5DSr + Otus 85 NOT SHARP, Live View Problem - NOT HAPPY WITH CANON!

[charleshansen]

Let me 1st preface with saying about 10 posts down from this one, was my original from last week. Many people contributed help, for which I'm grateful. I still have the problem. I also want to say the 5DSr is new - less than 300 exposures on it. I have many Canon lenses, with many being "L". I am a member of Canon Pro Services, CPS. Using the 5DSr + the Zeiss OTUS 85mm, there's $9000 invested, with no satisfaction. Let me also add that for the 1st time since I'm 18, and that's over 40 yrs ago, I have ventured outside of the Canon system and bought a Fujifilm X-T2 + their 56mm 1.2, and I am in all ways astounded at what else apparently is out there. Quality, performance, functionality - superb. The best money and satisfaction right out of the box (at 1/3 the money of the Canon + Otus) I have ever spent in photography. And the sharpest images I have ever taken. Enough of that, because I think I'm finished with the Canon ride. 

To recap the problem - when using Live View for any lens attached, Canon or Zeiss, the lens automatically stops-down to say 5 or 5.6 or thereabouts. It can be stopped. When the picture is snapped, the lens opens up to whatever I set, say 1.4, makes the exposure, and the  returns to a stopped-down position - cannot be averted in ANY way. Exposure simulation enable, disable, or "during" makes no difference. What brought my attention to this was trying to discover why images viewed at 16X during live-liew, looking razor, razor sharp on the LCD using "the finest lens in photography", the Otus 85mm, result in soft, out-of-focus throwaways every time, regardless of lens. I made the posting last week addressing this problem.

People contributed lots of input - all to no avail. I called CPS phone tech support, and ABSOLUTELY no help there. I attended the Photo Expo show yesterday in NYC, spoke to the Canon techs at their booth + the techs they had there at the CPS booth - and no help there. Nobody's even really sure what the functionality is supposed to be. The only real help, or at least corroboration was had at the Zeiss booth, where a rep mounted the Otus 85 on a 5DSr, in Live-View, and it performed as it should - lens stayed wide open, until the exposure was made. So, my original suspician that there was something wrong with the camera seems true. Luckily is well within warranty, so Canon will be getting it back.

The reason I'm so annoyed at this is because I believe Canon has become the Microsoft of the photo world. They find it hard to support what they're putting out. They'd rather put customers through the ringer, than admit a problem. I did other searches..for the last 4 yrs or so, many people have had issues with Canon DSLR live-view. Now whether it's a defect 100% of the time, is probably not true either, but apparently either the systems are not designed/implemented well, or they can't be explained well enough to operate. People having real frustrations after spending good money with live-view systems either improperly designed or too flaky to operate without lots of heartache. No one on this board will accuse Canon of being the industry's most innovative company, but Canon will have to accept the fact they're the largest and are getting sloppy and complacent. I dread sending this Camera back, as I've read other such horror stories. They don't listen to their customers. And since the 5DS series has been released - NO firmware updates either.

If you're happy in all ways in the Canon eco-system, God bless you - but I believe it's the end of the road for me with Canon.

(And I won't even talk about waiting 4 years for the Mark III's successor, only to see a half-hearted attempt at what should have been a winner)

[RobertTheFat]

---

@kvbarkley wrote:

I would guess it stops down so that the highlights are not blown out on the sensor at the minimum video shutter speed.

---

We.re not talking video here, so I guess I don't see how that applies.

[kvbarkley]

---

@RobertTheFat wrote:

---

@kvbarkley wrote:

I would guess it stops down so that the highlights are not blown out on the sensor at the minimum video shutter speed.

---

We.re not talking video here, so I guess I don't see how that applies.

---

Live view is essentially video without saving a file.  Hence, all the people asking for "clean" HDMI output.

[jrhoffman75]

"When Live-View is used, the aperture is closing down to some value, and from what I found out, this is correct and by design. Thus is influencing your photographic method, whatever that is. My beef is that no one at Canon knows what that stop-down value is, or how it is arrived at. Simple question, right? Deserves a simple answer."

Maybe the simple answer is that there is no one answer - it depends on the light intensity and ISO set in the camera. The camera is just trying to ensure that the LCD is not blown out so the user can see the display.

Since it looks like the process is understood now, at least to some degree, what if you crank down the lighting intensity until you see that aperture is not closing down, manually focus for the sharpest image on the detail you want, and then restore brighter lighting for your shoot?

[RobertTheFat]

---

@charleshansen wrote:
When Live-View is used, the aperture is closing down to some value, and from what I found out, this is correct and by design. Thus is influencing your photographic method, whatever that is. My beef is that no one at Canon knows what that stop-down value is, or how it is arrived at. Simple question, right? Deserves a simple answer.

---

Deserves an answer? Yes. Simple? I dunno. Somewhere there must be one or two guys who wrote that part of the firmware, and they probably know the full story. But finding them may not be easy. (Well, it isn't, or you would have.) Canon may even have contracted it out.

Is the issue unique to the 5Dsr? Conceivably, but probably not. In principle, the lack of an IR filter could put its sensor at greater risk of overheating. But common sense suggests that the intensity of the IR radiation to which only the 5Dsr would be subjected is pretty low. So if the issue spans, say, all recent members of the 5D series, or even the 1DX as well, who (outside of Canon) might know the answer?

Maybe the developers of Magic Lantern. They've obviously deconstructed the firmware of many Canon cameras and should have encountered the stop-down code. There must be at least one Magic Lantern forum with participants who know what they're doing. Why don't you lurk over there for a while and see if you can find someone knowledgeable enough to point you in the right direction? If you get really lucky, there may even be a version of ML that works on the 5Dsr, and it might even have a way of defeating the stop-down.

I don't think I'd have the nerve to use ML on an expensive camera, especially to defeat a "feature" I didn't fully understand. But then I don't have your problem; and if it's causing you as much grief as you say it is, then the reward may be worth the risk.

[charleshansen]

I don't think so...
Look, a professional photographer needs to be able to control every aspect of his instrument, or at least understand every aspect. And the manufacturer should be able to explain why their product functions as it does. If as people have speculated here there is a built-in “defense mechanism” that gets deployed when Live-View is activated, in high ambient lighting conditions, the Canon should be able to put forward a statement something like the following:
“Using the Live-View function under very bright lighting conditions and with very fast lenses, the sensor may be damaged”. To avert this, the lens will automatically stop-down to protect the sensor, despite how the camera is set and which is beyond your control”. This I could see, and go along with. However here’s my rebuttal: If this really is the case, then why does the firmware allow the lens to be FULLY OPENED TO MAX APERTURE when the DOF button is depressed? Why doesn’t the sensor get damaged then? The mirror is still up; the shutter is still open, the sensor is still bathed in light. The image on the LCD screen is still being presented through an f-stop NOT OF YOUR CHOOSING. This throws off your final DOF and critical focusing ability. If there is a defense mechanism somehow being deployed, then why is there no time-out on Live-View. None of it makes sense.
This is why I don’t buy the defense mechanism scenario.
What I describe is exactly what happens, and the LCD screen automatically compensates (assuming you have the LCD brightness control on AUTO). The other control which could remotely conceivably influence what we’re observing is the “Exposure Simulation”, but even this does not influence the diaphragm from having a mind of its own when Live-View is used…
I just don’t understand how Canon has designed this entire “Live-View” system – and so far no one can explain it either. This relates to the 5DSr I am using, but apparently others have corroborated it with other Canon EOS bodies/lenses as well.
Some have gotten annoyed with me for pursuing my observations – sorry, I won’t march in lock-step and pretend this all makes sense.
Bottom line – from everything I’ve observed, the functionality could have been preserved, as with all the cameras that have gone before it – film or digital - mount the lens, set the exposure settings, activate “Live-View” and observe all your settings, focus and aperture. Press the DOF preview button to do just that – “get a preview of how the image is going to be affected during exposure with a different f-stop”. From what I see, their system could have been designed to support this traditional photographic flow in a digital world. It even had a name – “open aperture metering”.

[charleshansen]

My post above was in response to KVBARKLEY.
Robertthefat - I would be scared to flash the 5DSr's firmware with ML code. But I appreciate your thought process, and the points you bring up.
I'm able to work around what Canon designed - I just don't like it

[jrhoffman75]

"And there is also a “fly in the ointment” when it comes to the use of the Live View function, namely the fact that many current camera models (e.g. the Nikon D700 and the Canon EOS 5D MkII) automatically control the aperture in the Live View mode based on the set ISO values and the ambient light in order to ensure that the image in the display is consistently shown with the correct brightness and contrast values. This means that it is no longer possible to close the aperture to the desired, fixed setting using the preview button. Correct evaluation of the depth of field and focus position is therefore equally out of reach using this method, since the aperture is only closed to the desired setting when the shot is actually taken"

Extracted from this article:

http://www.zeiss.com/camera-lenses/en_us/website/photography/what_makes_the_difference/manual_focusi...

[charleshansen]

wow - all great points. Gets even more complicated.
Now I'm wondering if there's any tie-in as to why no firmware update to the 5DS series yet..they may not know how to address with any backwards compatibilty across the product line's architecture

[charleshansen]

jrhoffman75 - thanks very much for the reference to that superb Zeiss article; they understand the whole deal, and can effectively communicate it. Canon should take some pointers

[charleshansen]

yes, this too, is a possibility. (but I believe the sensor is not as sensitive as thought).
When getting critical focus, it's really helpful to throw a lot of light on the subject.
On my Canon 1.4 50mm lens, when I press live view + DOF button, lens opens up to 1.4 and stays open as long as I want - no blowouts, no timeout, just about any ISO. And at this setting, camera computes a 1/4000 speed. So I really don't think sensor sensitivity is at play. Sure would be nice if Canon could do some commenting. But what you say about the lighting would work.