Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1      
2
       3       end
  

Archive 2012 · Re-assessing the D800 sensor as a Canon alternative
  
 
splathrop
Offline
• • •
Upload & Sell: Off
p.2 #1 · Re-assessing the D800 sensor as a Canon alternative


galenapass: A lot of words, but they don't really makes sense. As noted above the D800 has the highest score for color depth among full-frame SLRs. The information is there it's up to you what you want to do with it.


I must owe you an apology for my lack of clarity. Because you have not understood what I am trying to say.

The story so far. I say Nikon D800/D800e files look lacking in color/contrast, or something. Nobody argues with that, not yet anyway. Some people seem to agree. I suggest reasons why that problem might be inherent in the high dynamic range sensor, at least in part because of its high dynamic range. I also suggest reasons why those problems might be hard to deal with in post processing—reasons that might account for the fact that people posting images from those cameras do not seem to have fixed those problems—their images fairly uniformly seem to show those problems.

And then multiple commenters weigh in saying, in effect, "Don't worry, just fix it with post processing."





Oct 13, 2012 at 08:22 PM
snapsy
Online
• • • •
Upload & Sell: On
p.2 #2 · Re-assessing the D800 sensor as a Canon alternative


splathrop wrote:
The story so far. I say Nikon D800/D800e files look lacking in color/contrast, or something. Nobody argues with that, not yet anyway. Some people seem to agree. I suggest reasons why that problem might be inherent in the high dynamic range sensor, at least in part because of its high dynamic range. I also suggest reasons why those problems might be hard to deal with in post processing—reasons that might account for the fact that people posting images from those cameras do not seem to have fixed those problems—their images fairly uniformly seem to show those problems.

And then multiple
...Show more

Any perceived differences in contrast and color (especially contrast) is simply from differences in the processing defaults of the raw converter (or JPEG engine). If you've ever seen the source code of a demosaicing algorithm and its associated image processing pipeline you'd see there are countless decisions that need to be made about the default rendering of the image, and many times those decisions are either entirely arbitrary or just suited to the developer's taste. Any some of those default decisions result in different rendering between two different bodies.


Edited on Oct 13, 2012 at 09:04 PM · View previous versions



Oct 13, 2012 at 09:02 PM
AGeoJO
Offline
• • • • • •
Upload & Sell: On
p.2 #3 · Re-assessing the D800 sensor as a Canon alternative


splathrop wrote:
The story so far. I say Nikon D800/D800e files look lacking in color/contrast, or something. Nobody argues with that, not yet anyway. Some people seem to agree. I suggest reasons why that problem might be inherent in the high dynamic range sensor, at least in part because of its high dynamic range.


I have noticed that a lot of D800/e users want to show case the ability of their camera in rendering details in the shadow area and they went too much in lifting the shadow parts of their images. It is a great feature to have but a lot of them went overboard with that for my taste. On the other hand, a lot of Canon users did become gun shy in that respect for "fear" of showing banding or noise and what not. I do not have any issues in doing so. Frankly, I have not seen this as issue whatsoever in real life images in the 6 months or so of using my 5D Mark III. YMMV, of course.



Oct 13, 2012 at 09:04 PM
JameelH
Online
• • •
Upload & Sell: On
p.2 #4 · Re-assessing the D800 sensor as a Canon alternative


AGeoJO wrote:
I have noticed that a lot of D800/e users want to show case the ability of their camera in rendering details in the shadow area and they went too much in lifting the shadow parts of their images. It is a great feature to have but a lot of them went overboard with that for my taste. On the other hand, a lot of Canon users did become gun shy in that respect for "fear" of showing banding or noise and what not. I do not have any issues in doing so. Frankly, I have not seen this as issue
...Show more

Quite true. It is the notion of doing something because one can. However if one does pp for what the specific image they want to create, I would say either of the two systems would not be the limiting factor - most definitely in my case anyway.



Oct 13, 2012 at 09:13 PM
Lan11
Offline
• •
Upload & Sell: Off
p.2 #5 · Re-assessing the D800 sensor as a Canon alternative


I agree with galenapass.
A lack of basic skills and lot of words, mostly meaningless, is the problem.
What really counts is the DXO number : -)



Oct 13, 2012 at 09:54 PM
Lan11
Offline
• •
Upload & Sell: Off
p.2 #6 · Re-assessing the D800 sensor as a Canon alternative


Quote:
“Maybe because contrast boundaries are more fully occupied across the printable gamut—which could put at a premium the sensor's ability to deliver the color you prefer right out of the camera.”

Common, we were to get the perfect files right from the digital cameras. This was promised. So what happened?
We were also to have a paperless society with the advent of the computers among other things.
One has to distinguish reality from hype and BS, but only a few really can.
Beside your are talking to people who knows best. Most do not print but use the social me(r)dia. Since limited color recognition is common among the male population you’re literally trying to discuss colors with the blind.



Oct 13, 2012 at 10:01 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #7 · Re-assessing the D800 sensor as a Canon alternative


splathrop wrote:
I must owe you an apology for my lack of clarity. Because you have not understood what I am trying to say.

The story so far. I say Nikon D800/D800e files look lacking in color/contrast, or something. Nobody argues with that, not yet anyway. Some people seem to agree. I suggest reasons why that problem might be inherent in the high dynamic range sensor, at least in part because of its high dynamic range. I also suggest reasons why those problems might be hard to deal with in post processing—reasons that might account for the fact that people posting images from
...Show more


Sensor's capture in linear fashion. RAW images store a linear signal. What you are talking about it not true. It's just whoever developed the RAW files or in cam jpgs you didn't like used some software or altered the basic settigns in it to do something you didn't like. It's nothing at all wrong with the D800 sensor. You are totally off-base here. Nobody is saying to "fix" it since nothing is broken, they are just saying don't do what those particular people did when you develop your RAW files. The basic RAW file is linear and would be no different than the 5D3 signal other than having MORE subtle color gradations and less noise in the parts of the RAW file near signal level 0.




Oct 13, 2012 at 10:54 PM
akclimber
Offline
• • • •
Upload & Sell: On
p.2 #8 · Re-assessing the D800 sensor as a Canon alternative


I currently shoot with a Canon 5D2, 5D3 and Nikon D800e. The D800e sensor of course renders a more detailed file. At ISO 100 & 200 the DR of the D800e smokes the 5D2 and 5D3. I'm very, very impressed and hope Canon gets its act together soon to be able to match what the D800e sensor can do (I prefer Canon lenses and ergonomics - I don't see myself as a permanent Nikon guy). Interesting that the OP notes more subdued colors from D800e files. I tend to agree but that can be easily adjusted via your RAW converter of choice. The D800e files are capable of rendering colors just as rich as the Canon files. I think some of what is being seen (as I think was also pointed out up thread) is that D800 folks are creating more flat files either intentionally or unintentionally due to pulling out more shadow detail than has been previously possible. The curve of the final files may be "flatter". I have noticed I tend to like a bit more of a contrast enhancement bump in my RAW processing of the D800e files tho, as compared to the Canon files. I can't comment on the D800 skin tones - I'm a nature/critter/landscape/urbex kinda guy. I will say that at those low ISOs, 100 & 200, the files of the D800e leave the files coming from the Canons in the dust. Bump the D800 up to ISO 800 or beyond tho and Canon starts to pull ahead Altho given the D800's MP count, I'm pleasantly surprised at its performance at ISO 800 & 1600. At these ISOs, I prefer the Canon files but the Nikon isn't bad at all. I bit more grainy sure but it's easy to deal with in post (maybe easier then canon high ISO noise) and the D800 also looses its DR advantage.

To the OP, if you're on the fence, rent or borrow a D800 and judge for yourself. My bottom line is that for the stuff I enjoy photographing I'm blown away by the low ISO D800 files.

Cheers!



Oct 14, 2012 at 01:20 AM
splathrop
Offline
• • •
Upload & Sell: Off
p.2 #9 · Re-assessing the D800 sensor as a Canon alternative


akclimber, thank you for your thoughtful response. My own experience using the 5D II and post-processing files from that is that the right post processing solution is easier to find when the image and light conspire to leave large parts of the gamut and histogram empty, and progressively more difficult the fuller the gamut and the broader the histogram. In principle, of course, if the gamut were full and the histogram were full, then no edits could be made without corresponding sacrifices—a hard post processing problem.

I hadn't been thinking in those terms when I first considered the D800e, but it didn't take much looking at the D800e image thread to bring it back to mind. By the way, I have never doubted that D800e images could be rendered in colors as rich as the Canon files, and I'm fully convinced that in principle it ought to be able to do better. What I do question is whether it is going to be anything but difficult to do that to advantage while post processing the kinds of files that make the D800e seem so advanced. I wonder about more tradeoffs elsewhere in your image than Canon files demand—simply because the Nikon files are going to have more information elsewhere in the image. Sort of a paradox of excellence.

Someone might be tempted to ask how could that be bad? But it could be frustrating to have a file in front of you that showed color and contrast you wanted to improve, but that tended to punish editing with unwanted tradeoffs.

One thing the old modernist photographers understood to their advantage is that life is full of clipped shadows. Look around outdoors and you often see darker areas where you can't see detail. So choosing exposures to emphasize desired detail, while pushing unwanted distractions into deep shadow became an expert's tool that was very easy on the viewer's expectations. And perforce, that's what we have often been doing with digital images until these new Nikons came on the scene.

But that older, virtue-of-necessity technique is sort of the opposite of the D800/D800e's native tendency—to avoid clipping everywhere. Maybe all that would be needed would be a perverse willingness to say, screw the dynamic range, I'm going to clip me some shadows. You could still use the resolution to great advantage, and enjoy more control of color and contrast as well.

Of course your suggestion of trying one out is a good one. I'll probably look around a bit more and then go do that when I know enough to make efficient use of the experiment.

Also, are you content with live view focussing with your D800e? And am I correct in remembering that you sometimes use Zeiss lenses? If so, any comparative comments vis a vis Canon, or examples to illustrate results from those combinations would be appreciated.



Oct 14, 2012 at 03:39 AM
akclimber
Offline
• • • •
Upload & Sell: On
p.2 #10 · Re-assessing the D800 sensor as a Canon alternative


Stephen,

Lets see, regarding the D800 color and the difficulty or lack thereof in attaining a "Canon" look, I'm pretty sure that one could create some RAW conversion profiles in your converter of choice so that the D800's color response could be made pleasing with a quick button or two. I really enjoy raw conversions and playing with the added info from the D800 sensor but it may indeed require more time to dial things in the way you want - as you point out, it's like having a surplus of riches But heck, just because you have more shadow data (and it seems to me some added highlight data as well) it doesn't mean you can't clip that data to your heart's content.

Another interesting observation regarding the D800 files is that they seem to produce very nice B&W conversions (I use Nik Silver Efex Pro 2 as well as onOne's Perfect B&W). I can't explain it (other than pointing to added DR) but my D800 B&W conversions have a look to them that's distinct from similar Canon conversions. Normally, if I read an observation like that I'd think it was likely rubbish but my wife of all people confirmed what I thought I was seeing by asking one day whether the image she was viewing was from my "new camera". I said "Yeah, sure is, why?" My wife replied that she liked that B&W image much better than my other B&W images For whatever reason, she can reliably/consistently ID my D800 B&W images vs my Canon files. Go figure (and yet another pixel peeping thing to test when I get around to it!).

As for the D800's live view, frankly it sucks compared to Canon I find it good enough to focus to suit me (with the Zeiss) but it is more frustrating than Canon, that's for sure. Why neither Canon nor Nikon offer accurate, detailed focus peaking in liveview is beyond me. In fact, I just today downloaded the Magic Lantern software to use with my IR converted T3i to help with focusing (and to play with on my 5D2). I do occasionally resort to using my Hoodman loop to assist in both Canon and Nikon liveview focusing.

And yeah, if you have the means, rent a D800 - not only to check out the sensor and liveview but also the ergonomics. Nikon seems to prefer more dedicated buttons and less "customizability" while Canon seems to prefer a more menu driven and fewer, more customizable buttons. Neither is "better" really. I like the dedicated buttons of the Nikon but miss that more robust menu driven button customizability of the Canon. When first starting out with the D800 I found my hand cramping after a while but one I got used to the camera, all was well. I don't have a grip on the D800 but do on my 5D2 and 3. It makes a big comfort level difference for me but (1) I can't bring myself to go for another far too expensive DSLR grip and (2) I'm really trying to stay light and simple with the D800.

Regarding Canon vs Nikon Zeiss comparisons, to be honest I have lots of requests for just that very thing - from folks on the net and from our local photo club I'm sorry to say I don't have any yet. Our weather this summer and fall has been record rain and cold so the few times I have been able to get out have been spent enjoying my gear and surroundings, not testing stuff (altho I really do want to do some pixel peeping tests one of these days - I enjoy that kinda thing - and I'll post them when I do!).

Cheers!



Oct 14, 2012 at 06:39 AM
 

Search in Used Dept. 



splathrop
Offline
• • •
Upload & Sell: Off
p.2 #11 · Re-assessing the D800 sensor as a Canon alternative


akclimber, the very kind of info I hoped for when I posted. Thanks again.


Oct 14, 2012 at 11:57 AM
mmurph
Offline
• • • •
Upload & Sell: On
p.2 #12 · Re-assessing the D800 sensor as a Canon alternative


Splathrop,

I went to all digital processing (of 6x7 film scans) in 1998, and full digital acquisition in 2003/2004.

I profile all of my devices - including every camera - which in my mind is the **only** way to ensure a consistent, neutral, baseline file from which to begin color manipulations to achieve the "look" that you want. (Warmer/cloder/higher/lower saturation, higher/lower contrast, skin tone placement, shadow depth, highlight rolloff, etc.)

I have also mixed my own, custom B&W ink sets, made profiles for those, as well as run multiple custom ink sets through 3 large format printers (2x24" and 1x44"), making custom profiles for all of those too.

Many pros understand that their "Look" is a large part of their "Brand." I remember around 2004/2005 on Rob Galbraith, some very knowledgable and serious pros complaining that it took them 1 year to dial in the look that they wanted on the Canon 1Ds, only to have the (wonderful) Canon 1DsII come along, and now they had to try to adapt that to their workflow ...


Sort of the question: If I shot all Kodak Portra 160NC, why did other (less intelligent;> folks prefer Velvia?


All of which is to say that, yes, "In Theory", anything can be made to look like anything else.

In practice, on the artistic side, we find the tools that fit our style, adapt a workflow and process that suits us, and then go out and use our creative/right brain to create images (versus engineering color spaces and software workflows with our analytic/left brain.)

That is: I would not minimize the effort involved to find the look that you want when switching to a new camera system.

I had both the Canon 5DIII and the Nikon D800E on pre-order, the day that each was announced. I fully, fully intended to move to the D800E. I probably would have done that had it been released in November or December. But I decided I did not want to spend the summer (in Michigan), and my vacation time in the Upper Penninsula, with family, doing engineering tasks, buying and selling lenses, learning a whole new system, and NOT making images!

Then I realized that the marginal value to me was limited to large prints. Finally, in August, I shifted my focus (no pun) from large scale ISO 100 prints to low light, high ISO event photography, and realized I needed a 5DIII and bought that.

I still hope to buy a Nikon camera and a couple of lenses to explore toward spring. I just bought a Sony NEX 5n to play with the sensor and images as a "point & shoot".

So you are right to realize that in theory anything is possible, but in practice it is a lot of work to get to where you want your look to be, and where you know and are comfortable with your tools.

The only camera that I ever really felt like I knew and trusted completely was my Canon 1DsII, after about 40,000 images per year, for 2 years, and the time spent understaning those imges (and the results I got using the camera and lenses) in post.

If you do decide to make the move, and want to dial in the image, I would suggest this process:

1) Start with a Color Checker 24 square or Color Checker Digital camera calibration.

2) Open that profile, or create a new profile, in the Adobe DNG Profile editor.

3) Tweak and save the colors, especially skin tones, there to achive the type of color space mapping that you are looking for. You can adjust Saturation, Contrast, or Curves, etc. for each of the 24 color squares in the Color Checker, then "Export" that as a profile to be used in Adobe Camera Raw or Lightroom.

4) Use Lightroom "Presets" to create a set of user defined "Looks" for yourself: Slightly higher saturation/l;ower saturation, slightly wider DR/Narrower DR, etc.

Of course you can do teh same thing in Capture One, etc.

Only you can judge the realtaive effort involved, whether you like to do that kind of "color space engineering" work, etc. And what that down time or distraction means to running your day to day business, or doing creative work - making images.

Which is, after all, what this is supposed to be about.

Finally, here is a recent post on The Online Photographer from Ctein (and Eric Chan) on "Why RAW is not RAW" that may prove interesting reading. It is certainly the type of thing you will be dealing with at some level:

http://theonlinephotographer.typepad.com/the_online_photographer/ctein/

Good luck! have fun.

I didn't look to see where you are from, but if you have winter and grey and cold looming, like we do, you might have some down time to take on a project like this. I plan to edit 10+years of work, going back to 1998, create some portfoluios, a web site, and print.

Cheers!
Michael



Oct 14, 2012 at 03:37 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #13 · Re-assessing the D800 sensor as a Canon alternative


splathrop wrote:
akclimber, thank you for your thoughtful response. My own experience using the 5D II and post-processing files from that is that the right post processing solution is easier to find when the image and light conspire to leave large parts of the gamut and histogram empty, and progressively more difficult the fuller the gamut and the broader the histogram. In principle, of course, if the gamut were full and the histogram were full, then no edits could be made without corresponding sacrifices—a hard post processing problem.

I hadn't been thinking in those terms when I first considered the D800e, but it
...Show more

The D800 sensor doesn't spread the histogram out or alter exposure or anything all it simply has is less noise in the darkest parts at low ISO and it is less color blind under natural daylight color temps and that is the end of the story. It won't magically spread the histogram out or produce less contrast or anything.
Digital sensors of any basic design are all linear capture.

Now maybe Nikon Capture software or Adobe gives Nikon stuff a different post processing default tone curve, I have no idea, but that has nothing to do with the sensor and they could have give it one like the 5D3, with some time you can probably find how to pull the sliders to match the same starting point.



Oct 14, 2012 at 07:46 PM
splathrop
Offline
• • •
Upload & Sell: Off
p.2 #14 · Re-assessing the D800 sensor as a Canon alternative


akclimber, another question if I may. Have you made any systematized attempt to compare results in your D800e files with different exposure approaches? Given your last post, I suppose maybe not. But I would be intrigued to know what the overall effect on color and contrast is if you use a D800e more or less as if it had less dynamic range than it does, and try hard to expose to the right. I'm guessing that doing that would take some of the flatness out of the images, or at least facilitates post processing to accomplish that.


Oct 14, 2012 at 08:03 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #15 · Re-assessing the D800 sensor as a Canon alternative


splathrop wrote:
akclimber, another question if I may. Have you made any systematized attempt to compare results in your D800e files with different exposure approaches? Given your last post, I suppose maybe not. But I would be intrigued to know what the overall effect on color and contrast is if you use a D800e more or less as if it had less dynamic range than it does, and try hard to expose to the right. I'm guessing that doing that would take some of the flatness out of the images, or at least facilitates post processing to accomplish that.


RAW captures are linear flat. Exposing more to the right would just clip off more highlights. I don't think youa re understanding how the sensors capture things and behave.

If you don't want it flatter then simply cut change the blackpoint or if you do try to keep the extra details in the shadows and still wnat contrast apply a strong curve to midtones, maybe on the ACR sliders bring the second brightest slider up, bring the next one wayyy down and then bring the fourth and final darkest tone one up.



Oct 14, 2012 at 08:26 PM
Monito
Offline
• • • • • •
Upload & Sell: Off
p.2 #16 · Re-assessing the D800 sensor as a Canon alternative


Exposing to the Right doesn't change contrast. It takes advantage of the full dynamic range available to raise as many tones as possible up out of the noise floor.

I don't think skibum5 owns a D800e, but I could be mistaken.



Oct 14, 2012 at 08:40 PM
mttran
Offline
• • • • •
Upload & Sell: Off
p.2 #17 · Re-assessing the D800 sensor as a Canon alternative


skibum5 wrote:
RAW captures are linear flat. Exposing more to the right would just clip off more highlights. I don't think youa re understanding how the sensors capture things and behave.



+1,



Oct 14, 2012 at 08:46 PM
Monito
Offline
• • • • • •
Upload & Sell: Off
p.2 #18 · Re-assessing the D800 sensor as a Canon alternative


skibum5 wrote:
RAW captures are linear flat. Exposing more to the right would just clip off more highlights. I don't think youa re understanding how the sensors capture things and behave.

If you don't want it flatter then simply cut change the blackpoint or if you do try to keep the extra details in the shadows and still wnat contrast apply a strong curve to midtones, maybe on the ACR sliders bring the second brightest slider up, bring the next one wayyy down and then bring the fourth and final darkest tone one up.


mttran wrote:
+1,


I don't think you understand how Expose to the Right works.

First thing to understand is that if you wish to get the most noise free images, given whatever camera you have in your hand at the time, there are at least seven ways to obtain that.

1) Shoot on a very dull day, like a foggy day, a simple scene with no dark corners and no sun burning in. That will place the darker tones above the noise floor in an ordinary exposure meter setting and not have any highlights to blow out.

2) Shoot in a studio where you can control the light and all the tones and place them on the scale exactly where you want.

3) Add light outdoors by fill flash or reflector or by repositioning to be near a bright wall.

4) Diffusing bright light or subtracting light with a black surface (usually in a 5-in-1 reflector kit).

5) Multi-exposure HDR, even without a tripod. This includes manual blends.

6) The Zero Noise technique: one normal exposure and one exposure four stops over that is brought down in post and merged for noise free shadows.

7) Possibly use EttR. Outdoors or in uncontrolled lighting, if the lighting and scene are generating a contrasty image, then decide what tones are important to the composition and what tones can be sacrificed, especially what highlights can be sacrificed. It is important to note that even ordinary exposures outdoors will often have highlights that blow out like chrome, or window glass, or strong light sources, or snow, or pure white surfaces or objects, etc. So you always have to decide if there is detail in those areas that is important or not.

Sometimes you even have to EttL, Expose to the Left, to shift important highlights down so they don't clip. Then you increase in Raw conversion, but you have to sacrifice shadow tones.

But usually there are highlights that can be clipped without damaging the overall effect of the image after all Raw conversion and processing is said and done.

When that is the case, you shift the exposure up (more time, wider aperture) to raise the tones just enough but not too much, and then bring them back down again during Raw conversion. This does not affect contrast. Contrast is a separate post-processing operation, though simple adjustments to contrast are often done simultaneously in Raw conversion (DPP, LR, ACR).

So, yes, EttR will clip more highlights which is why it is not always applicable though often is, but no, EttR will not make the image flatter.

(I will not be surprised if we get a cryptic one line response from mttran and a lengthy response from skubum5 with details of sensor tests and references to sensor design, etc.)



Oct 14, 2012 at 09:16 PM
mmurph
Offline
• • • •
Upload & Sell: On
p.2 #19 · Re-assessing the D800 sensor as a Canon alternative


I don't want to enter into these wars, it is not worth my energy usually.

But here is some info published quite recently on why RAW files are not truly linear. Including input from Eric Chan, of Adobe:

http://theonlinephotographer.typepad.com/the_online_photographer/2012/10/raw-is-not-raw.html



Oct 14, 2012 at 10:45 PM
mttran
Offline
• • • • •
Upload & Sell: Off
p.2 #20 · Re-assessing the D800 sensor as a Canon alternative


Monitor, look at this sample here: http://www.fredmiranda.com/forum/topic/1155718/0#11029322
and tell us your masking (issue) processes still work in these cases. Another + for Skibum response:

I don't think you' re understanding how the sensors capture things and behave.




Oct 14, 2012 at 10:50 PM
1      
2
       3       end




FM Forums | Canon Forum | Join Upload & Sell

1      
2
       3       end
    
 

You are not logged in. Login or Register

Username   Password    Retrive password