Shutter speed test for digital camera (please move the thread if not correct section)

[Chan Tran]

There are a lot of ways to check shutter speed of film cameras but I wonder how do you check the shutter speed of a digital camera. Even the professional tool of the film era like The Kyoritsu EF-8000 would not work for digital.

 

[4season]

The only recent camera service manual that I have in my collection is for Nikon's ZFC, which has a vertically-running first curtain shutter. There are no technician-level adjustments for the shutter or image sensor.

 

[reddesert]

I guess one question is, to what end? To see if the shutter is running mechanically slow or hanging up? The electronic timing of the shutter is likely to be as reliable or better than one's test instrument.

If there is some reason to test it or just curiosity, and the main problem you face is lack of access to the back side of the shutter, and the camera has a mechanical shutter (many digital cameras no longer have a shutter) ... it may be possible to test the shutter by measuring the reflectance change off the sensor vs the shutter blades. Sensors tend to have a moderately reflective cover glass while shutter blades are dark. If it's a DSLR and you can't lock the mirror up, you could probably black out the focusing screen so that image is dark as well, so the mirror and shutter are dark-ish and your detector measures an increase in reflectance when the camera sensor is exposed. Clearly this would have to be homebrewed with perhaps: a light source, some photodiodes (with maybe a snoot or optics so they each "see" part of the sensor), and a storage oscilloscope.

 

[Chan Tran]

I guess one question is, to what end? To see if the shutter is running mechanically slow or hanging up? The electronic timing of the shutter is likely to be as reliable or better than one's test instrument.

If there is some reason to test it or just curiosity, and the main problem you face is lack of access to the back side of the shutter, and the camera has a mechanical shutter (many digital cameras no longer have a shutter) ... it may be possible to test the shutter by measuring the reflectance change off the sensor vs the shutter blades. Sensors tend to have a moderately reflective cover glass while shutter blades are dark. If it's a DSLR and you can't lock the mirror up, you could probably black out the focusing screen so that image is dark as well, so the mirror and shutter are dark-ish and your detector measures an increase in reflectance when the camera sensor is exposed. Clearly this would have to be homebrewed with perhaps: a light source, some photodiodes (with maybe a snoot or optics so they each "see" part of the sensor), and a storage oscilloscope.

Mechanical shutter even electronically controlled isn't very accurate because it depends a lot on the curtain speed so there is a need for checking. I wonder at final assembly don't they check it.

 

[monopix]

Mechanical shutter even electronically controlled isn't very accurate because it depends a lot on the curtain speed so there is a need for checking. I wonder at final assembly don't they check it.

So what you are saying is you don't trust the camera manufacturers to produce a camera with a shutter accurate enough for your needs. Maybe you should be looking at scientific cameras rather than ones made for the consumer market.

 

[Chan Tran]

So what you are saying is you don't trust the camera manufacturers to produce a camera with a shutter accurate enough for your needs. Maybe you should be looking at scientific cameras rather than ones made for the consumer market.

One just want to know the facts. You don't want to know or care that's up to you. If I compare 2 cameras even of same make and model they would give different exposure result (exposure may not be correct term here but rather the resulting brightness of the images) even with exact same scene, same lens at same aperture and same shutter speed. I want to know the difference is due to either the different shutter speed or the ISO gain they set.

 

[ic-racer]

check out running LED array:

https://pdfs.semanticscholar.org/c6a5/45876623c92353d27a52b94c49aa908634e8.pdf

 

[wiltw]

If you aim the camera at a blank wall, and manually set the indicated combination of shutter + aperture + ISO, and the histogram peaks at the midpoint of the histogram, that is an indication that the camera is achieving the 'proper exposure' for that combination.
So if you do that for every equivalent combination, using every shutter speed possible on the camera, and they all result in the same position of the peak of the histogram, then you see that every shutter speed is functioning properly, don't you?!

 

[4season]

You might be able to observe what the mechanical shutter is doing by recording it's action with high-speed video found in some digicams and mobile apps.

 

[reddesert]

Mechanical shutter even electronically controlled isn't very accurate because it depends a lot on the curtain speed so there is a need for checking. I wonder at final assembly don't they check it.

I'm sure they check it during assembly. However, whatever diagnostics and test fixtures they have access to are far beyond us. For all we know, in some cameras, the shutter module might have its own internal instrumentation and diagnostics - it might be able to measure something like when the second curtain fires or when it closes, and whether that's in spec. We know that many electronic SLRs have an option for second curtain flash sync, so they have access to at least two different timings (first and second curtain). That doesn't guarantee that the second curtain goes mechanically just after the timer fires the sync, but in practice it seems to work.

It is likely that when the shutter throws an error, the service recommendation is "replace the shutter unit." So the testing may be limited to "does it work within spec?" If not, they put in a new one, they don't adjust some gears.

It occurred to me that if you want to estimate the shutter timing at fast speeds, you could take advantage of the focal plane shutter slit vs electronic flash timing. Normally of course that's a disadvantage, but if you have a flash unit that can be set to a repeatable short time (like with a 1/128 power setting), you can take pictures at various different shutter speeds faster than the sync speed, and see how much of the frame is exposed. You wouldn't get an absolute measure of time unless you measure the duration of the flash pulse (perhaps with a fast photodiode and an oscilloscope). However, this would be a way to compare two different cameras' shutter speeds, if that is your concern.

 

[grain elevator]

If you aim the camera at a blank wall, and manually set the indicated combination of shutter + aperture + ISO, and the histogram peaks at the midpoint of the histogram, that is an indication that the camera is achieving the 'proper exposure' for that combination.
So if you do that for every equivalent combination, using every shutter speed possible on the camera, and they all result in the same position of the peak of the histogram, then you see that every shutter speed is functioning properly, don't you?!

This is the obvious practical answer.
There's no reason to assume a proportional error. Any likely error will show more in shorter exposure times. So make use of reciprocity.

 

[4season]

What makes this endeavor a bit trickier is that the sensor itself can perform shutter-type functions, and for all I know, this may be where the most critical timing is done.

Upon closer examination of the Nikon ZFC service guide, I'm struck by how simple it is: The procedures shown are mostly about replacing a damaged LCD, buttons, switches and parts of the outer shell most likely to be damaged if the camera were dropped. Should the shutter or image sensor be damaged, they tell you to replace the camera. Have seen this sort of thing before in some Sony service manuals for lower-end products.

 

[abruzzi]

My Leaf digital back actually gives a readout of the detected shutter speed, but that doesn't really work for a standard fixed-sensor camera. It sounds like you'd need to take photos in a very controlled environment, i.e. an 18% grey card in a studio, with a recently calibrated light meter, take a spot reading off the card, set the camera to that setting, take the photo, and review that the grey is recorded as middle grey. If its brighter, your shutter is too slow, if its darker your shutter is too fast. If you want to know by how much, you may want to take several shots with larger or smaller apertures, but the same shutter speed, so you see where the gey fall with exactly half or twice the light. from that you should be able to best guess the inaccuracy.