6 stop push: 5DsR vs A7R vs A7RII

[ajfotofilmagem]

It might be that people don´t think to much about how much a stop or two is. So here is an example.

To me this is at the extreme end of what I do. I have done more extremes than this, but I believe it is a good example.

First the straight RAW-file, just levelled, then the finished image, then two crops from each. If this was a more important picture, I would have worked more on the noise reduction, but that is a bit beside the point here.

The finished picture is Exposure: +1.5, Highlights: -100, Shadows: +100, Whites: +55. Colour luminance (blue): -34 and some rudimentary noise reduction. There´s a Looong way to 6 stop.

This is a good example of the usefulness of lifting shadows with low ISO.

But again I ask:
Does anyone have images showing the superiority of EXMOR to lift shadows in images above ISO1600? :

 

[gdanmitchell]

It is actually a bit worse than that. It isn't that people are incapable of making good photographs in the technical sense. In fact, any of these current cameras from Sony, Nikon, Canon, Pentax, Fujifilm, Olympus, and many others can produce outstanding photographic quality — more than good enough for just about any professional use from inclusion in printed and web media, through advertising, and on to fine art prints.

The real problem — and while it isn't new it has certainly been amplified — is the unending "brand humping" that sees photography as being all about Who Has The Best Thing, with the consequent effect of "confirmation bias" being a perverse focus on proving that the other guy bought The World's Crappiest Thing.

The lengths people now go to "demonstrate" the "inferiority" of the thing they chose not to get become astounding. Does anyone imagine even for a moment that the "test images" we see bandied about remotely represent what anyone is going to see in their photographs, no matter which of these cameras they might choose to use? That isn't going to happen.

Given a choice, I welcome improvements in dynamic range. And noise. And color quality. And sensor resolution. And camera interface. And lenses. And speed of operation. And flexibility. And adaptability. And reliability. And...

Camera choice is not about going nuts about this or that test chart. It is not about this brand or that. (I know photographers doing beautiful, professional work using cameras ranging from iPhones and 4/thirds gear up to 8 x 10 LF and everything in between, and using essentially every brand that is currently available.)

When it comes to this thing called "photography," more and more people are getting so distracted by gear humping that they are losing focus on photography and instead going of in some bizarre and unrelated direction...

Dan

By the way, if you want to see a real world example of just how bad you can make a decent exposure look if that is your goal, please read my article: http://www.gdanmitchell.com/2015/08/04/photographic-myths-and-platitudes-that-noise-is-awful

Here is the "awful" image from a 5Ds R... but you have to read what it actually represents! ;-)

Check this article that is somewhat a response to those "tests" (if it should be considered tests....) Tim Parkin did

http://www.gdanmitchell.com/2015/08/04/photographic-myths-and-platitudes-that-noise-is-awful

Ah, you found my article. Thanks...

Great article. Puts things into perspective.

I am honestly dumbfounded by the number of people on this forum who – judging by their comments – seem incapable of producing a quality photograph using equipment that is currently available today. It seems as though the narrowing gap between the lowest-priced and highest-priced cameras has generated a whole group of people desperate to identify miniscule differences and ascribe to those differences a significance that far exceeds rationality.

 

[anthonyd]

..I'd be interested to know which have been lifted by 6 stops in his portfolio.

On a terrible work computer at present so cant link properly but both the below contain 6 stop lifts in places. Also, many of my night-time macro shots contain similar - those are all shot with a Canon so I am doing these lifts with both Sony and Canon.

https://www.flickr.com/photos/kristianbell/16946600384
https://www.flickr.com/photos/kristianbell/17184708029

Its one small bugbear I have against some of these "only need to push 6+ stops if you're drunk or useless" comments is that they seem to imply the only use for such a lift is to lighten the entire image - for me it usually means one small patch that for whatever reason I find really distracting.

You take nice pictures ... you must have a nice camera
All jokes aside, your pictures are very nice (and I bet most of them would be just nice even without the 6+ push).

 

[Eldar]

It is actually a bit worse than that. It isn't that people are incapable of making good photographs in the technical sense. In fact, any of these current cameras from Sony, Nikon, Canon, Pentax, Fujifilm, Olympus, and many others can produce outstanding photographic quality — more than good enough for just about any professional use from inclusion in printed and web media, through advertising, and on to fine art prints.

The real problem — and while it isn't new it has certainly been amplified — is the unending "brand humping" that sees photography as being all about Who Has The Best Thing, with the consequent effect of "confirmation bias" being a perverse focus on proving that the other guy bought The World's Crappiest Thing.

The lengths people now go to "demonstrate" the "inferiority" of the thing they chose not to get become astounding. Does anyone imagine even for a moment that the "test images" we see bandied about remotely represent what anyone is going to see in their photographs, no matter which of these cameras they might choose to use? That isn't going to happen.

Given a choice, I welcome improvements in dynamic range. And noise. And color quality. And sensor resolution. And camera interface. And lenses. And speed of operation. And flexibility. And adaptability. And reliability. And...

Camera choice is not about going nuts about this or that test chart. It is not about this brand or that. (I know photographers doing beautiful, professional work using cameras ranging from iPhones and 4/thirds gear up to 8 x 10 LF and everything in between, and using essentially every brand that is currently available.)

When it comes to this thing called "photography," more and more people are getting so distracted by gear humping that they are losing focus on photography and instead going of in some bizarre and unrelated direction...

Dan

Thanks for the article, interesting read

 

[timparkin]

Who is going to lift shadows 6 stops in real life. This had nothing to do with the image quality the 5dsr can deliver.

You'd think it's absurd, but on the Sony forums and Fred Miranda the comparisons are occurring at 7 stops. They're also arguing about the 12/14 bit raw voodoo that occurs in bulb or electronic shutter mode.

I don't think I've pushed anything in my library more than 3 stops. If I am missing by 7 stops, something crazy happened.

I'm actually impressed with the 5DsR's performance considering how many words were spilled on the Sony vs Canon DR debate.

If you've ever used grads you'll have come across situations where you need a combo of two grads. Possibly a three and a two stop. Then add another stop for opening the shadows in general and there's your six stops. I agree it's unlikely but I posted an example in the article (in the updates at the bottom) showing a view outside my back door just as people were asking. the clouds were peaking at EV16 and the shadows under the tree were EV 4. I agree it won't happen that often and in most situations you can use grads but the idea is that you can do without grads.

Want to see another couple of examples..

http://static.timparkin.co.uk/static/tmp/dynamic-range-2.jpg
http://static.timparkin.co.uk/static/tmp/portra-test-pp.jpg

These are actually examples of unfiltered Portra 400 which has nearly 19 stops of dynamic range (unclippable highlights!).

The shadows here would need over a six stop push to get them back.

e.g. if you wanted this..

http://static.timparkin.co.uk/static/tmp/portra-test.jpg

 

[Sporgon]

Who is going to lift shadows 6 stops in real life. This had nothing to do with the image quality the 5dsr can deliver.

You'd think it's absurd, but on the Sony forums and Fred Miranda the comparisons are occurring at 7 stops. They're also arguing about the 12/14 bit raw voodoo that occurs in bulb or electronic shutter mode.

I don't think I've pushed anything in my library more than 3 stops. If I am missing by 7 stops, something crazy happened.

I'm actually impressed with the 5DsR's performance considering how many words were spilled on the Sony vs Canon DR debate.

If you've ever used grads you'll have come across situations where you need a combo of two grads. Possibly a three and a two stop. Then add another stop for opening the shadows in general and there's your six stops. I agree it's unlikely but I posted an example in the article (in the updates at the bottom) showing a view outside my back door just as people were asking. the clouds were peaking at EV16 and the shadows under the tree were EV 4. I agree it won't happen that often and in most situations you can use grads but the idea is that you can do without grads.

Want to see another couple of examples..

http://static.timparkin.co.uk/static/tmp/dynamic-range-2.jpg
http://static.timparkin.co.uk/static/tmp/portra-test-pp.jpg

These are actually examples of unfiltered Portra 400 which has nearly 19 stops of dynamic range (unclippable highlights!).

The shadows here would need over a six stop push to get them back.

e.g. if you wanted this..

http://static.timparkin.co.uk/static/tmp/portra-test.jpg

Tim, welcome to CR, which is sometimes mistaken for DR. Get your drum scanner out and polished; I've some 6x7s for scanning.

 

[weixing]

Hi,

It might be that people don´t think to much about how much a stop or two is. So here is an example.

To me this is at the extreme end of what I do. I have done more extremes than this, but I believe it is a good example.

First the straight RAW-file, just levelled, then the finished image, then two crops from each. If this was a more important picture, I would have worked more on the noise reduction, but that is a bit beside the point here.

The finished picture is Exposure: +1.5, Highlights: -100, Shadows: +100, Whites: +55. Colour luminance (blue): -34 and some rudimentary noise reduction. There´s a Looong way to 6 stop.

The problem is the photo become like some CAD generated image...

Have a nice day.

 

[gdanmitchell]

It is actually a bit worse than that...

Dan

Thanks for the article, interesting read

Thank you for taking the time to read it! I just wrote and posted another follow-up article that is intended to bring a bit of perspective to the the flood of hyperbole. Take a look: http://www.gdanmitchell.com/2015/08/06/photographic-myths-and-platitudes-the-best-camera-part-1 — "Photographic Myths and Platitudes: The Best Camera! (Par)t 1)"

Dan

 

[timparkin]

Hi,

It might be that people don´t think to much about how much a stop or two is. So here is an example.

To me this is at the extreme end of what I do. I have done more extremes than this, but I believe it is a good example.

First the straight RAW-file, just levelled, then the finished image, then two crops from each. If this was a more important picture, I would have worked more on the noise reduction, but that is a bit beside the point here.

The finished picture is Exposure: +1.5, Highlights: -100, Shadows: +100, Whites: +55. Colour luminance (blue): -34 and some rudimentary noise reduction. There´s a Looong way to 6 stop.

The problem is the photo become like some CAD generated image...

Have a nice day.

Hi Sporg, at last someone who knows which way I'm biased (the proper 'full frame')

 

[romanr74]

Who is going to lift shadows 6 stops in real life.

I do, regularly.

OK, I'm intrigued. Could you post some examples?

You should check out Kristian's work, really striking images. I particularly like the (predominantly tan) desert shots, not that those require 6-stop shadow lifting.

I agree that Kris's work is impressive, but I'd be interested to know which have been lifted by 6 stops in his portfolio.

I agree too, but everybody here will have to agree that many of these pictures look very "unnatural".

 

[timparkin]

Sony A7Rii compared with Sony A7R and Canon 5DSr

Note that, in this comparison, the A7R and 5DSR were exposed for 1/2500s, while the A7R2, at the same aperture, was exposed for 1/1000s.

Well spotted, but Tim Parkin did say that they exposed as far to the right as possible in holding highlights, then under exposed from there by 6 stops. So, this suggests the a7RII has more highlight range. Interesting.

However Tim getting involved in under exposing by 6 stops is a real worry. He lives just up the road from me and I think I'll have to drop in and have a serious word with him.

Once we have to resort to 5 or 6 stops under to show differences in sensors it's game over really; they're all the same in practical use.

It would be great to have a beer to discuss. The exposure difference was ambient light changes I think. As for 6 stop push. I've used a 4 stop grad before AND opened the shadows - that's six stops! The idea with the Sony is the possibility of shooting without grads. I shall have to find a few 6 stop push examples

 

[3kramd5]

Sony A7Rii compared with Sony A7R and Canon 5DSr

Note that, in this comparison, the A7R and 5DSR were exposed for 1/2500s, while the A7R2, at the same aperture, was exposed for 1/1000s.

Well spotted, but Tim Parkin did say that they exposed as far to the right as possible in holding highlights, then under exposed from there by 6 stops. So, this suggests the a7RII has more highlight range. Interesting.

However Tim getting involved in under exposing by 6 stops is a real worry. He lives just up the road from me and I think I'll have to drop in and have a serious word with him.

Once we have to resort to 5 or 6 stops under to show differences in sensors it's game over really; they're all the same in practical use.

The exposure difference was ambient light changes I think.

So basically each shot was independent, and the range of tones within the scene were not constant (e.g. the shadows may have been deeper relative to similar highlights when the A7R was used versus the A7R2).

It's certainly interesting and probably correlates well with how an individual camera holds up to huge shadow pushes. I'm not sure I'd use it as a tool of comparison, however.

I hope the weather cooperates and I can play with mine tonight (it's in a box at my house). ;D

 

[timparkin]

Sony A7Rii compared with Sony A7R and Canon 5DSr

Note that, in this comparison, the A7R and 5DSR were exposed for 1/2500s, while the A7R2, at the same aperture, was exposed for 1/1000s.

Well spotted, but Tim Parkin did say that they exposed as far to the right as possible in holding highlights, then under exposed from there by 6 stops. So, this suggests the a7RII has more highlight range. Interesting.

However Tim getting involved in under exposing by 6 stops is a real worry. He lives just up the road from me and I think I'll have to drop in and have a serious word with him.

Once we have to resort to 5 or 6 stops under to show differences in sensors it's game over really; they're all the same in practical use.

The exposure difference was ambient light changes I think.

So basically each shot was independent, and the range of tones within the scene were not constant (e.g. the shadows may have been deeper relative to similar highlights when the A7R was used versus the A7R2).

It's certainly interesting and probably correlates well with how an individual camera holds up to huge shadow pushes. I'm not sure I'd use it as a tool of comparison, however.

I hope the weather cooperates and I can play with mine tonight (it's in a box at my house). ;D

Unlikely that the distribution of light changed as it was as flat as a pancake and you'd see those variations in the raw files. A minor contribution and way below the 6 stops we are working with.

 

[3kramd5]

Unlikely that the distribution of light changed as it was as flat as a pancake and you'd see those variations in the raw files. A minor contribution and way below the 6 stops we are working with.

Yes, it's quite clearly flat, that's why I was so incredulous about the differently required exposure times when there aren't any obvious clouds and the sun was still close to directly overhead (although maybe not as much as far north you are).

 

[timparkin]

Unlikely that the distribution of light changed as it was as flat as a pancake and you'd see those variations in the raw files. A minor contribution and way below the 6 stops we are working with.

Yes, it's quite clearly flat, that's why I was so incredulous about the differently required exposure times when there aren't any obvious clouds and the sun was still close to directly overhead (although maybe not as much as far north you are).

All I can think is that there were very high level clouds above the lower clouds we could see. I suppose one of those going over the sun would moderate the flat light. I know from working large format cameras that light levels can change even though you don't notice it because it's so flat and your eyes adjust

 

[3kramd5]

Unlikely that the distribution of light changed as it was as flat as a pancake and you'd see those variations in the raw files. A minor contribution and way below the 6 stops we are working with.

Yes, it's quite clearly flat, that's why I was so incredulous about the differently required exposure times when there aren't any obvious clouds and the sun was still close to directly overhead (although maybe not as much as far north you are).

All I can think is that there were very high level clouds above the lower clouds we could see. I suppose one of those going over the sun would moderate the flat light. I know from working large format cameras that light levels can change even though you don't notice it because it's so flat and your eyes adjust

Absolutely. It didn't look cloudy at all (quite harsh), but that's a definite possibility.

 

[benperrin]

https://www.flickr.com/photos/kristianbell/16946600384
https://www.flickr.com/photos/kristianbell/17184708029

Its one small bugbear I have against some of these "only need to push 6+ stops if you're drunk or useless" comments is that they seem to imply the only use for such a lift is to lighten the entire image - for me it usually means one small patch that for whatever reason I find really distracting.

Yes this is correct. Often a large push is confined to a small area of the image. I do have to ask though; why don't you just bracket exposures and use luminosity masks to lift the shadows in the offending area? I mean I know everyone has their own methodology and your images are certainly better than mine so I'm just asking. I like your post processing you really have some awesome images.

 

[timparkin]

Who is going to lift shadows 6 stops in real life.

I do, regularly.

OK, I'm intrigued. Could you post some examples?

You should check out Kristian's work, really striking images. I particularly like the (predominantly tan) desert shots, not that those require 6-stop shadow lifting.

I agree that Kris's work is impressive, but I'd be interested to know which have been lifted by 6 stops in his portfolio.

I agree too, but everybody here will have to agree that many of these pictures look very "unnatural".

Here's an example from my business partner Joe Cornish. First the before and after with histogram

http://static.timparkin.co.uk/onlandscape/SonyLakes/dynamic-range-example.jpg

and then the lightroom adjustments to get that far..

http://static.timparkin.co.uk/onlandscape/SonyLakes/lightroom-adjustments.jpg

Tim

 

[timparkin]

https://www.flickr.com/photos/kristianbell/16946600384
https://www.flickr.com/photos/kristianbell/17184708029

Its one small bugbear I have against some of these "only need to push 6+ stops if you're drunk or useless" comments is that they seem to imply the only use for such a lift is to lighten the entire image - for me it usually means one small patch that for whatever reason I find really distracting.

Yes this is correct. Often a large push is confined to a small area of the image. I do have to ask though; why don't you just bracket exposures and use luminosity masks to lift the shadows in the offending area? I mean I know everyone has their own methodology and your images are certainly better than mine so I'm just asking. I like your post processing you really have some awesome images.

The obvious answer is - why do that when you don't have to? It sounds like bracketing, blending and luminosity masks is just just making it harder for yourself if you have a camera that can do it in a single exposure.

 

[benperrin]

The obvious answer is - why do that when you don't have to? It sounds like bracketing, blending and luminosity masks is just just making it harder for yourself if you have a camera that can do it in a single exposure.

Because you lose quality by lifting the shadows by so much. If it's a tiny area it might not matter but it's something that would bug me.