A real-life worthwhile use of low ISO shadow lifting

Your example is not indicative of the problem and I suggest you give it a rest as you apparently don't understand the issue or the need. It's like someone who uses a car strictly for transportation not understanding why someone else might want one with 350 HP.

Bob

--
http://www.pbase.com/rwbaron
Thanks for that contribution, Bob. Helpful, as always. At least I gave a real-life example of a practical use of shadow lifting. Of course it's not indicative of any "problem". That was part of the point. And you'll notice I have "given a rest", as you put it, to my requests for examples. Since none have ever been forthcoming, I realize that's it's pointless to ask anymore. That's why I focused on something practical, like showing what you actually can do under difficult conditions with the least forgiving output (OOC JPEGs) of an appallingly bad sensor. I am sure that there are plenty of "photographers" who simply squirm with delight when they pull their shadows 5 or 6 stops with little noise and no banding. I'm sure that, to them, it's just as much of a thrill as gunning their 350 HP car along a country road, kicking up the dust, and congratulating themselves on how much power they have under the hood. Whatever gets you off. It's fine by me.
Someday in the not too distant future my hunch is you'll think back on this post and realize how myopic your view was.

Discussing this with you is moot because you don't want to learn,
Incredible that you would say that, after I've repeatedly asked for examples to teach me, and never been given any. The condescension that drips from your posts and from those of other members of the club is palpable. I know what shadow lifting looks like. I know what extreme shadow lifting looks like, both with and without noise and/or banding. What I don't know is what an actual worthwhile shot that requires far more shadow lifting than I can currently achieve looks like. That's what I used to ask for. That's what I was never given.
Apparently you missed the dozens or hundreds of thread on this.

OK Alastair now you can prove my point.

http://www.dpreview.com/forums/post/53983635

--
http://www.pbase.com/rwbaron
I vaguely remember that post, but it didn't seem like you tried to process the 7D file for optimal results, but I'm certainly willing to try myself. Where is the RAW?

--
As the length of a thread approaches 150, the probability that someone will make the obvious "it's not the camera, it's the photographer" remark approaches 1.
Alastair
http://anorcross.smugmug.com
Equipment in profile
 
Last edited:
I was just posting to show what could be done under difficult circumstances, even with OOC JPEGs from the old 18MP sensor. This seems to have made a lot of people defensive, to the point of being quite insulting about the fairly modest results. Everyone is entitled to their own opinion, of course, and I'm sure they will continue with their condescending "you couldn't possibly understand, until you get inducted into the club" replies. That's fine. My post was actually intended for those who are currently shooting with Canon, and perhaps wondering whether to go for the 7DII eventually. My point was, if you can get these results with bad conditions and the worst settings for DR with this old camera, imagine what you'll be able to get with the new camera when shooting RAW. I actually think this result is fine. As does everybody I have shown it to outside of this forum. These are people who look at pictures and see images, not test results of so much DR and such and such noise levels. They are actually the kind of people I take pictures for.

--
As the length of a thread approaches 150, the probability that someone will make the obvious "it's not the camera, it's the photographer" remark approaches 1.
Alastair
http://anorcross.smugmug.com
Equipment in profile
Yea ! What the heck Alastair !?!? Lifting shadows ! Ha !

Next thing we know, you will shooting with some ridiculous technique, like HDR ;) LOL

Geeez ! You oughta' know real photographers would never consider such a thing ! :) LOL

Fish
 
Apparently you missed the dozens or hundreds of thread on this.

OK Alastair now you can prove my point.

http://www.dpreview.com/forums/post/53983635
Looks to me like he had insufficient color NR. This is typical and is in fact the reason for the big difference seen in Fred Miranda and Tony Northup's examples. (They turn all NR off!) I only see this in ACR if I turn all NR off and push shadows.
That post was some time ago but to my recollection NR at LR default for both images.

How would you propose removing the pattern noise in the 7D file?
Did he post RAWs any where in the thread for others to process? Or at least post the pre-processed histograms and his settings?
The intent was to simply show an example where the high read and pattern noise of the 7D at ISO 100 presents a problem with processing a 7D RAW file compared to a camera with lower read noise and no pattern noise.

Do you think I'm massaging the files to the 7D's disadvantage?

Bob
 
Your example is not indicative of the problem and I suggest you give it a rest as you apparently don't understand the issue or the need. It's like someone who uses a car strictly for transportation not understanding why someone else might want one with 350 HP.

Bob

--
http://www.pbase.com/rwbaron
Thanks for that contribution, Bob. Helpful, as always. At least I gave a real-life example of a practical use of shadow lifting. Of course it's not indicative of any "problem". That was part of the point. And you'll notice I have "given a rest", as you put it, to my requests for examples. Since none have ever been forthcoming, I realize that's it's pointless to ask anymore. That's why I focused on something practical, like showing what you actually can do under difficult conditions with the least forgiving output (OOC JPEGs) of an appallingly bad sensor. I am sure that there are plenty of "photographers" who simply squirm with delight when they pull their shadows 5 or 6 stops with little noise and no banding. I'm sure that, to them, it's just as much of a thrill as gunning their 350 HP car along a country road, kicking up the dust, and congratulating themselves on how much power they have under the hood. Whatever gets you off. It's fine by me.
Someday in the not too distant future my hunch is you'll think back on this post and realize how myopic your view was.

Discussing this with you is moot because you don't want to learn,
Incredible that you would say that, after I've repeatedly asked for examples to teach me, and never been given any. The condescension that drips from your posts and from those of other members of the club is palpable.
Don't forget their club is only good below ISO 400. At ISO 400 and above even DxOMark shows little to no difference in the dynamic range between Canon and other cameras.
Why do you view this as "their club"? Why is this an "us against them" for you? That IMO says a lot about you and some bizarre tribal brand loyalty issue. We're all Canon shooters (at least I am) who are affected by this issue and would like to see it change.
I know what shadow lifting looks like. I know what extreme shadow lifting looks like, both with and without noise and/or banding. What I don't know is what an actual worthwhile shot that requires far more shadow lifting than I can currently achieve looks like. That's what I used to ask for. That's what I was never given.
you only want to push your brand biased agenda. I've wasted my time with others like TT Martin on this issue and posted a valid example and it went no where and that's a big reason why you won't get any takers.
Your obsession with making this a personal contest between TTMartin and rwbaron, rather than Fuji vs Canon is the reason I ended up blocking you. And it seems to me you still are obsessed about me. Otherwise, why do you keep using my name in posts?
I have no obsession with anything. It was you who called me out in the original thread stating you could provide a better end result. You laid down the challenge to me (which you failed miserable at) but for some strange reason you think I'm obsessed with you? I think it's you and Norcross who have the obsession as you're the ones starting threads about it. Please show where I've ever started a thread on this issue.
As I have said before, start a thread where you share the RAW files with everyone, and I will be happy to post my results.
When I have access to the computer where the RAW files are and if I can find them I will post both in a new thread on a neutral forum. You, Taylor and Norcross will all be welcome to join in and take your best shot.

Bob

--
http://www.pbase.com/rwbaron
Thanks. I don't think of it as a shot. I'm certainly not trying to kill anything. But I will see what can be done with Lightroom, which is what I use for pretty much all my processing. Oh, and which forum?

--
As the length of a thread approaches 150, the probability that someone will make the obvious "it's not the camera, it's the photographer" remark approaches 1.
Alastair
http://anorcross.smugmug.com
Equipment in profile
 
Last edited:
That post was some time ago but to my recollection NR at LR default for both images.
Default is not sufficient. You have to play with the NR sliders. Sometimes I think people are scared of those things. You've got plenty of room before you destroy detail, especially on color NR.
How would you propose removing the pattern noise in the 7D file?
In ACR that kind of thing cleans up nicely with more NR. At +2 to +3 you might end up with a hint of it remaining at 100% view, but not in print.

Push much harder and it rears its ugly head again, at least with the 7D.

Also: there are some NR plugins that specifically tackle pattern noise.
Do you think I'm massaging the files to the 7D's disadvantage?
Not at all. But when you process for maximum IQ possible (as opposed to 'equal' to see how the hardware compares) I don't think the difference is all that great or relevant in general real world use, especially when looking at realistic sizes and not pixels.

If in 5 years Sony is shipping 20-bit ADCs with 19 stops of noise free DR and Canon is still at this level, THAT would warrant hand wringing on the part of Canon users.
 
Last edited:
That post was some time ago but to my recollection NR at LR default for both images.
Default is not sufficient. You have to play with the NR sliders. Sometimes I think people are scared of those things. You've got plenty of room before you destroy detail, especially on color NR.
I do not like the effects of LR NR so I use the Denoise plug-in which IMO is much better. I may be able to clean up some of the noise but also some detail in doing so. I use Denoise extensively with my 7D files and know what it can and can't do.
How would you propose removing the pattern noise in the 7D file?
In ACR that kind of thing cleans up nicely with more NR. At +2 to +3 you might end up with a hint of it remaining at 100% view, but not in print.
Not my experience and the Denoise debanding tool doesn't do an effective job either.
Push much harder and it rears its ugly head again, at least with the 7D.

Also: there are some NR plugins that specifically tackle pattern noise.
Maybe there's one better than Denoise.

Why do you want to go through all that effort when there are cameras that give you a clean file to begin with?
Do you think I'm massaging the files to the 7D's disadvantage?
Not at all. But when you process for maximum IQ possible (as opposed to 'equal' to see how the hardware compares) I don't think the difference is all that great or relevant in general real world use, especially when looking at realistic sizes and not pixels.
I shoot only RAW and only Canon up until recently when I added a Fuji XT1. With my 10D I used ACR/PS exclusively and then moved to DPP for my XTi, 40D and 50D for most of the conversions except for those beyond DPP's capabilities. More recently I've moved back to LR and PS so I think I know something about working with Canon CR2's. I also own 3 Canon A3+ printers and have done a lot of printing so I know how a processed file looks as a medium sized print.
If in 5 years Sony is shipping 20-bit ADCs with 19 stops of noise free DR and Canon is still at this level, THAT would warrant hand wringing on the part of Canon users.
The problem is real and unfortunately Canon is lagging. Canon is currently the only manufacturer of imaging sensors for DSLR's using off-chip ADC. Sony, Toshiba, Panasonic, Aptina, Samsung and others all have some design of on-chip ADC with the resultant lower read noise. People talk about Sony but Panasonic with their recent MFT's sensor has surpassed Sony and it appears Samsung may have the new APS C sensor to beat but time and the reviews will tell.

Canon designs excellent "sensors" as their silicon and pixel design is among the best. They'd most likely have the best overall sensor/ADC IQ with an updated on-chip design and that is why this is such a shame. In some ways they remind me of GM back in the 80's and 90's. The big market leader that relies on inertia through a loyal customer base but look how that turned out in the long run.

I have to ask what cameras have you personally shot with and processed the RAW files from to your taste? Have you ever worked with a camera that provides a few added stops of clean shadows? If so do you shoot in such a way that you need to recover deep shadows? I never did until buying the Fuji and it's not even that good compared to other current APS C cameras but I'm impressed with how much better it is than my 7D. Bill Claff's analysis rates the XT1 PDR at only 1 stop over the 7D at base ISO so I can only guess how good a 2 stop improvement with the D7100 or K5 would be.

I used to be on the same side of this argument as you and defending Canon against the "shadow lifters" as I used to refer to them. Having experience with another camera has opened my eyes to what can be achieved. My suggestion is don't knock it until you've tried it.

Bob

--
http://www.pbase.com/rwbaron
 
Last edited:
I do not like the effects of LR NR so I use the Denoise plug-in which IMO is much better. I may be able to clean up some of the noise but also some detail in doing so. I use Denoise extensively with my 7D files and know what it can and can't do.
Don't know what to tell you. Aren't ACR and LR noise reduction routines the same? Maybe not.
In ACR that kind of thing cleans up nicely with more NR. At +2 to +3 you might end up with a hint of it remaining at 100% view, but not in print.
Not my experience
So post RAWs and see what other people produce. Maybe there are some RAWs which clean up and some which do not. The example I posted to this thread cleaned up pretty darn well with no detail loss.
Why do you want to go through all that effort when there are cameras that give you a clean file to begin with?
What effort? I move the NR sliders in my RAW converter.
Not at all. But when you process for maximum IQ possible (as opposed to 'equal' to see how the hardware compares) I don't think the difference is all that great or relevant in general real world use, especially when looking at realistic sizes and not pixels.
I shoot only RAW and only Canon up until recently when I added a Fuji XT1. With my 10D I used ACR/PS exclusively and then moved to DPP for my XTi, 40D and 50D for most of the conversions except for those beyond DPP's capabilities. More recently I've moved back to LR and PS so I think I know something about working with Canon CR2's. I also own 3 Canon A3+ printers and have done a lot of printing so I know how a processed file looks as a medium sized print.
Nobody wants your resume. Do you have RAW files that you want other people to try their hand at? Post them. Do you have a a set of comparison RAW files that you believe illustrate some fantastic advantage of another camera? Post them.
The problem is real and unfortunately Canon is lagging.
The "problem" is that Canon DSLRs get "only" 12.5 - 13 stops of total DR (5D3 and 70D; 7D was about 11.5) with "only" 2.5 - 3 stops of shadow latitude.

In other words: their "problem" is that they are better then Kodak Portra but not quite as good as Sony Exmor :-)

Oh and this is at low ISO because at 400 and above it doesn't matter.
I used to be on the same side of this argument as you and defending Canon against the "shadow lifters" as I used to refer to them. Having experience with another camera has opened my eyes to what can be achieved. My suggestion is don't knock it until you've tried it.
Because the D800 and Sony A7's I've shot with didn't have this. Oh wait :-)
 
Gee thanks a lot for that.
 
FWIW, here's an example of a shot where I think I was helped by the higher dynamic range of the Exmor sensor in my RX1r. I was able to preserve the highlights in the sky while having no difficulty raising the shadow to show the detail in the ice.

Now, I'm sure some will say that a Canon could do this. And I'm sure it could to some extent. But if I'd shot this with my 5D3, I very much doubt that the shadow areas would be as clean as this.

BTW, this was shot in an ice cave, looking up through a hole in the top of the cave.



 
Good for You. As for myself, I shadow lift high and low ISO pic's all the time. I'm always pushing my histogram as far to the right without clipping as I can.
 
Last edited:
Here's another example where I believe I was helped by the DR of my RX1r's sensor. The scene was strongly backlit, making the lanterns almost glow. But the front of the building was in deep shadow. As you can see, the lifted shadows are very clean and reveal a lot of detail.

 
FWIW, here's an example of a shot where I think I was helped by the higher dynamic range of the Exmor sensor in my RX1r. I was able to preserve the highlights in the sky while having no difficulty raising the shadow to show the detail in the ice.

Now, I'm sure some will say that a Canon could do this. And I'm sure it could to some extent. But if I'd shot this with my 5D3, I very much doubt that the shadow areas would be as clean as this.

BTW, this was shot in an ice cave, looking up through a hole in the top of the cave.

That's really cool. How big was the cave?

--
 
aftab said:
bhollis said:
FWIW, here's an example of a shot where I think I was helped by the higher dynamic range of the Exmor sensor in my RX1r. I was able to preserve the highlights in the sky while having no difficulty raising the shadow to show the detail in the ice.

Now, I'm sure some will say that a Canon could do this. And I'm sure it could to some extent. But if I'd shot this with my 5D3, I very much doubt that the shadow areas would be as clean as this.

BTW, this was shot in an ice cave, looking up through a hole in the top of the cave.

That's really cool. How big was the cave?
Pretty big.









In the next to last shot, that guy is standing under the hole in the roof of the cave.
 
Last edited:
Thanks. Looks like an awesome place. Where is it?

He must be trying the same shot you did. :)
 
Thanks. Looks like an awesome place. Where is it?

He must be trying the same shot you did. :)
It was under the Mendenhall Glacier near Juneau, Alaska. Sadly, it recently collapsed.
 
It's like someone who uses a car strictly for transportation not understanding why someone else might want one with 350 HP.
I think another analogy is called for. To me it seems that all those that argue in favor of the extended dynamic range are the ones that bought the car with launch control and advocate it's use at every red light just because it guarantees a 0.1 seconds faster 0-100 mph time. To them it doesn't matter if the competition does better on the normal road, for them only that insane and in every day life irrelevant single feature outweighs all rational argument.
 

Keyboard shortcuts

Back
Top