Shutter shock is an issue that has plagued early models of mirrorless cameras. Unlike in a DSLRs, the shutter must wind to cover the sensor very quickly, then release. This additional movement causes mirrorless cameras to be more susceptible to the shutter action. DSLR cameras are not immune, and have their own host of problems in the form of mirror slap, but it was in mirrorless that this issue came to the forefront. This site’s interest in shutter shock should be clear. To what extent can a tripod minimize the effects of shutter shock? In order to test this I am going to use a modified version of the MTF testing I described previously.
Previously I used a backlit razor blade as an edge to test sharpness. The razor blade is an excellent high contrast, high resolution target, but the images are time consuming to post process through MTFmapper’s single ROI function. Now, I am going to use one of the tilted square targets that MTFmapper automatically identifies. This has two advantages. First, I don’t have to crop the images before feeding them into MTFmapper. Second, the tilted square gives me measurements of both the horizontal and vertical movement of the camera. The test target is shown below.
I am still illuminating it by a pair of adjustable color temperature LED lights. The camera setup is shown below:
The camera has a Singh-Ray 77mm Variable ND filter attached to the lens. In this particular case, I am using a Fuji X-T2 with a Rokinon 85mm 1.8 manual focus lens. This isn’t the ideal camera for testing shutter shock, but it is the one I have on hand to begin with. I am using a RRS TFC-14 tripod with the BH-30 head. This is middle of the range in terms of stiffness.
The procedure for testing the shutter shock is as follows. Focus the lens to the target wide open, then stop down to f/4. Maximum light on the target. ISO 400, Raw. Set the desired shutter speed on the camera. Adjust the variable ND filter to the point that correct exposure is achieved. I started at 1/500s and worked down to 1s in 1/3 stop increments. Import images to Lightroom and export to tiff with default settings. Open images in MTFmapper to analyze edge acuity.
There is a very small effect of shutter shock right around 1/100 s shutter speed, and only in the vertical direction. The effect is so small that it would never be noticeable in every day images. It isn’t even noticeable when visually inspecting the test images. Despite being small, it is measurable. This is shown by comparing the vertical MTF for the mechanical shutter to the electronic shutter, as shown below:
The mechanical shutter is consistently slightly worse than the electronic shutter, and particularly so right around 1/100s.
To complete this test, I repeated it with the camera in portrait mode. The system should be more susceptible to shutter shock in this configuration, as the shutter motion is no longer pushing straight up and down on the tripod apex. In portrait mode, the motion of the shutter will cause a torque about the tripod apex in a motion corresponding to the pitch stiffness of the tripod. I have set up portrait mode by rotating the ball head into the side pocket, as shown below. Alternatively, I could use an L bracket to set up portrait orientation. I will try that in a later post.
The results are below. This time, I have just plotted both the MS and ES results on the same graph, as there is enough separation to be readable.
The effect of the shutter shock is much greater when the camera is mounted in portrait orientation as expected. Again we see near perfect results from the ES. The MS deteriorates the results, but again not in a way that the average user would likely notice. For example, here is the raw 100% output from MTFmapper of a mechanical shutter shot at 1/500s:
and here is one from a “bad” 1/60s shot:
The effect is hardly visible to the naked eye. Importantly though, this proves that this method is quite sensitive. We should be able to measure very small deteriorations in image quality.
The Fuji X-T2 when mounted directly to the tripod is clearly susceptible to a small amount of shutter shock, but not in a manner that would be particularly noticeable in everyday images. However, when in doubt, just use the electronic shutter. There is no downside at the shutter speeds where the system can detrimentally affected by the mechanical shutter. There are still a lot of questions that I have left to answer:
- How much does using a stiffer, or flimsier tripod affect the shutter shock?
- How much worse is the shutter shock when using a telephoto lens mounted on the tripod with a lens foot?
- Does using an L bracket help or hurt when the camera is in portrait mode?
- Does the shutter damping mechanism in the Fuji X-H1 help?
- Does the IBIS of the Fuji X-H1 help or hurt?
- Shutter Shock is a high frequency motion. Does adding mass help more than tripod stiffness?
- How much does changing the focal length affect the shutter shock?
Please post any additional ideas in the comments!
For an extension of the telephoto
lens foot versus camera plate idea,
you could use a macro rail mounted on
the camera plate to adjust the offset
between the shutter and the center of
the tripod.
You seem to have most of the bases
covered there, though.
I have a macro rail sitting on the desk for that purpose :).
I notice that the measured sharpness
begins falling off slightly as the
exposure time keeps going up in the
range of 1 second to 10 seconds.
It would be interesting to test how
this varies with tripod/head
stiffness and out to longer exposure
times; does it level off at some point?
I only tested up to 1s. It does fall off slightly, and I think that it is a result of the variable ND filter going outside of its normal operating range. As I approach 1s the images tilt heavily magenta and this can start to throw off the MTF measurement. Its a small effect, though good eye.
I just thought of another thing to test:
How does the additional joint of an
add-on battery grip affect this?
How does it affect it when you have a
battery grip equipped but are using
the lens foot?
Good question. I have a battery grip for the X-H1. I suspect it would make things worse when mounted directly to the tripod, but better when a lens foot is being used.
Very interesting results and consistent with expectations. I’m a little surprised there is not more difference
between ES and MS.
One potential way to look at telephoto lenses is based on the length of the lens or the extent to which the lens
extends while zoomed. Especially with lighter long lenses like the 200-500 and 150-600 models, the vibration is
a major problem without long lens technique. Long Lens Technique is a mitigating factor that reduces vibration
significantly.
Looks like the shutter shock problem is even amplified if you use OIS lenses, especially in the standard zoom
range with smaller and lighter moving OIS elements. There is big discussion now about that with the new XF16-
80 4.0 but this also happens with the 18-55 and 15-45. In this case the pictures are visibly unsharp in the range
of 1/80s – 1/200s and definitely not only “hardly visible to the naked eye” as you mentioned above….
Interesting article. I’ve been trying to learn more about shutter shock with my X-T2 because I recently
bought a 16-80mm that seems to be more prone to it, but, in my case, mostly at 1/30 to 1/60th second
handheld whether OIS is on or off. Otherwise, this lens using electronic shutter is quite sharp at most focal
lengths and I very satisfied with it.
My tests of this with my new 16-80 were all done handheld because that’s the way I take most of my
pictures. But no matter how many test images you shoot, human handholding adds another variable
because of the human element. Yet, a camera on a tripod can’t move the way a camera in your hands can.
In the very old days, Popular Photography used to test SLR cameras for mirror slap simulating
handholding by putting the camera on a piece of foam rubber and then exposing with the self timer,
theoretically eliminating the human variable but still allowing the camera to move as it would in your hands.
Seems to me this would still be useful.