Canon 5d Mark III Shadow recovery

[sach100]

Btw, pushing shadows in LR/CR imho isn't really a substitute for +ev (with the adjustment brush or graduated filters) because it looks often artificial to me if shadows go over +50 in PV2012.

Agreed. I was just trying to see how much information is in there. The recovered part does look 'plasticy'

I think you could lift a LOT more than that. Try boosting the exposure slider by at least one stop, if not two. I push around my 7D shots a lot, sometimes a +80 shadow lift on top of a +1, +1.33, or +1.5 exposure lift (and usually some negative black shift to maintain contrast) when I botch a shot, and experience zero banding.

Thanks for the tip!

[mirekti]

If you cross your eye on the two images, focus on the one 'created' in the middle you get a 3D HDR image!   

HAHA! it does! +1

My eyes hurt!!

[Cgdillan]

If you cross your eye on the two images, focus on the one 'created' in the middle you get a 3D HDR image!   

HAHA! it does! +1

My eyes hurt!!

Off topic but..
"5D MkIII or 6D??? | EF 35f1.4 | EF 50f1.4 | EF 70-200f2.8L MkII | TC 1.4x III | TC 2x III"
Unless all you do is video.. Definitely the 5D mkIII. My opinion...

[Marsu42]

"5D MkIII or 6D??? | EF 35f1.4 | EF 50f1.4 | EF 70-200f2.8L MkII | TC 1.4x III | TC 2x III"
Unless all you do is video.. Definitely the 5D mkIII. My opinion...

... and if you do video: definitely 5d3, too, unless it's proven Magic Lantern will also run on the 6d. And the 6d hasn't even got the swivel screen, doh.

[dswatson83]

This video compares the D600 & 7D for shadow recovery and they also show a 5dIII file. It's clear to me that there is banding and grain in both the Canon shots more than the Nikon but only when viewed at 100% or more and only when pushing it all the way.

Nikon D600 Review Part 4 - Dynamic Range Testing...WOW!	Small | Large

[Cgdillan]

"5D MkIII or 6D??? | EF 35f1.4 | EF 50f1.4 | EF 70-200f2.8L MkII | TC 1.4x III | TC 2x III"
Unless all you do is video.. Definitely the 5D mkIII. My opinion...

... and if you do video: definitely 5d3, too, unless it's proven Magic Lantern will also run on the 6d. And the 6d hasn't even got the swivel screen, doh.

I agree mostly. If all you do is video and don't use magic lantern I would probably say 6D just for the price. unless of course you also need the headphone jack and silent controls. But IQ and ergonomics for video other than that is about the same.at $1,300 cheaper.

If you have the money then 5D mkIII will be the one that will last you without finding yourself wanting more. The 6D has it's limitations that you could easily outgrow.

[Cgdillan]

This video compares the D600 & 7D for shadow recovery and they also show a 5dIII file. It's clear to me that there is banding and grain in both the Canon shots more than the Nikon but only when viewed at 100% or more and only when pushing it all the way.
Nikon D600 Review Part 4 - Dynamic Range Testing...WOW!

I disagreed with a huge portion of this video. I understand where he was coming from but still I disagree with his approach. and the 5D mmiii scene was so different than the sunset scene there was no way to actually compare the difference. Although yes a huge difference between the D600 and 7D. I wonder what the difference is between the 7D and D7000 in this area.

[jrista]

This video compares the D600 & 7D for shadow recovery and they also show a 5dIII file. It's clear to me that there is banding and grain in both the Canon shots more than the Nikon but only when viewed at 100% or more and only when pushing it all the way.
Nikon D600 Review Part 4 - Dynamic Range Testing...WOW!

I disagreed with a huge portion of this video. I understand where he was coming from but still I disagree with his approach. and the 5D mmiii scene was so different than the sunset scene there was no way to actually compare the difference. Although yes a huge difference between the D600 and 7D. I wonder what the difference is between the 7D and D7000 in this area.

I agree about the D600 vs. 7D comparison. He kept saying that there was "so much detail" in the D600 shadows, but not in the 7D shadows. As far as I could tell, the following is actually what occurred:

 1. The 7D photo was over-exposed, which blew out the highlights of the sky, and captured full detail in the shadows.
 2. The 7D shadows appeared to be fully recoverable, I did not see any LOSS of detail, even if it was a bit noisy detail. The 7D shadow detail in the 100% crop simply looked brighter and a bit noiser than the D600 100% crop, but I wouldn't have said there was LESS detail.
 3. The D600 shot was a tiny bit under-exposed (or possibly just "properly" exposed), which preserved the highlights of the sky better.
 4. The D600 DOES preserve additional detail in the shadows, so they were able to be recovered despite the lower exposure than the 7D.
 5. The 7D shadows looked dull and muddy because a 100% shadow recovery resulted in a higher shadow exposure overall than in the D600 recovery.
 6. The D600 appeared to have the same general DETAIL in the shadows, however since it was not over-exposed with a 100% shadow push, it appeared to have much richer color.
 7. The D600 was shot second, after the sun had sunk a little lower into the horizon, which would have reduced the dynamic range of the scene a bit. Combined with the D600's additional DR to start, the odds were overly stacked in favor of the D600.

I would offer that one of two things could be done with the 7D image. First, he could have exposed it properly. He just popped out the camera and snapped a shot, without any concern in-camera for preserving those highlights. That is NOT a fault of the camera, that was operator error. A proper 7D exposure would have resulted in much the same kind of sky and midtines as the D600. Probably would have REALLY lost some detail in the shadows if that was done, however with a bit of NR much of it could have been recovered. Wouldn't be as good as real DR, but it would have been better than a botched exposure.

Alternatively, the Blacks of the 7D could have been pulled down a bit, which would have had the effect of increasing shadow contrast, and enriched the shadow colors to look as good as the D600 from a color fidelity standpoint. Nothing can be done about the blown sky, however I would be willing to bet LR4 could do some wondrous things with it regardless, and produce a final image that looked a lot closer to what the D600 produced.

I really wish I could get my hands on a D600 to compare with my 7D. It wouldn't take much more than a LITTLE bit of extra care with the 7D to capture a similar scene with much better results. Again, nothing is an alternative for additional real-world sensor DR, but I think the video above was far too ad-hoc and careless to be a viable test. That blown sky was the result of crappy camera use by a hapless camera user.

[Cgdillan]

This video compares the D600 & 7D for shadow recovery and they also show a 5dIII file. It's clear to me that there is banding and grain in both the Canon shots more than the Nikon but only when viewed at 100% or more and only when pushing it all the way.
Nikon D600 Review Part 4 - Dynamic Range Testing...WOW!

I disagreed with a huge portion of this video. I understand where he was coming from but still I disagree with his approach. and the 5D mmiii scene was so different than the sunset scene there was no way to actually compare the difference. Although yes a huge difference between the D600 and 7D. I wonder what the difference is between the 7D and D7000 in this area.

I agree about the D600 vs. 7D comparison. He kept saying that there was "so much detail" in the D600 shadows, but not in the 7D shadows. As far as I could tell, the following is actually what occurred:

 1. The 7D photo was over-exposed, which blew out the highlights of the sky, and captured full detail in the shadows.
 2. The 7D shadows appeared to be fully recoverable, I did not see any LOSS of detail, even if it was a bit noisy detail. The 7D shadow detail in the 100% crop simply looked brighter and a bit noiser than the D600 100% crop, but I wouldn't have said there was LESS detail.
 3. The D600 shot was a tiny bit under-exposed (or possibly just "properly" exposed), which preserved the highlights of the sky better.
 4. The D600 DOES preserve additional detail in the shadows, so they were able to be recovered despite the lower exposure than the 7D.
 5. The 7D shadows looked dull and muddy because a 100% shadow recovery resulted in a higher shadow exposure overall than in the D600 recovery.
 6. The D600 appeared to have the same general DETAIL in the shadows, however since it was not over-exposed with a 100% shadow push, it appeared to have much richer color.
 7. The D600 was shot second, after the sun had sunk a little lower into the horizon, which would have reduced the dynamic range of the scene a bit. Combined with the D600's additional DR to start, the odds were overly stacked in favor of the D600.

I would offer that one of two things could be done with the 7D image. First, he could have exposed it properly. He just popped out the camera and snapped a shot, without any concern in-camera for preserving those highlights. That is NOT a fault of the camera, that was operator error. A proper 7D exposure would have resulted in much the same kind of sky and midtines as the D600. Probably would have REALLY lost some detail in the shadows if that was done, however with a bit of NR much of it could have been recovered. Wouldn't be as good as real DR, but it would have been better than a botched exposure.

Alternatively, the Blacks of the 7D could have been pulled down a bit, which would have had the effect of increasing shadow contrast, and enriched the shadow colors to look as good as the D600 from a color fidelity standpoint. Nothing can be done about the blown sky, however I would be willing to bet LR4 could do some wondrous things with it regardless, and produce a final image that looked a lot closer to what the D600 produced.

I really wish I could get my hands on a D600 to compare with my 7D. It wouldn't take much more than a LITTLE bit of extra care with the 7D to capture a similar scene with much better results. Again, nothing is an alternative for additional real-world sensor DR, but I think the video above was far too ad-hoc and careless to be a viable test. That blown sky was the result of crappy camera use by a hapless camera user.

I understand that the D600 is better IQ. I meant that I disagree with the fact that he even compared the two cameras. It's a comparison that makes no sense since the two cameras would be used for two totally difference purposes. Neither one replaces the other.

[sach100]

Off topic but..
"5D MkIII or 6D??? | EF 35f1.4 | EF 50f1.4 | EF 70-200f2.8L MkII | TC 1.4x III | TC 2x III"
Unless all you do is video.. Definitely the 5D mkIII. My opinion...

From my experience in using 5d3 - it won't disappoint you at all. The video tends to be mushy with respect to fine details (outdoor - on grass for example) as other tests have shown but i don't know if 6d will be any better.

Whether 5d3 is worth the difference in price versus still untested 6D?  well its not an easy call until you can test both the bodies.

[RLPhoto]

ETTR.

[TrumpetPower!]

If the 5DIII has insufficient dynamic range for whatever you're doing, whatever you're doing is extreme. Basically, you're either trying to salvage a severely underexposed image, or you're trying to turn deep shadows into midtones (or even highlights -- I've actually seen people try that).

Here's an example of light that's as bad as one might reasonably want to shoot in -- midafternoon Arizona direct desert sunlight from the bottom of a steep ravine half-and-half lit and shadowed. And there's no dynamic range problems. No clipped highlights, no noisy or blocked shadows; just the expected image quality awesomeness from Canon's ultimate jack-of-all-trades camera.

http://www.canonrumors.com/forum/index.php?topic=13771.msg249243#msg249243

It's not a particularly beautiful image, though...but that's in no small part because the light itself is ugly.

Cheers,

b&

[distant.star]

.
This looks like Salt Lake City.

This shot was taken while i was adjusting for correct exposure. The sky looked okay but the bottom half is obviously underexposed. I wanted to see how much i can push the shadows in post.  I increased the shadow recovery slider to +100 in camera raw keeping everything else unaltered. If i can do this in post then this is plenty of available DR in shadows for me.
 

[art_d]

If the 5DIII has insufficient dynamic range for whatever you're doing, whatever you're doing is extreme. Basically, you're either trying to salvage a severely underexposed image, or you're trying to turn deep shadows into midtones (or even highlights -- I've actually seen people try that).

Not true. There are circumstances where you cannot properly expose all parts of the image due dyanmic range variations in the shooting conditions. You can expose properly for one part or the other and either blow out the highlights or block up the shadows. The lesser of two evils is to expose so the highlights don't get blown and if you still need to you can lift the shadows...this works ok in a lot of circumstances, but not all. In particular,  scenarios where you have larger areas of smooth shadows, pattern noise can become a very real problem that will be visible even on moderate prints sizes.

[TrumpetPower!]

If the 5DIII has insufficient dynamic range for whatever you're doing, whatever you're doing is extreme. Basically, you're either trying to salvage a severely underexposed image, or you're trying to turn deep shadows into midtones (or even highlights -- I've actually seen people try that).

Not true. There are circumstances where you cannot properly expose all parts of the image due dyanmic range variations in the shooting conditions. You can expose properly for one part or the other and either blow out the highlights or block up the shadows. The lesser of two evils is to expose so the highlights don't get blown and if you still need to you can lift the shadows...this works ok in a lot of circumstances, but not all. In particular,  scenarios where you have larger areas of smooth shadows, pattern noise can become a very real problem that will be visible even on moderate prints sizes.

You know, I keep coming across these complaints, very much like the one you just made, but I've yet to experience anything remotely like this problem in my shooting. Perhaps you could post an example of a properly-exposed image shot with a 5DIII with excessive shadow noise? And, please, not Fred Miranda's infamous page two example. That shot is at least a stop or two underexposed, in harsh noonday Sun, and he's lifting deep Zone II shade almost to midtones. That's the textbook definition of "extreme." And of "bad light." And, too, "poor technique."

Cheers,

b&