E-P5 and "shtter Shock"

Started Oct 6, 2013 | Discussions thread
Anders W
Forum ProPosts: 18,703Gear list
Re: "sh_tter Shock": multiple cases and causes?
In reply to Ken Strain, Oct 9, 2013

Ken Strain wrote:

Anders W wrote:

3DrJ wrote:

Anders W wrote:

3DrJ wrote:

Anders W wrote:

3DrJ wrote:

Great term! Especially the "fill-in-the-blank" missing vowel between 'h' and 't'.

Putting a 'u' in there, and reading the numerous posts about it, quite a few hypotheses have been put forward that might account for unexpectedly blurry images. So it seems not a settled issue.

iple "explanations" for something, leathan one thing going on. Perhaps the "cause" of blur is different in e so easy.

So it's conceivable the problem could be "shutter button jab", shutter vibration, IS or AF mechanism "looseness", other factors not yet teased out, or >=2 of these at the same time.

As a side note, it's good to see this being investigated rather calmly and patiently, without "hysterics" that sometimes have been evident with other issues.

I've had instances of mysterious blur with the E-M5, but I've just started to review the situations re: lens, shutter speeds, etc., and consistent patterns not yet emerged. However, I'm going to keep looking at it, and maybe will have more to report later.

If you search for "shutter shock" on this forum (which you potentially might do even more effectively via google than the internal search function), you will encounter a large number of prior threads on the subject as well as quite a few tests. Many of the comments in the threads on the subject that have surfaced during the past few days are poorly informed because the posters aren't informed about what has been said and done in the past. In reality, I think we have a pretty good understanding of what's going on and not.

Can't say how informed posters have been, but to me it's pretty clear subtly blurred images can occur via an array of factors. Shutter-induced blur is not necessarily easy to distinguish from camera or subject motion blur.

Here's an example: I had a series of shots at different shutter speeds, holding the camera at a distance, using the monitor for framing. Of course, this is likely to give inconsistent results (vs. camera on a tripod), but illustrates what one might encounter in "real-world" camera use.

This scene was the test subject: ...

.... I am among those who on the forum who have been testing for this phenomenon most extensively. Consequently, I am of course perfectly aware of all of the above. In order to test in a manner that allows you to draw reasonably firm conclusions, you have to shoot quite a few images (especially if you are testing hand-held since the problem is probabilistic rather than deterministic; it isn't perfectly reproducible on a shot-to-shot basis), use a target that readily lets you see any minor blur such as the one exemplified here (the print-screen structure is particularly revealing)


inspect the results carefully and average them across the sample, and carefully compare differences in shooting conditions (e.g., the shutter speed used), taking care to vary only one condition at a time.

Can't disagree at all. I didn't regard my off the cuff results as definitive tests, but illustrating the possible confusing foms of blur that can appear at random in images.

Here's an example of fairly recent thread (two months ago) which contributes some new tests as well as links to previous ones:


Thanks for the reference. However I've seen the topic and read a good number of threads over the last 3-4 years or so. The fact that the topic emerges fairly regularly, and new tests are invented, seems to suggest variability of the shutter shake phenomenon. Naturally, interest in it arises with new camera models as each introduction potentially brings with it twists in how the problem presents and where it stems from.

Yes, the phenomenon is likely to vary not only on a shot-to-shot basis but also depending on factors such as for example holding technique, body used, and lens used (not only FL but also other things like length, weight, and OIS might make a difference) to name but a few. The fact that the results vary depending on such parameters does not mean that we do not or cannot have a reasonably firm grasp of the general nature of the problem. It just makes it more difficult to achieve it.

I wonder if the difficulty in pinning it down has to do with intrinsic variability of mechanical or optical devices, but perhaps also there are a number of unknown variables that remain unaccounted for. Obviously factors we cannot parameterize skew results unpredictably and make conclusions tenuous.

What makes this problem particularly difficult to pin down is the human factor as far as hand-held shooting is concerned plus the variability of tripods and the surface they are standing on in tripod-based tests. The variability across body models, lens models, and combinations thereof is of course a problem too. Finally, there is the possibility of copy-to-copy variations within bodies/lenses but I doubt that this is in reality a factor of major importance.

I think such variation should not be ruled out. I only have weak evidence: my E-PL2 had shock approaching an order of magnitude worse than my E-PM2 and E-PM1 which are roughly similar. This is based on similar photographs taken at different times with the same lenses and holding technique. The older camera was close to useless due to vertical double images. Your suggested holding technique helped. Fortunately a monopod also helped.

OK, those are different bodies, but similar size and perhaps similar IBIS design. Most E-PL2s CANNOT have been that bad or that would have been talked about.


Of course, the possibility of copy-to-copy variation can't be ruled out completely. In this case, as in many others, the matter can ultimately be settled only by having several copies of the same body and/or lens tested by the same user (or hand-shake simulating machine) in an identical manner. Regrettably, such tests are few and far between. Right now, I can't think of a single example as far as shutter-shock is concerned (except possibly with regard to the Pany X 14-42 and X 45-175 where I think there might perhaps be some examples out there from the time the special problems with these lenses figured very prominently in the discussion).

As to your E-PL2 versus your E-PM2 and E-PM1: Suppose it is indeed a matter of copy-to-copy variation in this case rather than a matter of design differences. How would you try to account for that variation? What are your thoughts about what the explanation might be?

Personally, I have a hard time thinking that shutters could be so different from one copy to the next (with regards to aspects relevant in this case) that it could lead to any variation to speak of when it comes to shutter-induced blur.

In recent threads it was discussed how DPR seemed to obfuscate the issue in the EP5, stating differences when using touch screen vs. shutter button. In other words, if we believe DPR's idea, the source of "shake" that affects images is broader than just the mechanical action of the shutter, thus confusion among camera owners as well as testers.

I wouldn't use a word like "obfuscate" but I am somewhat sceptical of DPR's idea that the shutter button and the way it is pressed is the main culprit as far as the E-P5 is concerned (particularly since some recent reports by Andy Westlake here on the forum indicates that the problem persists even with a self-release or anti-shock delay of as much as two seconds, which is sufficiently long to remove any shake due to the button-press itself).

Obviously, camera shake has many sources and that fact may well confuse users as well as testers. However, for reasons spelled out above, it remains possible to isolate the impact of shutter-shock from other sources of blur, provided you are ready to think carefully about how to do it and then carry out the test in a sufficiently rigorous and extensive way.

Here you are alluding to a great dilemma of experimental science. It's necessary to eliminate variables in order to test the hypothesis, but then it is unclear what the test result means in "real-world" circumstances when other variables are uncontrolled. This is seen in medical research all the time.

Considering the fuzziness that seems to be associated with shutter shock testing, even if specific tests point to factors that produce image blur, narrowness of the test environment and the stochastic nature of results makes it hard to know how to apply the test findings to everyday camera use.

While what you say about experimental science is true in general, the only major factor at play in this regard in the present case, as I see it, is the variability with regard to hand-holding techniques (and tripod characteristics). What you'd ideally want is a set of testers/tripods or a machine that can emulate various techniques. But that ideal is rarely meet. Nevertheless, multiple tests/testers plus a bit of discussion on the forum appear to take us at least some steps forward.

In other regards, it's not difficult to make the test scenario sufficiently realistic to be in a position to generalize, I think, at least if we restrict the generalization to the particular body and lens models actually tested.

The best we can do may be formulating a partial solution. That is, to refine knowledge of factors we can determine, e.g., shutter shock, techniques of releasing shutter, lens/OIS/IBIS effects and so on. Apply this knowledge to the best of our abilities to reduce undesirable blur, and this will be helpful. But in the end, we have to accept that photography is kind of a gamble, there is a certain probability a shot will be blurry, and some images will unaccountably show it.

It never pays to argue with mother nature ....

I have no problem agreeing with any of that.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow