I'm studying computer science and I'm looking for possible bachelor thesis topics. I was wondering if there is an application for making 24 hour or multiple day timelapses automatically, e.g. program running on computer like raspberry pi will determine ideal exposure for a connected camera based on scene illumination, take pictures by usb, download them from camera etc. It would be open source.

Do you know about something like this that exists?

As an example, I shot this manually: https://www.youtube.com/watch?v=l0Gb9S644qk.

I would like to have a program running on a small computer (raspberry) doing the same.

I like photography and CS, do you think this would be a good topic for a bachelor thesis or complete waste of effort?

  • \$\begingroup\$ Maybe you want to check this question to get idea about what is required in sense of exposure for timelapse: photo.stackexchange.com/questions/73412/… And CS mean CounterStrike :)? \$\endgroup\$ Commented Feb 10, 2016 at 18:49
  • \$\begingroup\$ I think the main idea in this question isn't really about photography, so this might be off topic here. The difference between a loop that runs for an hour and one that runs for a week is really just the upper limit of the loop counter, right? So that part of the project doesn't seem very interesting from a CS standpoint. Building a system for automatically compressing, storing, and easily retrieving all those images, on the other hand, might be worthwhile. \$\endgroup\$
    – Caleb
    Commented Feb 10, 2016 at 18:50
  • \$\begingroup\$ @Caleb I edited my question with link, what I would expect. It isn't about looping, but about adjusting expositions with various levels of light, so that result would look pleasing - no flickering, no sharp bumps of under or overexposure, but smooth transitions between day and night. \$\endgroup\$
    – Martin
    Commented Feb 10, 2016 at 18:55
  • \$\begingroup\$ It's probably been done for a pi with a picam. You may have trouble communicating between a pi and the camera as the software is often windows /mac only and the communications not openly described. \$\endgroup\$
    – Chris H
    Commented Feb 10, 2016 at 19:09
  • 2
    \$\begingroup\$ I don't think that auto exposure could produce effect with any good quality, It will have serious flickering. \$\endgroup\$
    – Martin
    Commented Feb 10, 2016 at 19:27

2 Answers 2


What you are trying to do is accomplished by a technique known as bulb ramping, or bramping (see: "What is bulb ramping?"). There are several commercial products and open source projects that bulb ramp cameras to create smooth timelapse sequences without jarring discontinuities in exposure.

Realize that basically, bramping requires prediction, or assumption, about what the lighting conditions will be. In the large, this works well: you know that at sunset, you will predictably have certain rough lighting conditions, on average. What you cannot predict as easily is specific conditions, such as cloudiness that would reduce your exposure more than anticipated.

In my opinion, an interesting bachelor thesis topic would be to incorporate short-term and longer-term weather forecasts to adaptively steer the bramping slope levels. But assuming away weather-induced or other unexpected exposure variations, outdoors timelapse bramping is predictable, and essentially, a solved problem (i.e., probably not an interesting thesis topic).


Here's an alternative approach, which may or may not work depending on just how much the light levels in your scene vary, and how much movement there might be.

Fix your exposure to the minimum exposure you're going to use. Using a simple trigger take a number of shots at each interval in proportion to the light you need. E.g. at noon take one shot at 1/125, after dark take 125 of them and add together for a 1s equivalent exposure.

This could be more or less noisy depending on your noise sources, and I suggest you shoot in raw mode if at all possible to maximise your dynamic range.

You could calibrate the number of frames manually or with a light meter on your control computer (or by image analysis on the fly if you can access the images).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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