5D2 FW 2.0.4 - Iris jitter in M movie mode

Home · Register · Search · Software · Hosting · Software · Join Upload & Sell

FM Forum Rules
Canon SLRs, primes, and zooms lenses reviews

FM Forums \| Canon-mount SLRs \| Join Upload & Sell
1 2 3 end

Archive 2010 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

wickerprints
Offline
• • • •
Upload & Sell: Off

p.1 #1 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

Read the following report:

http://www.canonrumors.com/2010/04/5d-mark-ii-firmware-2-0-4-faulty/

I just tested this with my 5D2 (FW 2.0.4) and with each of the five lenses I own. Here are the results:

EF 24-105/4L IS USM - normal
EF 70-200/2.8L IS Mark I USM - normal
EF 85/1.8 USM - normal
EF 100/2.8L macro IS USM - normal
EF 300/4L IS USM - ABNORMAL

I did not try any of the lenses in combination with the EF Extender 1.4x II.

Some people are reporting that this is a stopping down of the iris. It is not. Rather, it is a rapid fluctuation of the blades that is not more than 1/3 stop. From my testing, it is clear that it is not affected by scene EV or contrast. It does not affect all lenses. Once I noticed the issue in the 300/4L IS, I tried very hard to duplicate the phenomenon with all of the other four lenses, to no avail. It also does not appear to be restricted to a particular focus distance, since immediately after observing the jitter, turning the focusing ring back does not necessarily cause it to occur again.

The issue, in my eyes, is clearly a bug. The good news is that it appears to affect certain lenses only.

How about other 5D2 owners? Can you perform your own testing? I'd like to hear your results with your lenses.

Oh, and BTW, I noticed for the first time that the 100L macro has two aperture stops (only one is variable). That was something I could've seen from the block diagram but it was kinda neat to actually see it. So rarely do we look through the front of our lenses while they are operating.

Apr 06, 2010 at 01:15 AM

dancam
Offline
• • •
Upload & Sell: On

p.1 #2 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

I did the 2.0.4 update recently and haven't noticed a problem so far. But, I also haven't used my 300/4 IS since I did the update. I'd love to test it out but, my 5D2 is atCanon getting fixed

Apr 06, 2010 at 03:50 AM

skibum5
Offline
• • • • • •
Upload & Sell: Off

p.1 #3 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

dancam wrote:
I did the 2.0.4 update recently and haven't noticed a problem so far. But, I also haven't used my 300/4 IS since I did the update. I'd love to test it out but, my 5D2 is atCanon getting fixed

i didn't look over the report in detail but they mention changing iris in a macro when focusing.... isn't that mandatory once you focus closer than a certain distance (with a canon macro lens)

Edited on Apr 06, 2010 at 07:27 PM · View previous versions

Apr 06, 2010 at 04:09 AM

wickerprints
Offline
• • • •
Upload & Sell: Off

p.1 #4 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

skibum5 wrote:
they mention changing iris in a macro when focusing.... isn't that mandatory once you focus closer than a certain distance

I think whoever made that statement misunderstood the situation with internal focus macro lenses, and the relationship between f-number (relative aperture) and effective aperture.

Recall that effective f-number is given by

N[eff] = N (1+M),

where N = f-number, and M = magnification ratio. So, at 1:1 and f-number set to f/2.8, the effective f-number is actually f/5.6, two stops less light. But the size of the entrance pupil is no different; the effective f-number describes the loss of light due to the fact that the subject in focus is not infinitely far away (and thus not projecting parallel light rays into the lens).

However, when you change the focus on an internal focus macro lens, you will notice that the entrance pupil diameter also appears to change. (To see this, mount the lens on the body, set the aperture to, say, f/8, look through the front of the lens, and while holding the DOF preview button, turn the focusing ring from infinity to MFD.) The reason for this change is simple--the true focal length of the lens does not remain constant in an internal focus design. In non-macro lenses, the focusing helical doesn't really move far enough to make this change in focal length apparent, so you won't really notice a significant change in focal length for an internal focus zoom like the 70-200/2.8L IS (though I can't say that about the Nikon 70-200/2.8 VR II, which has a serious problem @ 200mm @ MFD). But in an IF macro, at 1:1 your focal length really is more like 65mm.

Now, none of this is really a suitable explanation for why FW 2.0.4 exhibits this "iris jitter" problem. The reason why is simple:

1. I could not reproduce the issue with the 100/2.8L macro IS, which of course is an IF macro lens.
2. I was able to cause the issue on the 300/4L IS, which is not a macro lens.
3. The change in effective f-number only relates to the actual light-gathering ability, not the size of the entrance pupil.
4. The change in entrance pupil diameter as a function of focus distance for an IF lens is an intrinsic optical property of the lens' construction--you can see it even if the lens is not mounted to the body. It has no relationship to the rapid fluctuation of the iris caused by the video mode.

Apr 06, 2010 at 04:41 AM

Mike V
Offline
• • •
Upload & Sell: Off

p.1 #5 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

My 100L does it, my 24L and 135L don't do it.

I think it is intentional.
I think it is auto exposure compensation.

Apr 06, 2010 at 05:36 AM

soniclove
Offline
•
Upload & Sell: Off

p.1 #6 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

it seems really random, like a twitch...doesnt seem intentional to me...I could get my 5DII to do it with my 100mm...not normal.

Apr 06, 2010 at 09:13 AM

dvarnav
Offline
• •
Upload & Sell: Off

p.1 #7 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

Its a stable problem for the 100 2.8 Macro lens and I reproduced the problem easily by using Manual mode 3.2F and by playing with focus the jitter is apearing. I tried the same with my 24-70 and I had absoloutely no jitter. I will try to reproduced it also with 50 1.8 and 100-300 and 70-200 2.8L and I will report

Apr 06, 2010 at 05:44 PM

dvarnav
Offline
• •
Upload & Sell: Off

p.1 #8 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

Just to report that all my lens was not reproducing the aperture jitter phenomenon except from 100 2.8Macro
50 1.8 is OK
24-70 2.8 is OK
70-200 2.8 is OK
28-105 is OK
100-300 is OK
100 2.8Macro is jitter affected (this lens is not the L series)

Apr 06, 2010 at 05:56 PM

wickerprints
Offline
• • • •
Upload & Sell: Off

p.1 #9 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

This is getting stupid:

http://www.canonrumors.com/2010/04/5d-mark-ii-firmware-2-0-4-update/

The explanation by "Dave" is incorrect, as I already explained in my earlier post. Anyone who knows anything about lenses will tell you that the effective aperture has nothing to do with the diaphragm stopping down. The reason why the entrance pupil appears to get smaller as you focus toward 1:1 is because these particular macro lenses are INTERNAL FOCUS. Furthermore, the jitter that occurs is not smooth, but intermittent.

The jitter also occurs on my EF 300/4L IS, which is not a macro lens. It happens even when it's not near MFD.

Confusion is what happens when you have people who make conclusions about things they don't really understand. When I read the comments, I even saw people talking about the EXIT pupil, which is completely irrelevant. The relationship of the exit pupil to the entrance pupil tells you the pupil factor, which in turn tells you about the symmetry of the lens (i.e. retrofocus vs. telephoto).

Apr 06, 2010 at 06:27 PM

kodakeos
Offline
• • •
Upload & Sell: On

p.1 #10 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

Wow, someone over at CR needs an optics lesson! Wickerprints is on the right track about Eff. aperture vs actual aperture. Lens elements get farther away from the image, the less energy they have.
MFD always looses a little bit of aperture compared to inf, but usually its nominal, and we dont notice - Ask any LF shooter..
But a jittery iris != effective aperture, or it would happen on EVERY LENS.

Apr 06, 2010 at 06:59 PM

wickerprints
Offline
• • • •
Upload & Sell: Off

p.1 #11 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

I just posted a video showing the problem:

http://www.youtube.com/watch?v=APgK5gdM69M

Apr 06, 2010 at 07:42 PM

wickerprints
Offline
• • • •
Upload & Sell: Off

p.1 #12 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

At the request of the group that initially discovered this issue, here is the link to the blog describing the problem in greater detail:

http://5dmark2.wordpress.com/2010/04/05/malfunction-in-firmware-2-0-4/

Here you can see ample evidence of the problem. The bottom line--in full Manual mode, the camera shouldn't be moving the iris *at all* unless the user is changing it.

Apr 06, 2010 at 08:31 PM

dvarnav
Offline
• •
Upload & Sell: Off

p.1 #13 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

wickerprints wrote:
This is getting stupid:

http://www.canonrumors.com/2010/04/5d-mark-ii-firmware-2-0-4-update/

The explanation by "Dave" is incorrect, as I already explained in my earlier post. Anyone who knows anything about lenses will tell you that the effective aperture has nothing to do with the diaphragm stopping down. The reason why the entrance pupil appears to get smaller as you focus toward 1:1 is because these particular macro lenses are INTERNAL FOCUS. Furthermore, the jitter that occurs is not smooth, but intermittent.

The jitter also occurs on my EF 300/4L IS, which is not a macro lens. It happens even when it's not near MFD.

Confusion is what happens when you have people who make conclusions about things they don't really understand. When I read the comments, I even saw people talking about the EXIT pupil, which is completely irrelevant. The relationship of the exit pupil to the entrance pupil tells you the pupil factor, which in turn tells you about the symmetry of the lens (i.e. retrofocus vs. telephoto).

What do you mean about things we do not understand ? Its a really funny discussion about the aperture. The problem is just the diaphragm fuzziness that 100 2.8 Macro Lens makes in Manual mode and nothing else. Any other discusion is a philology to how we see in true life this problem. The bug is always reproduced and we have nothing but a clear bug that occurs with these conditions

Apr 07, 2010 at 08:00 AM

chrisdee
Offline
• •
Upload & Sell: Off

p.1 #14 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

So let me get this straight. You have a situation where all these factors combined:

1) you are using the 100mm nonL macro or the 300mm f/4
2) you are using a smaller aperture than max the lens is capable of
3) the camera is in M mode
4) you are using the video capture
5) for some reason you keep rotating the focus ring back and forth while capturing

will cause what you call a "twitching" of the aperture blades between wide open and your set aperture which ruins your movie. Right?

Man, you guys have lots of time on your hands, I really envy others sometimes

Apr 07, 2010 at 08:18 AM

wickerprints
Offline
• • • •
Upload & Sell: Off

p.1 #15 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

dvarnav wrote:
What do you mean about things we do not understand ? Its a really funny discussion about the aperture. The problem is just the diaphragm fuzziness that 100 2.8 Macro Lens makes in Manual mode and nothing else. Any other discusion is a philology to how we see in true life this problem. The bug is always reproduced and we have nothing but a clear bug that occurs with these conditions

My comment was not directed at you. It was directed at the people commenting on the blog entry at Canon Rumors--specifically, the ones who were claiming that this "iris jitter" problem was simply a normal consequence of the effective aperture reducing light at macro distances.

That was proved to be so obviously incorrect as to be absurd. This phenomenon has nothing to do with effective aperture or macro lenses. My testing with the 300/4L IS showed as much.

The worst part about this topic is that there is a group of people who responded in a very hostile manner to the original reporter of this issue. They basically mobbed him, accused him of spreading fear and paranoia, and then when I posted my evidence supporting him, I was tarred with the same brush. Someone even accused me of being a sock puppet / alias.

It's really quite incredible, considering that I have barely used video recording on my 5D2--I am far more comfortable with stills!--and so the iris jitter isn't such a huge deal for me. But I conducted my own testing and made the YouTube video because I saw that someone was being unjustly criticized for trying to bring a legitimate issue to the attention of 5D2 users. And I felt the need to set the record straight.

I am still awaiting a credible and knowledgeable explanation for what is happening in the video I posted. No answer from the mob yet.

For what it's worth, I am a VERY SATISFIED 5D2 user. I freaking LOVE my camera. Seriously. After a year of having it, I literally wonder how I ever managed to live without it.

Apr 07, 2010 at 08:33 AM

mbaumser
Offline
• • •
Upload & Sell: Off

p.1 #16 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

chrisdee wrote:
So let me get this straight. You have a situation where all these factors combined:

1) you are using the 100mm nonL macro or the 300mm f/4
2) you are using a smaller aperture than max the lens is capable of
3) the camera is in M mode
4) you are using the video capture
5) for some reason you keep rotating the focus ring back and forth while capturing

will cause what you call a "twitching" of the aperture blades between wide open and your set aperture which ruins your movie. Right?

Man, you guys have lots of time on your hands, I really envy others sometimes

Can you explain the point of this post?

I don't have a 5d2 nor do I shoot video but it makes perfect sense to me that someone might shoot a video in full manual, with the lens stopped down and change the focus while doing so.

I guess if Toyota accelerator pedals only stuck when making a right turn, while going up a hill and the radio tuned to 101.9 FM than it's really not a problem...right?

I doubt this issue was found by someone testing their gear in every possible manner with nothing better to do. It was probably fist found in a real world situation by someone using their gear in exactly the conditions described and saw a problem.

Marc

Apr 07, 2010 at 12:56 PM

Pablov
Offline
•
Upload & Sell: Off

p.1 #17 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

It was probably fist found in a real world situation by someone using their gear in exactly the conditions described and saw a problem.

Marc

This is exactly how it happened (making it hard or even impossible to use the lens when filming for serious purposes with the new firmware). Then the information was shared with other 5D2 users who work on video, and confirmed it was not an isolated case.

Then a member of 5D Mark II Team made a detailed report including two videos, one showing the iris jitter and another showing what can be seen on the camera's LCD (or footage) when that happens.

wickerprints has helped a lot to show it is not limited to "macro" lenses. Not to mention that he also helped to make some people understand what it was all about, especially at CR.

As one pro user said, once you hit record, and your aperture is fixed to any value, the aperture should NOT be changing in full manual mode.

The result of the "iris jitter" when filming can be seen on one of the videos posted in the original report.

Apr 07, 2010 at 03:16 PM

ippikiokami
Offline
• •
Upload & Sell: Off

p.1 #18 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

Not that I use the 100mm macro or a 300 f/4 for video at all but that's really strange. Is there something about these two lenses that would make it only happen on them?

Apr 07, 2010 at 03:38 PM

scalesusa
Online
• • • •
Upload & Sell: Off

p.1 #19 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

If only one of your lenses does this, I would suspect a lens issue.

Apr 07, 2010 at 03:39 PM

dvarnav
Offline
• •
Upload & Sell: Off

p.1 #20 · 5D2 FW 2.0.4 - Iris jitter in M movie mode

Its firmware bug and it happens in any 100 Macro L or not L and also 300L f4 ... end of story WE WAIT CANON TO ANSWER. For any other that have doubts about that its so easy reproducable bug that noone could miss to make it again.

Apr 07, 2010 at 06:16 PM

1 2 3 end
FM Forums \| Canon-mount SLRs \| Join Upload & Sell 1 2 3 end