5D Mark II Shutter Exposed!
 

My son Nathan and I dug into the 5D Mark II shutter and found some surprises:

1) Just because the camera says it's shooting a given interval, that doesn't mean that it actually is - even between 1/40 and 1/120.

2) Does it do faster shutter speeds in video? Heck yeah.

Some notes:
* Our tests were run with the Canon lens twist method, so this also applies to Nikon lenses on an adapter.

* We shot everything at 100 ISO, because that lets us control the shutter speed on the display.

* We didn't shoot at 1/30 because that automatically defaulted to 3200 ISO.

To see the video check out our new site: ~ The Murder of Dirk Snowglobe - A Not Dead Detective Series ~. We're promoting our next video project there. After the article drops off the front page, you can access it here: ~ The Murder of Dirk Snowglobe - Article: 5D Mark II Shutter Exposed! ~

Hi Jon, Nathan

Good work. But what do you make of this:

As I am in the land of PAL and get flicker on some shutter speeds.

Using lens twist and bright light to get down to ISO100 I find if I lock off with a shutter speed range of 1/60 to 1/800 I get no flicker until I hit 1/100 but if I change the light input and then re lock to the shutter range 1/50 to 1/800 I get no flicker on 1/100 and flicker will start at 1/125.

Seems the camera adjust the shutter speed relative to the shutter range.
Is there anyway you can try this with your turntable?

Thanks, James

P.S.

I myself have been messing around with the camera for two weeks and I do find it a good challenge.

Nice work!....

Any tips on locking the shutter down???

Here's the process we used.

1) Set your EOS lens to the desired shutter, then twist - or use a Nikon lens on adapter. The advantage with the Nikon is that you can set aperture at any time.
2) Enter Live Mode.
3) Turn the exposure adjustment to -2 using the large dial on the back of the camera.
4) Shine enough light into the camera to get it down to 100 ISO with a 1/250th (or so) shutter. Hit "Exposure Lock" immediately.
5) If you didn't get 100 ISO and ~1/250, press the WB button on the top of the camera twice to exit exposure lock. You can then repeat step 4.
6) Once properly locked, turn the large dial to adjust the shutter.

The problem with this in the field is that, while you can control your shutter speed, you are stuck at 100 ISO. In the scenario where you want everything in your film at the same shutter speed (say 1/80), you can only adjust your exposure with aperture, as your ISO is fixed. With the Nikon, you can adjust at any time. With an EOS lens, you need to twist the lens back on, tweak your aperture, and start the process over.

James,

I'm not sure why your flicker changed. I did a test at various ranges, and everything looked identical to the naked eye.

I tested 1/80 and 1/100 at these ranges:
* 40 - 640
* 50 - 800
* 60 - 1000
* 80 - 1250

PLEASE confirm if you are using exposure simulation mode in the live view settings

??????????

Hello Nicky,

We did not use Exposure Simulation. We used "Movie Display" Mode for the video. The Shooting mode was "M" for "Manual." (However, "M" for "Murder" is a much better title. :))

I wish we could use Exposure Simulation. It's a crime that the settings are lost when shooting video.

Thanks for this. It's very useful.

(Odd how you read off shutter speeds—I've always expressed them as fractions of seconds. "A hundred twenty-fifth," "one one-hundredth," e.g.)

Thanks, John,

I think the way I read the numbers shows my background as an engineer, rather than a photographer.

I'm not quite sure what you mean, John - I've always seen shutter speeds represented as 1/30, 1/60, etc.

I think the graphics are fine. John was probably talking about our voiceover.

Did you notice that the shutter speed being displayed on the screen while recording video is the shutter speed that would be used if you take a still photo (in the middle of recording) ?

For example, if you point the camera at a very dark subject, it would display 1" (1 second), but it's recording at the slowest possible speed for video which is 1/30 (and it would be underexposed); and if you shoot a still photo, the photo will be exposed at 1 second (properly exposed).

Similarly for speeds higher than 1/125, it would display 1/320 on the screen (when you mount a lens with focal length=300mm), but the fastest speed for video (with an EF lens) is 1/125; and if you shoot a still it would be exposed at 1/320.


And I believe the same thing is true for speeds between 1/30 and 1/125, where, as Jon mentioned, the speed displayed on the LCD is not the actual speed used on the video, but the speed that will be used when you take a still photo.

It does seem to use the same aperture and ISO for the photo as the one that's displayed on the LCD while recording video.


This is with "Movie display" of the live view, which is what I always use when recording video.
I haven't tried "Stills display" or "Exposure simulation".

Is this mean that you can't know the shutter speed before or when you lock the exposure?

For example . . .

http://www.vimeo.com/2530166 (here?)

Based on Jon's test video (first post on this thread) and my tests, I would think that
the speeds displayed on the LCD equate to the following actual speeds:
--- Anything below 1/60 (including those below 1/30) = 1/30 actual
--- 1/60, 1/80, 1/100 = 1/60 actual
--- 1/125 or higher (with an EF lens) = 1/125 actual

With a non EF-lens, you can go higher (in 1 stop increments):
--- 1/250
--- 1/500
--- etc.

I don't know the intermediate speeds that will be displayed which equate to each one of the actual speeds higher than 1/125.
For example I don't know if 1/160 means 1/125 actual, or 1/250 actual.
But if it follows the previous pattern, it would be:
--- 1/125, 1/160, 1/200 = 1/125 actual
--- 1/250, 1/320, 1/400 = 1/250 actual
--- 1/500, 1/640, 1/800 = 1/500 actual
---etc.
But I'm not sure -- Jon's video doesn't include these intermediate speeds.


In all cases, the speed displayed on the LCD will be the speed used when you take a still photo in the middle of the video.


Edit: After watching Jon's video again, my table may need to be adjusted. And Jon's test does not include 1/30.
But it's the same idea.

It would be great if the tests can be done with a faster turntable so that you can see a bigger difference (in the distance travelled) using different speeds.

Right … & it was charming, BTW. You and your offspring make a fine team.