used to indicate the aperture are in fractions of the focal length of the lens. You should be using the same lens and the same focus setting for all exposures so the focal length doesn't change. An aperture of f/2.8 lets in twice as much light as f/4. Why? The amount of light you let in depends on the area of the opening, and the area is proportional to the diameter squared. (f/2.8) squared is approximately (f^2/8) and (f/4) squared is (f^2/16). (Note: f^2 means f squared. This editor doesn't support exponents.)
   The Film Speed
   Of course, a digital camera doesn't have film, but it usually has a control that serves the same purpose--it changes the light sensitivity. A setting of 200 ASA is twice as sensitive as 100 ASA.

Once you learn and understand these factors that control the sensitivity of your camera, you can get an average exposure (the mode in ImageJ) of 128 for any light conditions. For the white paper, just continue to shorten the exposure time and/or close down the aperture and/or reduce the film speed until the mode drops to 128 or below. Of course if one exposure gives you over 128 and another exposure with the sensitivity dropped by half gives you under 128, the settings to get 128 would be in between the ones you used.
When photographing the sky, just continue to increase the sensitivity until some exposure exceeds 128.
If you get a mode of 128 for the white paper, and you get a mode of 128 for the sky when the camera is set to 1000 times the sensitivity, then the sky is 1000 times as dark as the paper.

All the information given here is also in the materials you already have, but perhaps this version helps.

WW

Hello. Im kinda doing this experiment as my research topic here in the Philippines.
I was just wondering how to do the EEQ Graph. Since the only variable that changes in your skyglow images are the "Mean Average Pixel" (Because the exposure time is constant 30 secs). How will I get the value of EEQ? (ex. The 0.003 based on the explanation)
Thank you for answering. This would be really helpful for my research.

Hi xfbangeles,

For this project you “calibrate” your camera by making a series of pictures at different shutter speeds. Next you use image analysis software to find the average pixel data for each of your pictures. At the end of this step you should have a data table showing the shutter times and average pixel values.

The next step is to make a graph of the data from your table. You'll put the shutter times on the x-axis and the average pixel values on the y-axis. You should use a logarithmic (log) scale for the x-axis because of the very wide range of shutter times.

If you are using a spreadsheet program or other graphing software you can choose the log scale option for the x-axis. If you are making your graph by hand you'll need to buy “semi-log” graph paper which has a log scale on the x-axis.

There is an example calibration graph near the end of the procedure section of the project. Your completed graph should look something like this example.

You use the calibration graph to get the EET (equivalent exposure time) values for your skyglow pictures. First you find the average pixel value for a skyglow picture with the image analysis software.
Now you locate the average pixel value of your skyglow picture on the y-axis of your calibration graph.
Next you find the shutter time on the x-axis of the calibration graph that corresponds with your picture's average pixel value.

I hope this helps and good luck with your project. Please post again if you have more questions.

A. Norman