Strange flare with 5dII. Please check.

Hi Philip.
Maybe my wording was confusing.

40D and 5DII (firmware 1.0.7) behave the same only on 5x magnification. On 5x the 5DII will go all the way to the edge.

With 10x it will not, it misses a few pixels...

Cheers
George

Philip Harle wrote:

Are you sure they are the same? On my 5DII, 5x live view doesn't go
anywhere near the edge of the frame, and 10x is even worse.

Hi Ramesh !
It's OK, the nice guys here more than make it up for the annoying ones !

It's all good fun, and as they say if you can't take the heat get out of the kitchen

sunhorse wrote:

George, thank you for letting us all know about this issue. You
showed a lot of patience in spite of some of the ridiculous fanboy
postings.

I will try this on my 5D II as well, since I bought it only a month ago.

I thought of having some fun with ray tracing modeling and this is a nice opportunity.
I though of sharing this 'cause I think some will find interesting.

The objective is this :

• How much of the image circle hits the 40D mirror back ?

• How much the 5DII back ?

and most important

• Why is the reflection so elongated ? For a point light source coming from an angle close to 90 degrees in relation to the sensor, it should be very close to circular (after been bounced off the mirror assembly)

I did a very basic modeling of the camera mirror boxes and also a test model of the mirror's back face.
The dimensions are as close as I could measure them.

For the image circle scenes:

-The light source is positioned at the registration distance of the EF mount(44mm).
-It more closely resembles shooting with very small aperture.

-The light projection was set to the minimum required to cover the FF size sensor.

For the reflection model, I set three lights shooting very narrow beams on three surfaces.

The first is a close approximation to the mirror's back housing with the distinctive horizontal grooves.

The second is a flat surface. This should definitely produce an almost-circular reflection.

The third is the same mirror back but rotated 90 degrees so the grooves are now vertical.

The extra two surfaces serve as a test/verification for the ray tracing accuracy.

Here are the results :

Please note that the bottom and right sides of the mirror box are not included so we can see "inside" easier.

Image circle and the two sensors combined for reference:

40D mirror box:

5DII mirror box:

Observe how the 40D has a lot more mirror area covered by the incoming light. This puts 40D with its smaller sensor and mirror box in a worse situation than the 5D.

And this is the "why this shape" test.

• The modeling is capable of resolving reflection, refraction, diffraction and caustics.

Recognize the shape ?

The way I understand it (it's probably a nice educative discussion) is that reflections from the curved surfaces of the tip of each groove are all mixed and projected on the sensor.
I'm not sure if diffraction plays a major role here.
I could play with this model and deduce many more info, but time is running out.

Observe how the flat surface produces a circular pattern.

The vertical grooves produce a totally different pattern. The non uniformity is because the grooves are not symmetrical.
It takes the horizontal grooves to get what we see in reality.

This may not be the ultimate proof, but (at least for me) very strong evidence as to the mirror causing this issue.

I hope it was an interesting reading

Cheers
George

Thanks, that is good news - I am using firmware 1.0.6 so hopefully it will change when I upgrade.
Philip

elgeorgio wrote:
Hi Philip.
Maybe my wording was confusing.
40D and 5DII (firmware 1.0.7) behave the same only on 5x
magnification. On 5x the 5DII will go all the way to the edge.

With 10x it will not, it misses a few pixels...

Cheers
George

Philip Harle wrote:

Are you sure they are the same? On my 5DII, 5x live view doesn't go
anywhere near the edge of the frame, and 10x is even worse.

Gary, you have posted a lot to this thread but clearly have not read much of it. The OP said early on he had tried live view. As the light is reflecting off the back of the mirror it clearly makes no difference.

GaryJP wrote:

A) Mirror lock
B) Live View

Better? Worse?
--
http://garyp.zenfolio.com/p518883873/

GaryJP wrote:

xLnT wrote:

No hood can solve it.

Not true. If the light sources cannot hit the lens, the mirror cannot
pick them up.

If the light sources aren't hitting the lens then the lens cannot image!

These reflections from the mirror are only 1mm or so off of the edge of the focal plane - say 13mm off axis as opposed to the 12mm to the focal plane edge. So, for a 100mm, f/2.8 lens, these sources are only 0.6deg off of the edge of the frame. At f/2.8, the physical aperture of the lens is approximately 36mm at the lens stop, and the hood must block all light from these sources without vignetting the actual image. Thus the hood must be at least 36/tan(0.6) mm long - around 3.6m! And that assumes that the stop is actually at the front objective of the lens, not internal! With a 70mm front objective and an internal stop, you need a 7m lens hood to achieve your goal!

Sure, you can use a hood to prevent the problem, but no practical hood will block the light from these sources from reaching the lens and still permit light from the rest of the scene to be imaged without severe vignetting.
--
Its RKM

Redteg94 wrote:

If this is what I think it is, I've had it happen with my 40D; there
is light coming in through your viewfinder from behind.

It isn't what you think it is. Why would turning the camera upside down prevent the light coming into the viewfinder and thus solving the problem.

Too many people throwing ideas and solutions in before reading the symptoms.

This is reflection off the back of the mirror, and it occurs with many full frame SLRs - I have even experienced the effect with film SLR cameras.
--
Its RKM

GaryJP wrote:

I don't see a "problem" here, that is not incredibly easily resolvable.

As posted elsewhere in the thread, your "incredibly easily" solution is total BS.

Nobody carried 22ft lens hoods around with them and you have never seen a baffle that size, let alone used one!

Yes, 22ft is the length of the lens hood required to isolate those light sources from the lens without vignetting the image for a 100mm lens with a 70mm front aperture!

A lens hood is NOT the answer and, if you had bothered to read the thread, you would have realised that the OP was already using your "incredibly" simple solution. The OP is the very proof that your "solution" doesn't work. Hoods work for lights that are tens of degrees out of frame, and are completely useless for sources that are a degree or less out of frame!
--
Its RKM

GaryJP wrote:

These light sources are out of shot. As someone pointed out, when you
zoom out a little to bring them fully into shot, the flare disappears.

So it isn't a solution if you don't want the lights in the frame!
--
Its RKM

I just upgraded to firmware 1.0.7 and my 5DII still does not go to the edge of the frame at 5x magnification. Does the 40D?

elgeorgio wrote:
Hi Philip.
Maybe my wording was confusing.
40D and 5DII (firmware 1.0.7) behave the same only on 5x
magnification. On 5x the 5DII will go all the way to the edge.

With 10x it will not, it misses a few pixels...

Cheers
George

Philip Harle wrote:

Are you sure they are the same? On my 5DII, 5x live view doesn't go
anywhere near the edge of the frame, and 10x is even worse.

GaryJP wrote:

Its RKM wrote:

GaryJP wrote:

These light sources are out of shot. As someone pointed out, when you
zoom out a little to bring them fully into shot, the flare disappears.

So it isn't a solution if you don't want the lights in the frame!

Personally I can think of very few situations where it's crucial to
me to frame bright lights right off the bottom edge off the frame and
refuse to either zoom in a bit to deal with it or zoom out and crop.

Well one example is where there are a lot of such lights in the scene and including some means that others are just outside of the frame. Another is where the lights themselves are in motion and it is impossible to choose just the right moment to take the shot that would avoid these internal reflections. One parituclar example that includes both of these situations, and which has already been cited in this thread, is contra-jour lighting across open sea: the waves exist all over the place and the sun reflections off the crests are continuously changing.

Still I'm interested to see how this develops.

I don't expect to see any developments. This is a problem that has been around for ages. It isn't unique to the 5D-II. I have replicated it on the original 5D and my first encounter with it was in 1976 using an old Zenith E film camera. Most full frame SLRs, whether digital or film, demonstrate it to one degree or another.
--
Its RKM

GaryJP wrote:

Its RKM wrote:

GaryJP wrote:

I don't see a "problem" here, that is not incredibly easily resolvable.

As someone pointed out, zooming out a bit loses it. Unless you are
shooting video.

That only works if only one light is causing the problem - and since you can't see this effect in the viewfinder it is nigh on impossible to predict until after the event unless you use live view. After the problem occurs, you can get the exact shot you want if you turn the camera upside down - you don't need to zoom out to stop the light hitting the mirror: an even simpler solution than your suggestion, but equally absurd!

And it seems it doesn't occur in the hands of every shooter even then.

True, it only occurs to those that shoot images with bright lights below the frame that can bounce off the back of the mirror.
--
Its RKM