I often hear photographers preaching the gospel of mirror lockup as a way to achieve sharper images. In fact, when shooting at a popular landscape photography destination like Yellowstone, you'll often hear most of the cameras using mirror lockup.
Mirror lockup is a function available in most cameras (though not some of the entry level Nikons and not in most Sony cameras due to the pelicle mirror). Mirror lockup makes the mirror flip up for a moment before you activate the shutter. The idea is that the vibrations caused by the fast-flipping mirror can cause sharpness problems. Generally, mirror lockup is used by landscape photographers and macro photographers who are shooting from a tripod.
So, in theory, mirror lockup makes sense. It stands to reason that getting rid of vibration would cause sharper pictures–especially for longer exposures. But I've always felt like it was a conspiracy of the camera manufacturers. I think they got together in some high rise in Japan and decided to put one feature in our cameras that was nothing more than a placebo, and they've been laughing about it ever since. “What fools!” they say, as we tout the placebo and use it so we can feel like our photos are sharper than the blockhead next to us who hasn't yet discovered it.
I'm not the only photographer who has doubted the practical effect of mirror lockup (MLU). A previous test was done that is specific to macro photography (and the result was the same as mine). I have periodically used mirror lockup, but to be honest, I've never really seen the difference. I just did what everyone else told me to do.
Mirror lockup has been in cameras for many decades. In fact, some early SLR wide-angle lenses couldn't mount without the mirror locked up. Mirror lockup made more sense in the film days when the mirror flopped with greater force. Since then, efforts were put in the design of the mirror to reduce the vibrations it caused.
The Mirror Lockup Test
This week, I put it to the test. I shot comparison photos for hours on different camera systems, different lenses, different focal lengths, different apertures, different tripods, different distances, and every imaginable shutter speed. I wanted to know once and for all if it was a useful feature, or if the Japanese high rise meeting was real.
I created a spreadsheet and catalogued the camera settings, distance, and many other relevant data points for each of the shots so that no human error could impact the results. After the test was finished, I took the photos to Dustin and had him look at each of the pictures to guage sharpness in a blind test without telling him which shot was and wasn't shot with mirror lockup.
All photos were shot using contrast detection focus with a perfectly well-lit focusing target. Sharpness on the unprocessed photos was judged at 100% view.
The Results of the Mirror Lock-up Test
Mirror Lockup makes absolutely zero visible difference in the sharpness of the photo IF you shoot from a rock solid tripod and ballhead. Period.
When the test was performed on my favorite tripod, there was no difference in the sharpness of the photos at all. However, when we tested lighter and cheaper tripods, poor quality mounts, or if the ballhead was not tightened down properly, mirror lockup made an obvious and profound difference.
First lesson learned from this test: If you use a high quality tripod, mirror lockup makes no difference at all. When I say “no difference” I mean the photos look identical to my eye. This was tested dozens of times at shutter speeds from 30 seconds to 1/2000th and EVERY shutter speed in between.
However, I also tested cheaper camera bases and heads. Second lesson learned from this test: On unsteady or cheap tripods and heads, mirror lockup made a profound difference but the difference was ONLY noticeable between 1/80th down to 3″ shutter speeds. The shutter speed most prone to vibration problems was 1/40th on the cameras I tested. All other shutter speeds showed no difference in sharpness with or without mirror lockup. At fast shutter speeds, mirror lockup makes no real difference because there simply isn't much time for the vibration to impact the photo. At longer shutter speeds such as those used for night photography, the vibration only occurs during a tiny fraction of the overall exposure time, so it isn't noticed.
For advanced photographers who use extremely heavy (and expensive) lenses that have a mount on the bottom of the lens, you should know that in my testing of the 70-200mm f/2.8 lens, I saw that mounting the lens to the ballhead proved to be a very unstable platform. With the lens mounted to the ballhead, MLU made a profound difference at 1/40, but when I mounted the camera to the ballhead with the same settings, there was no difference between the MLU and non-MLU photos. Obviously, the danger in doing this is that the weight of the lens pulls on the lens mount and destroys your camera… so I don't recommend it at all. I only did this for testing.
Conclusion
I'm done with mirror lockup. It is one more setting I can ignore when in a shooting situation. To me, it's important to pare down my gear, settings, and everything else during a shoot so that I can focus on creativity and lighting. It is important to remember that using mirror lockup certainly will not hurt the sharpness of your pictures, but I find that being overly burdened with too many camera settings negatively impacts my ability to make a great photo.
However, if you don't yet have a professional tripod and ballhead (click here to see my good, better, and best tripod and ballhead recommendations), then MLU is one setting you should pay attention to when using shutter speeds between 1/80 and 3 seconds.
As mentioned previously, advanced photographers should know that mirror lockup does make a difference when shooting extremely heavy lenses where the lens is mounted to the ballhead instead of the camera.
hi guys
l just got the nikion ds3100 dslr l wont do astrophoto what be the best shetting for the moon
I have a fairly good Manfrotto tripod & head ($700 worth) but when taking star photos, the effect of mirror lockup is very obvious. So, no, not a conspiracy. For me, essential.
Are you sure the movement in the pix wasn’t due to the tripod ball head creeping during the exposure, thus blurring the shot??
How can you be sure it was the mirror that caused it? What kind of flooring were you on?
Two thing that are not addressed in this article. Those that shoot multiple exposures for HDR processing. Even though I’m not usually worried about the effect of the mirror flip degrading any single frame, I do find that the repeated 3 or 5 frame snap can cause each frame to be slightly out of alignment with the other frames in the set. Now I know that the HDR processing software usually does a pretty good job aligning the images before blending but occasionally it misses and I find it more convenient to use the MLU feature. I also shoot astro-photography where even a fractional hiccup from the lens snap can cause a pin-point star to do the jig (and in this case the tripod used is of a very sturdy build – and in the thousands of dollars price range).
When I am doing more conventional photography I do not bother with the MLU feature though.
Congratulations on writing an informed, brave, and useful article. You hit the nail on the head when you observed that the original function of a mirror lock-up was to accommodate wide angle lenses, which intruded into the camera body, and their rear element could otherwise foul the mirror. Historically, that’s why so many 35mm SLRs didn’t have a mirror lock-up … because it wasn’t necessary. Including such a facility to improve slow speed shooting would have been an indictment of mirror dampening performance … like Ford motors including wheel chocks to supplement a parking brake.
Thank you for you efforts … I love it when people ask questions rather than repeat the same old story.
Thanks for this tests, I just got my shutter release cable , and 10 stop ND filter, and want to use it with the MLU, as I have read it could prevent blur in the photo. Now, I know when and where to use it
Curious about if MLU will make any differnece now they are removing the blur filter from 24mp aps c cameras. Can you be bothered to do it all again?
In fact curious about if there is a resolution or pixel density threshold that would benefit from it as some of the reviews of these high density sensors seem to indicate they are more easily affected by camera shake.
Thank you for the information, I found your artcile while googling because I was a bit suspicious about MLU.
I assume you did all tests without IS but did you try IS in the range 1/80 – 3 s?
Mirror Lock-Up (MLU) is not the same as Mirror Up (MUP).
MUP is a one-time temporary lift of the mirror as a prelude to a single exposure
during which time the viewfinder is inactive. Occasionally one hears of multiple
exposures, burst/continuous exposures, or bracketing exposures being done
after a MUP; but a MUP mirror ALWAYS flips back down after such exposure cycles.
MLU is a mirror up condition in which the mirror is locked up until purposefully released. During this state the viewfinder is also inactive. However, you are able
to make an arbitrary number of successive exposures. This is useful for focus-stacking because you want to do focus-bracketing without the vibration from
flopping the mirror up and down each time.
Unfortunately many DSLR firmware tend not to support MLU (except for such things as manually cleaning the sensor). They usually don’t allow you to make an arbitrary number of focus-bracketed exposures during MLU. There may be a way to do this through USB-tethered remote control software … But for Nikons it seems, the closest you can get is MUP.
Tripod dilima, I would like to see the procedure for determining if my tripod is moving or not! Can’t seem to find an article that indicates tripod movement. I don’t want buy a tripod if mine is doing the job. Most tripod articles praise features other than camera movement. I have an old aluminum tripod which does make clearer pictures when I use it. So how do I test the tripod for movement!
It’s an interesting read and has clearly been helpful to you,so obviously, not a waste of your time. I’m just wary of accepting your results are universal, though. There are a lot of different camera/lens combinations on the market.
Do not overlook the point that not all shutters are created equal, either. That’s a separate point from mirror vibrations—of course—but the problem is that the shutter cannot be excluded from the image capture process and, hence, quality of the results (assuming it’s a physical shutter, naturally).
I’d like to take the opportunity to note that the larger the camera and mirror become, the more beneficial pre-firing a single lens reflex one may be. There is a good reason why relatively few of the best medium format film SLRs lacked some means of lifting the mirror prior to exposure. A 6×6 or 6×7 sized mirror has a darned sight more mass, no matter how well damped it’s mechanism is. For instance Rollei’s SL66 was one of the smoother mirrors with its neat pneumatic damping—but a pre-release was still included for use if desired. I can’t speak to the characteristics of a modern medium format digital SLR, not having used one—except to note that they tend to almost without exception be smaller “formats” than their film counterparts, and perhaps also their mirrors, in some cases, at least.
Lastly, you are referring to “MLU”—but, is the functionality your camera incorporates actually a MLU? I suspect it may not be.
You cited a prime reason yourself why some of the better 35mm SLRs did indeed have a mirror lock up system—Ie that their ultra wide angle reflex lenses were of non-retrofocus design, and hence unable to be fitted, with the camera mirror in its ordinary rest position. Thus, an additional mechanism was designed that overrode the spring which lowered the mirror after exposure, to mechanically “lock” the mirror in the retracted position. By describing this feature as a “mirror lock up” their manufacturers were exactly right—once engaged, the mirror would be physically prevented from descending again, until such time as the MLU was deactivated. “Locked up”, indeed.
Less frequently a MLU might be provided for other purposes besides fitting ultra wide glass—for instance the original Bolskey-conceived ALPA Reflex and its derivative the ALPA Prisma had a MLU lever to permit certain collapsible lenses (such as their 50mm Angenieuxs) to be collapsed, without crunching the mirror in the process (and woe betide any owner who failed to use it).
On the other hand, quite a few SLRs would pre-fire their reflex mirrors if the shutter was tripped using the self timer. Thus, not facilitating the use of non-retrofocus wide angles, but still conferring the benefits of reduced vibration when utilised on a tripod. This was more common than a true MLU, which in most (but not all), cases, tended to be found only in the most expensive, professional models.
Cameras which feature a genuine MLU include: the early ALPA Reflexes until the Alnea 4 (some later types such as 11si for instance also had it); the Zeiss Ikon Contarex and Contarex Special; the Nikon F; Canon FT; Minolta SRT-101 and certain later SRTs; Nikon Nikomat/Nikkormat; Yashica TL Super & TL Electro X; Olympus OM-1; and the original Leicaflex. This is not an exhaustive list, obviously. A few of these cameras (Eg the ALPA 11si and Electro X) incorporate both MLU *and* pre-release with self timer.
When you set your DSLR(s) to lift the mirror prior to exposure—if said mirrors do not remain up, until you subsequently release them—they’re more accurately defined as having a pre-fire or pre-release function. Although the term MLU seems to be used generically today, as a term for any system enabling the mirror to be raised before exposure, literally—mirror lock up—as mentioned above does actually mean, precisely that.
Thanks for an interesting report!