Canon Mirror lock-up ... is it so hard?

[Paul Sumi]

I am not going to mention the rumors of Canon releasing news of the 12 MP, 1.3 crop-factor, "tweener" (3D?) on 8/23 with a stationary pelical mirror that doesn't flip.

No way. My lips are seeled.

Aren't they actually talking about a FF 5D?  Not that I'm spreading rumors, either.   :cool:

[sergio]

Canon MLU is very poorly designed. I keep forgetting all the time it is on and missing precious opportunities. It really is a pain. And I don't like have the drive in multishot mode. I really think a pro camera deserves to have a dedicated button for this task, just like in the old days.

[Ray]

I guess the reason Canon have not allocated a separate button to this function is because they've dampened mirror vibration so well they believe MLU is now rarely needed.

I've always found that a particularly sturdy tripod in a totally breeze-free environment and a shutter speed between 2 secs and 1/30th are prerequisites for MLU to serve any purpose. I find that I am rarely in such a situation. In fact most of the occasions I've used MLU have been during testing of lenses (to be sure, to be sure) and testing for the conditions under which MLU might serve a purpose   .

[crspe]

So ... is it only hacked 300D's where MLU is disabled in multi-shot mode? There was only 1 reply about the 20D that it is not disabled there.  How about with other Canon cameras? It seems amazing to me that MLU is not switched off in multi-shot mode. What happens when MLU is enabled, in multi-shot mode if you hold down your finger? Do you get multiple shots? Does the viewfinder just stay black the whole time? what is the shot rate? MLU delay then full speed, or MLU delay between every shot?

Its almost ironic to think that in a hacked 300D, MLU is fixed whereas in all other Canon cameras, it is effectively broken / hardly usable.

[rokkitan]

My suggestion (to Canon) is to make MLU part of a 2 second timer setting. Then it should be easy to switch to: single, continous, 10 second timer and 2 second timer w/MLU. A Custom Function could enable/disable it: disabled, 2 sec. timer enabled, 2 sec. timer with MLU enabled, both enabled. -RKS.

[pfigen]

"As well-dampened as Canon's mirrors are, they still impart notable image blur between 1 sec and 1/30th sec."

If the mirrors were indeed well dampened, this statement wouldn't be needed. Canon's film camera mirrors appear to be dampened far better than their digital cousins. The Nikon film cameras I used to shoot with were so well dampened, that I rarely used MLU at all.

"You can get away with a single button press if you are using either the 2-seoond or 10-second delay drive modes. When you press the shutter, the mirror pups up, the timer counts down, and the shutter trip"

Unless I'm missing how to do this, and I just referenced the entire manual, neither my 1Ds or 1DsMKII, when operated in either 2 or 10 second delay, flip the mirror up at the beginning of the sequence, only at the end of the delay just prior to the shutter opening. This would be great if it actually worked.

[Ray]

However, haing taken just recently a few shots of my neighbour's fence with my 100-400 securely fixed to a tripod, at a range of apertures ISO's and shutter speeds, all at 400mm, I've been able to extract from the confusing mass of variables some direct comparisons.

At 1/8th sec, at ISO 100, MLU on produces a marginally sharper image that would be apparent on a very large print; much larger than one would normally print from a 20D image.

At 1/125th sec and ISO 1600, MLU on also produces a very,very marginal increase in sharpness at 300% on screen which just might be visible on a 4x6ft print scrutinised closely. I would consider this to be extreme pixel peeping and of no practical significance.

I see no point in comparing images shot at different ISO's and apertures (well there was a point. I've just done it and found there are too many variables.)

Of course, I understand this test is not the last word. The 100-400 is not the best of lenses. It may well be that a $10,000 600mm lens on a 1Ds Mkll will show greater differences in the same circumstances. But I'm unable to test such lenses. I'm content knowing the limitations of my own equipment   . All else is academic.

[Ray]

I don't think mirror slap is much better than it was 20 years ago. In fact my 20D has more slap than my A1 and not much less than my Ftb.

I'm sure mirror slap will behave differently on different cameras. One would certainly hope that modern materials and design would improve the situation. The 20D mirror is a particularly noisy one, yet it's smaller and lighter than that of a full frame 35mm DSLR.

Noise of itself does not degrade image quality (I mean, audible sound waves as opposed to lower frequency vibration). Shout as loud as you like when pressing the shutter but I doubt that image quality will be affected. (Although I have to admit I haven't tried this  ). I wouldn't be surprised if the noisy mirror slap of the 20D is actually a deliberate engineering trick to transform harmful low frequency vibration into harmless (but audible) high frequency vibration.

I took a number of shots today with the 100-400/20D combination to see if the telephoto effect extended the normal range of slow shutter speeds where MLU can be beneficial. It doesn't appear to. At 1/60th and 1/125th there was essentially no difference between MLU on and MLU off, at ISO 100 and f11. At 1/30th and slower, there was  noticeable but slight degradation in 100% crops on screen, with MLU off, but it's doubtful that such differences would be noticeable on an A3 size print after the usual processing and sharpening.

The tripod I used is nothing special, just an average, inexpensive Slik Able 300DX. I used a cable release. Some shots were taken inside and some outside. It was a particularly calm day.

[Jack Flesher]

Jack,
Where are the tests or reports that show this?

Nowhere, I compared them for myself...

But to be clear I compared my 1Ds2 to my Rebel XT, not a 20D; the 20D could certainly be better than the Rebel.

[jani]

FWIW, mirror slap has always been better -- meaning less of it -- in heavier cameras.  I assume this is a function of inertia providing added damping for the harmonic oscillations.  Anyway, it would explain why the mirror slap in the 20D is more visible than that from a 1-series camera.

Well, the FOV factor could also come into play here.

[jani]

Also, on the 20D (at least my camera, firmware 1.1.0), setting the camera to continuous shooting does not turn off the mirror lock-up.

It annoys me a lot.

There is just one thing that makes it slightly better than destructively distracting, and that's the fact that when I press "menu", at least "custom function" is already selected, and entering there puts me right at mirror lock-up, because that's where I was last. So it's "only" the following steps to disable:

Menu, set, set, turn wheel, set.

Of course, if I was shooting with auto-bracketing, I'd have to go to two different menus to get a sensible configuration for hand-held shots again.

This way of handling the user interface ruined some wedding shots, when a friend was taking images because I was the best man. He had to change to one of the non-RAW modes in order to take images at all (because he couldn't find what was doing that mirror lock-up in the menus, no wonder ...), was thoroughly distracted, and unfortunately the white balance was way off in the few pictures he had a chance to take.

Canon needs to fix the user interface for mirror lock-up. And auto-bracketing.

[Ray]

Yes, I do that often. But still, when the wind is blowing and you are waiting for that break so the flower stops wagging, the 2/10 sec delay does not work...

This is the sort of situation where the advantages of MLU are highly dubious. The sorts of long exposures that create a problem without MLU, also create a problem with the slightest breeze.

I suppose it's quite likely one would need to use a really slow shutter speed in moderately low light if the flower was the foreground to a wider scene that one also wanted to be in focus, ie. one is using f16 and a low ISO. However, in such a situation, when the flower is still, other parts of the image are usually affected by breeze, causing blur with or without MLU.

[Stealthfixr]

Having played with a Minolta 7D recently, I found my 20D's MLU to be a clunky & slow in comparison.  Canon makes some great cameras, no doubt.  They also seem slow to adopt user driven changes/inputs.  I think Jack's idea above is a great one ... Canon, can you hear us?!