Is Your Camera Lying?

It is easy to rely on the ratings marked on different tools, whether it is a power supply, scale, or speedometer. However calibration is essential for any part that is relied upon either professionally or for a hobby. [Jeremy] wanted to see if his Lomography camera shutter really was only open for 1/100ths of a second when set to that. In order to test his rig, he set up an LED on one side of the shutter, and a high speed phototransistor to gauge the time spent open, using an oscilloscope to measure the time the reference point was pulled low. In his case, when the camera was set to 1/100, the shutter was actually open for closer to 1/150th of a second (the mean was 1/148ths of a second, with a standard deviation of 417 uSecs). This difference can make a large difference in picture brightness.

Be sure to check his blog for more pictures of the setup, as well as some useful part references and circuit diagrams.

25 thoughts on “Is Your Camera Lying?

  1. Interesting read, and I like the idea of designing an experiment to test a possible problem. At the same time I wish he had a solution to actually calibrate the shutter speed. As it is he just measures it, right?

  2. Ah, but did he also test his oscilloscope to make sure it was giving accurate readings? I kid…

    Even though the idea is pretty basic, this is a cool and fairly clever hack (and yes, I do consider it a hack). Good job!

  3. Well, it is worth asking what kind of shutter he has in the camera. It’s a cool experiment, but different shutters act differently. One of the most popular high speed mechanisms is called:

    http://en.wikipedia.org/wiki/Focal-plane_shutter

    It is just a little slit that slides across the frame. Given that he didn’t have a perfect point light source, it could skew his results.

    If you look at the Rigol capture, you would expect that the drop would be perfectly vertical (in a perfect world, with a perfect light source, etc.) It isn’t which implies that it is actually letting in more light.

    I’m not criticizing the hack at all – this is a cool experiment. The most interesting part is that he got a range of results in his 75 image sample.

    I wonder how a a higher end camera would react.

  4. Film has a LOT of forgiveness when it comes to exposure. The cheapie junky camera he uses is not designed for precise photography. It’s not lying, it’s just horribly inaccurate.

    P.S. his claim you cant get that kind of photo effect with a Digital camera… He never tried. a lensbaby does just that for dirt.

  5. @biozz, Now that you mentioned it, I went back and found a similar story of someone completely different calibrating shutter speed, However this story provides a little more detail and some excellent pictures of the setup.

    Great spot though, even some of the writers miss each others stories some time.

  6. The comments reveal that people are taking the wrong message away from this article. The test is not uncovering a flaw in the timing of real film or digital cameras that a normal person might buy in a store. It’s showing that $2 toy cameras being used for taking technically horrible photos under the guise of art have the sorts of problems you would expect them to. People buying proper cameras don’t have to worry about this sort of thing. There’s no scandal here.

  7. If your light meter lies it is either a: not calibrated correctly, or b: not being used properly. The zone system is a good thing to get acquainted with. Being empirical is a fun way to approach photography but it rarely produces consistent results.

  8. You don’t need a high speed phototransistor. I calibrated a old mechanical SLR using a LED as the detector and a house hold lamp, along with my oscilloscope. Tweaked the cams and set screws and got it working pretty well.

    The rise and fall time for a typical LED are in the low microsecond range, many are sub microsecond. You can get a ns speed light detector with a run of the mill 850 nm laser diode…

  9. @Brian,

    It’s not a focal plane shutter, it’s a very simple leaf shutter. I’m sure a focal plane shutter would behave differently. You’d probably have to design the experiment differently, too; perhaps recording three sensors, one each at the far edges of the film plane and one in the center.

    But this was cool for what he was testing. He learned that his shutter is half a stop darker than expected, not that it really matters with a Lomo.

  10. @loans

    The point is not to take terrible pictures, the point is to take lots of pictures.

    The fact that they’re being shot through a Lomo makes most of them terrible. But occasionally the distortions line up in a particularly pleasant (though not accurate) way with the subject material, and you get an artistic shot worth keeping.

    For the cost of all the film processing required to get that one decent shot out of the thousands printed, though, it’s probably more economical to buy a decent DSLR and learn how to take pictures that don’t require a freak accident to appreciate.

  11. @jaded

    It doesn’t matter if it is leaf or focal plane. The method is the same. I am just saying anyone with a 7 cent LED can use it as a detector if they want to do something like this.

  12. I actually thought about doing this with my Diana F+

    I’m not really a fan of the Lomo philosophy and I generally dislike the idea of just blowing through film to end up with a happy accident.

    However, if you know what you’re doing with the camera and film you can actually turn out repeatable and artful results.

    Can it be done with photoshop? Depends, but once you know what you’re doing it’s much easier to just hit the shutter than to go through menu’s and settings and layer after layer in PS.

    That and I still firmly believe that film produces a finer tonality than digital.

    I have cheap throw away cameras, decent DSLR’s a scanner camera and some pinholes. They’re just tools to be used.

  13. 50% out is not bad at all. I service cameras from time to time and with some mechanical timing mechanisms, all the speeds affect each other – so to be within 50% might be the best you can do for some of the speeds.

  14. Thats really intertesting thought I am more keen on researching about the cameras but this is really something which i think is poretty impressive as I would have neverthougt of this things being of any importance especially when the time difference is like really microscopic. Jesus, i feel like reading F1 timings :). Anyways all i know about Cannon or Nikon is through som sites i visited like Testfreaks etc. Pretty neat info though.

  15. Is your test equipment lying?

    Such a simple setup with a photo transistor and a 10k pullup resistor (and long leads with unknown capacitance) could easily mean that the camera has an exact 1/100s shutter speed with an extra 50% error from your measurement setup.

    Sensor speed could be calibrated with a simple dc motor and a cardboard disk with some slots in it.

  16. I’m sorry, I must be sleeping.

    1/150s is shorter than 1/100s and as far as I know the turn-off delay of photogransistors is usually longer than the turn-on delay.

    But still. Without any information about sensor calibration or accurarcy I would not consider this fun experiment to have any “scientific” value as he claims it to have.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.