July 23, 2008

Using a DSLR as a Shutter-Tester

What good is a DSLR if you can't test the shutter of your Holga with it? Ok, I'm kidding, but there is a fairly simple way to use a DSLR as a shutter speed tester for any lens that has a leaf shutter. In fact, with a little care, this method can be used to test practically any shutter. Don't expect pinpoint accuracy or rush to put your lab equipment on eBay, but this method should be accurate to within at least a third of a stop - certainly good enough for average everyday use.

Think for a moment what a shutter does. It is simply a way to block the path of light falling on the film/sensor and then to remove that blockage for a certain known amount of time to let light fall on the film/sensor. It's a pretty simple concept really - a mechanized and repeatable version of the old hat-on-lens technique. Problem is, shutter speeds go off - they slow down, they speed up and do all sorts of funny things. Often, as in the case of mechanical marvels like the Holga, they are simply unknown, or vary from camera to camera. But if there is one good thing to be said of modern electronically controlled shutters in DSLRs and or other electronic thingamagigs, it is that they are remarkably accurate and consistent. So let's get about measuring a mechanical shutter by directly comparing it to the known shutter speeds of a DSLR.

Our first task would be to mount whatever lens/camera/shutter we want to test in front of a DSLR. Now, we don't need to focus anything or permanently mount anything. We just need to make sure that we can hold our shutter in front of the DSLR in such a way that when the shutter is closed (and the DSLR's shutter is open), no light reaches the sensor. We also, however, need to make sure that nothing sticks into the body of the DSLR so that we don't end up damaging the moving mirror. This can be done in a few ways, but the easiest for me was to use a hollow extension tube (or set of tubes) used for macro photography. I used a Nikon PK-13, mounted a BR-2a reversing ring on it and then used a simple hollow black tube that is available for less than $5 at camerafilters.com. You can use whatever you have handy - the basic idea is to provide a reasonable bit of distance between the shutter and the DSLR's body. Once done, the setup should resemble the picture on the right (I used a film body in the pictures, because I was using the digital body to make the pictures).

Now, point the whole setup towards something reasonably bright, uniform in color and preferably near middle gray in tone - a wall, the sidewalk, distant trees and the sky all work well. Nothing needs to be in focus - in fact, we want things as blurry as possible. Now the optical path has two shutters obstructing the light - the shutter to be tested, and the focal plane shutter of the DSLR. Light will reach the sensor only when both shutters are open. If we hold open either shutter in bulb mode, the other shutter's speed will be the sole determinant of the amount of light reaching the sensor. Quickly test that the optical path is truly light-tight by keeping the shutter to be tested closed and holding the DSLR's shutter open for a few seconds. If no stray light reaches the sensor, you will get a spike at the far left of the histogram as in the first histogram of the screen-capture to the right.

So now, let's put the DSLR on bulb (having a locking remote release is helpful) and fire the shutter to be tested. If you look at the resulting histogram you will see a sharp localized spike at one point. Adjust the ISO (if the lens you are testing has an aperture iris, you can also adjust that - just make sure nothing changes throughout the test) so that the spike is somewhere along the middle like the second histogram in the illustration and we are all set. Note the position of this histogram and now put the shutter on bulb and make a series of exposures with the DSLR shutter around the speed you are testing. So if you are testing to see whether the speed that says 1/100 on your shutter is accurate, then do the following:

  • Make the first exposure with the DSLR on bulb and with the shutter you are testing fired at the 1/100 setting.
  • Now, put the shutter being tested on bulb and make exposures with the DSLR at 1/200, 1/160, 1/125, 1/100, 1/80, 1/60, 1/50.
Each exposure is a third of a stop more than the previous one. If you observe the histogram, you will notice that the spike gradually travels to the right as you increase exposure. Try to pick out the one that has the spike closest to the one in the first frame - that is your shutter speed. Simple.

A word about RGB histograms. You don't need to bother with them for the test. If your subject is middle gray and your white balance is set correctly, your RGB spikes will be more-or-less at the same point, otherwise just use any of the channels or the combined histogram. In this test we are just comparing exposure times by comparing the amounts of light let in by either shutter, so as long as every other factor is consistent between those two exposures, we should do fine.

So that is it. Once you try it out, it's a pretty simple test and you should be all set to make precise exposures on Velvia with your Holga. Have fun!