Movie mode won't work because the frame rate is way too low for any shutter speed above 1/20 or so. The photoplug is probably the best thing, if you still have a headphone jack, but I also want something zero-budget.
I might have an idea that might not give you exact measurements but could still be useful for checking correctness and approximately measure the actual shutter speeds with 1/3 of a light stop or so accuracy (which is all I'd need in practice). I haven't tried it yet so I don't know if it will actually work but I'd love to hear from you.
I'm not aware of an app for this but I was just thinking you could manually already do the following, all you need is a smartphone (or any digital camera) with reliable manual camera controls and a light and maybe a few optional things, like a few strips of something you can safely stick to cover parts of the film gate (this is for a bit more advanced diagnostics). A black bag that doesn't let through light is probably advisable, as well as something to keep the smartphone and the SLR firmly in place during the tests.
Basically the idea is to compare the exposure of the shutter speed of the SLR with the exposure of the same shutter speed on the phone.
Setup: you place your phone behind the film gate, like you originally suggested, and open the camera app and go to fully manual mode. The only light source should be in front of the SLR's lens, and the only light going into the phone camera should come through the SLR's film gate (this is where some sort of a dark bag could help; alternatively, it would probably also work if you use a (cheap LED) flashlight pointed into the SLR lens in a dark room).
How to test: the test consists of three parts: (1) creating a reference photo using the smartphone's shutter and (2) creating a test photo using the SLR's shutter and (3) comparing. These photos don't need to be focused because we will be comparing average brightness anyway, so it's even best to be as blurry as possible. The (1)
reference photo can be created by fully opening the SLR shutter while taking a picture with the phone's shutter speed set to the speed you want to test. Then, (2) the
test photo can be created by (i) setting the SLR's shutter speed to the same shutter speed you used in step (1), (ii) taking a picture with the phone's shutter speed set to multiple seconds so you have time to (iii) press the SLR's shutter release button. If the average brightness of the reference and test images is the same, and there was no other light coming into the phone's camera that didn't pass through the SLR's shutter, and there was no difference in the physical setup between the two shots, I think we can conclude that the shutter speeds are the same. Now you can run this test across all shutter speeds, even the fastest ones.
Calibration: to find out what the actual shutter speed is closer to (up to 1/3 stop accuracy), you can play around with different shutter speeds on the phone for the reference photo and different ISO's on the phone for the test photo. When you match the average brightness between reference and test by adjusting shutter speed or ISO on the phone, this will tell you how fast the SLR shutter actually is. For example with ISO, if the reference had ISO 100 and shutter speed 1/100 and the ISO needed to match it in test was 200, this means the shutter speed of the SLR was actually 1/200. You should get the same result by adjusting the shutter speed to 1/200 when taking the reference picture and keeping ISO fixed. It would probably be nice to do both to verify.
A bit more advanced testing: I found this
article, which points out some problems with the shutter that I hadn't considered before. For example, maybe the shutter curtains are closer together on the left than on the right, which could give the same average brightness but would result in a photo that is underexposed on the left and overexposed on the right (there is a photo in the article illustrating this, at the end of "Level 3". The solution in the article is "Level 4" testing with multiple diodes (btw, you can't do this testing with a photoplug on the phone). But if you have some stripes of something completely opaque and sticky but that doesn't leave traces, you could also try to detect these kinds of problems using our setup. You can simply cover the middle and left part of the film gate (careful not to obstruct the shutter curtain movements!) to test the right side of the shutter curtains, etc. If the shutter curtains are not parallel, the test photos for the right part uncovered will have a different average brightness than the test photo for the left part uncovered.
Testing light metering and auto-exposure: I haven't thought about this, but probably you can use a similar testing methodology. I don't think it would work reliably as-is to test automatic aperture because different apertures would affect the geometry of the light beam, and which part of the beam goes into the phone camera. You would need to somehow optically average the light that would fall on the film (using milky plexiglass maybe?). But to test automatic shutter speed and metering involved there, you could try getting the reference photos using the phone's automatic shutter speed, and compare that to the SLR's. While doing this, the ISO settings on the phone and SLR should be the same.
What do you think?
You could probably simplify this process a bit by developing an app, in particular, for comparing average brightness more scientifically than just eye-balling it, but this procedure would still require a lot of manual work (setting shutter speeds, pressing buttons) so I think an app isn't crucial.
I see it's an old thread but I just ordered my first film SLR on eBay and was wondering how to test it for light leaks, shutter times and light meter and auto-exposure modes on zero budget and without ruining film, which is now getting more and more expensive to buy, develop and digitize. So if you have any tips on zero-budget film SLR testing, please let me know!