13

Exposition

I know what mirror lock-up is and what it's good for, but I'm curious about the range of shutter speeds where it provides a real benefit.

A little background

I use a nice, sturdy tripod for shooting still life photos, and sometimes portraits. More often than not for still life shooting, I use live view either because the camera is at an odd height or angle, or because I'm shooting in very low light that makes it difficult or impossible to compose and focus through the viewfinder.

There are previous questions that ask about whether the mirror flips back down and then up again during live view shooting. In my camera the answer is yes no, but there's more to be aware of (detail further down).

The Blanston Hypothesis

It seems that if the shutter speed is fast enough, then any vibration of the camera would be insignificant because the image is captured too fast for the camera to move too much during the exposure. And it seems that if the shutter speed is slow enough, the short amount of time that the camera vibrates wouldn't matter because it would be buried below the noise of the capture (assuming very low light, no flash, etc.). So I figure there must be a range of shutter speeds where mirror lock-up makes a difference in image quality. It wouldn't surprise me if it's related to focal length, sort of like the 1/(focal length x crop factor) guideline for non-IS handheld shooting.

Recapitulation

So, as the title states, at what shutter speeds is mirror lock-up worthwhile? Is my reasoning correct (or at least sane)?


The detail I promised you earlier

This answer indicates that live view does accomplish mirror lock-up using a Canon 70D, and it appears the same is true for my 80D, but there is an additional menu setting I needed to be aware of: Silent LV mode.

When I use mirror lock-up in normal (non-live view) mode, I can clearly hear that the first curtain noise at the beginning of the exposure is a very minor "tick" sound, which makes sense. If I do this with a suitably long shutter speed (say 1 second or more), I can clearly separate the sounds at the beginning of the exposure from the sounds at the end, when the mirror flips back down.

However, when I use live view, I can very clearly hear noises that I first thought was the mirror moving at the beginning of the exposure. It turns out the mirror isn't moving, but something about the way the shutter activates in this mode created enough noise and vibration to trick me into thinking that's what was happening. When I found the Silent LV option in the menu, it was set to "Disable." Once I changed that to either Mode 1 or Mode 2, the noise and vibration more closely matched what I'd observed when using lock-up and viewfinder shooting.


Coda

This question was initially posted because I thought that I couldn't get mirror lock-up in live view mode. It turns out I can, but I'm still curious about the answer.

  • 2
    (1) What do you find in the menu under Silent LV shooting? Mode 1, 2 or disabled? (2) I don't own the 80D. With the 5D3 I can set phase AF for LV. That requires mirror flipping for each AF measurement. Does that exist in the 80D? – bogl Mar 15 at 15:29
  • 3
    If there was an award for a damn good, well written question, I'd give it to you. – Hueco Mar 15 at 15:44
  • 1
    @Hueco Thanks! As a show of good faith, I'm willing to accept this in lieu of an official award. – Gern Blanston Mar 15 at 15:46
  • 2
    I'm afraid that I've blown my quarterly gear and film budget on an RB67ProS. You'll have to post again next quarter I guess :-). – Hueco Mar 15 at 15:49
  • 1
    @bogl Thanks for pointing out that menu item to me. It turns out that the mirror wasn't flipping back down before shutter release, but with that menu set to "Disabled" as it was, something about the shutter actuation in LV created about the same level of noise and camera vibration that I get from a normal, non-mirror-locked shot using the viewfinder. Setting either Mode 1 or Mode 2 reduces this vibration substantially. I'm still going to leave the question up because I'm still curious about this. – Gern Blanston Mar 15 at 22:43
2

There are a few other factors not mentioned in your question that affect when mirror lockup may be beneficial:

  • Focal length. The same vibrations affect images that use longer focal lengths more than images that use shorter focal lengths.
  • Sturdiness of the camera mount. A very sturdy tripod will dampen vibrations faster than a less sturdy tripod, which can actually amplify them in some scenarios.
  • The contents of the scene. An evenly lit scene will be less affected than a mostly dark scene with a few bright light sources. They can leave squiggly trails, even if there's only one second of vibration during a 30 second exposure.

Although it varies based on camera design, most researches that have thoroughly tested such vibrations place shutter speeds between about 1/100 second and 1 second as the most vulnerable to mirror vibration affecting the image.There have been a few controlled tests that found the maximum effect of mirror slap with the cameras tested was around 1/15-1/30 second. How quickly the effect drops off on either side of 1/15-1/30 second depends on several factors including those listed above and in the question.

Mirror lockup is most useful when using a very long telephoto lens or when doing high magnification macro work. Because the field of view in these cases is limited to a very small angle, they are the two situations that are most likely to result in vibration caused by the mirror to be visible in the exposure. It should go without saying that to get any benefit from mirror lockup the camera should be tripod mounted and you should use a cable release, wireless remote, or self timer to prevent vibration caused by pressing the shutter button.

One application I use it for is taking pictures of the night sky. Even with shutter speeds as high as 1/125 to 1/250 sec for the moon I get sharper results locking up the mirror when using an effective focal length of 640mm. Canon Super Telephoto lenses include an IS mode that is designed to be used while tripod mounted that will compensate for mirror vibration.

In general, once exposure times are over 1-2 seconds mirror slap becomes much less of an issue because the vibration duration is a much lower percentage of the total exposure time. Note that the duration of the vibration will be affected by the sturdiness of the camera mount. A sturdy tripod will kill the vibration much faster than a less sturdy tripod.

But mirror lockup can also be useful for exposures longer than one second or so if there are bright light sources in an otherwise fairly dark frame. Stars in the night sky, streetlights over a dark street, etc. can leave squiggly light trails due to that one second or so of mirror vibration even on a 30 second or longer exposure!


Addendum:

Depending on the autofocus method being used, the mirror can cycle down and then back up just before an exposure even when shooting in Live View. This is to allow the PDAF system, rather than the main imaging sensor's CDAF, to focus the camera.

Depending on the exact Canon camera model and the 'Silent Shooting' mode selected, particularly in Live View, the shutter may or may not operate as you expect.

  • 2
    Do you have any references? Not doubting you, just interested in any research that's been done in the area. Also reference are always good :). – damned truths Mar 17 at 12:29
  • In fact I did mention focal length and tripod sturdiness (briefly) in my question. But I'm glad you included them in your answer for context. – Gern Blanston Mar 17 at 15:25
  • @GernBlanston I do see where you mention using a tripod (but not anything about how different tripods might affect the answer to your basic question), but I don't see anything about focal length in the question. – Michael C Mar 18 at 21:14
  • Different Canon models do some things re: silent shooting differently. I've not played around with an 80D, but based on my experience with other Canon cameras, particularly the 7D Mark II that preceded the 80D by about 18 months, my heavy hunch is that in LV when SS is set to disable, the second curtain is closing, then both shutter curtains are reset to their normal (non-LV) position before the first curtain is opened in the conventional direction to begin exposure. – Michael C Mar 18 at 21:22
  • When in LV the first curtain is open at the bottom of the light box (top of inverted image) and the second curtain is open at the top of the light box. Electronic shutter begins from bottom to top of the camera (top to bottom of image), then EITHER first curtain closes by moving up in the camera (the direction it normally travels during a reset after the image has been taken when using the viewfinder) OR second curtain closes by moving down across the sensor (bottom to top of image). Whichever curtain closes then reopens to enable LV. – Michael C Mar 18 at 21:28
5

Your hypothesis is correct.

  • If the exposure is short enough, vibrations caused by the mirror will be irrelevant.
  • If the exposure is long enough, the vibrations from the mirror will decay soon enough to have little impact on the exposure in total.

For a specific camera, the Pentacon Six, a relatively heavy medium format SLR, this document states (in German) that vibrations from the mirror are e.g. only an issue with the shutter speeds 1/15 and 1/30 seconds. With 1/8 (and longer) or 1/60 (and shorter), mirror induced vibrations should not be a problem.

These are however numbers for a specific camera and it will be very difficult to give a general advice, since the vibrations' amplitude, frequency and decay curve will depend on several variables like e.g:

  • The weight of the mirror will matter. A heavy mirror will cause more vibrations than a light mirror.

  • The speed of the mirror will matter. The faster the mirror moves, the more vibrations will be caused.

  • The weight of the camera (and everything attached to it) will matter. A heavy camera will vibrate less than a light camera. This means that even if a longer lens and a narrower field of view in general makes the exposure more prone to shake- or vibration induced blur, the extra weight of the longer lens and hence the higher inertia, may improve the behaviour.

  • Using a tripod may or may not help. The increased weight of the overall system may help (see last point), but on the other hand, many tripods are not overly stable and especially if the camera is attached to a long, extended centre column, the impulse from the mirror may cause the column to swing more and for a longer time, than if you tried to hold the camera steady in your hand.

So if you really need to know when you have to use the mirror lockup: Setup your camera in a specific configuration (with lens and tripod) and make test shots. Be aware though, that you will have to do this for any configuration combination you are going to use.

  • At higher shutter speeds the effect of mirror slap is reduced. Here's a chart showing the effect on mtf at 1/500 sec. It's significant and you need to go to 1/1000th or faster to pretty much eliminate it. Resonant freqs alter things below 1/100 and there you get more system dependencies. At the faster speeds it's pretty much a factor of the mass of the lens/camera/focal length and gets worse linearly as the speed decreases. See dpreview.com/forums/post/56681755 – doug Mar 16 at 3:01
  • It could also depends on using "Silent" mode (which, AFAIK either moves the mirror more slowly or somewhat earlier than the shutter operation). – xenoid Mar 16 at 8:53
  • @bogi No, this is not directly related to the resonance frequency. If you bump into an object and it moves, e.g. if the mirror bumps into the camera and the camera moves, the velocity of the bumped object's movement is related to the conservation of energy and not to the resonance frequency of the objects. This is relevant for the shorter shutter speeds. For the longer shutter speeds, it is relevant how long it takes to dampen the vibrations. The vibration frequency can be far above the reciprocal of the shutter speed and still cause blur. – jarnbjo Mar 16 at 10:47
  • Thank you! I like the technical explanation here. Unfortunately, my laser vibrometer is out for calibration right now, so I need to make do with a more general, rule-of-thumb answer. – Gern Blanston Mar 17 at 15:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.