fredmiranda.com

Moderated by: Fred Miranda

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 … 9

11 end

Archive 2012 · Canon 5D Mark III and Nikon D800 Review

n0b0
Offline
• • • • •
Upload & Sell: Off

p.10 #1 · Canon 5D Mark III and Nikon D800 Review

It's the same with DR. Before D800 people managed to take beautiful photos but now people talk about the D800 high DR sensor like they can't take beautiful photos anymore without it.

Apr 23, 2012 at 11:10 PM

cameron12x
Offline
• • • •
Upload & Sell: On

p.10 #2 · Canon 5D Mark III and Nikon D800 Review

n0b0 wrote:
It's the same with DR. Before D800 people managed to take beautiful photos but now people talk about the D800 high DR sensor like they can't take beautiful photos anymore without it.

Exactly.

Although I will say that having the D800's DR would be a Godsend in many difficult lighting conditions.

In a normal contrast scene this capability is less important.

It's the difficult-lit and high contrast scenes that bring attention to this extraordinary capability of the D800 sensor.

Apr 23, 2012 at 11:18 PM

jamesf99
Offline
• • • • •
Upload & Sell: Off

p.10 #3 · Canon 5D Mark III and Nikon D800 Review

cameron12x wrote:
The choice could be pretty simple:

1. Determine the average viewing distance that you believe that your prints will be viewed from.
2. Determine the average print size that you believe you will be printing at.
3. Determine the needed density/resolution of the print (e.g. 300dpi) in order to provide adequate viewing.
4. Calculate the needed sensor resolution needed to accomodate 1-3 above.

To repeat, so often, we work in the wrong direction when analyzing a given situation.

It's improper to assume for example that 36mp will automatically provide any usable resolution advantage over 22mp until the final output medium, viewing distance, and needed density/resolution...Show more →

who mentioned MP? Not me. But if you're satisfied with 22MP, or even wanted 25MP, you can always get that out of the Nikon. I don't know how it works though, if it's a crop (my guess), if they drop pixels, or what.

I don't use LV, and I know lots of others that shoot without it too, but it *is* nice to have. it wasn't on any camera before 2007, and i regularly use bodies predating that so it's not an option. Sure, Nikon probably made a mistake, but there's a chance they can fix it. To me, the sensor IQ is what matters more, and with digital, the sensor is like film that can't be changed.

Canon provided some of the things that many people have wanted (finally a working AF implementation, no longer an offensive joke on the FPS, and they finally dropped the bull shit 95-98% VF) but then they take away VF screen replacements, but most seriously they didn't do anything on the sensor for 4-5 years. They give us video crap that I absolutely could not care less about if i tried. Digression - Anyone that was remotely satisfied with the 5d2 should be happy, they'll have the best of everything in a D800.

If you watch the YouTube video, the guy is satisfied it won't be a problem for him, but he's working during the day. i don't know if low light/night would cause problems, which is my one minor concern.

Anyway, LV aside, the issue comes down to the sensor for me. I'd love to shoot more B&W, and with two more stops of DR, I pretty much guarantee you that the images coming from the Nikon are going to be superior; that kind of added range can do wonders for B&W (and color of course).

I don't think anyone can argue with the results, and Fred showed it pretty effectively, as have other people before him..

Apr 23, 2012 at 11:30 PM

jamesf99
Offline
• • • • •
Upload & Sell: Off

p.10 #4 · Canon 5D Mark III and Nikon D800 Review

cameron12x wrote:
Exactly.

Although I will say that having the D800's DR would be a Godsend in many difficult lighting conditions.

In a normal contrast scene this capability is less important.

It's the difficult-lit and high contrast scenes that bring attention to this extraordinary capability of the D800 sensor.

Actually, people just put up with it, including me. It has been a long 14 years of dealing with insufficient digital DR (though it's not as bad as it was in the 90's, that's for sure). Tonal range is something we lost, and until Sony/Nikon began returning it, it has been something we always need to over come through pp gyrations (and we probably still will to some extent). I shoot a LOT of high contrast scenes and always have to compromise on what I'm going to ruin, the highlights or the shadows and it's extremely wearing.

If you've seen B&W film photographs, you'll know what you're missing. The reason I keep mentioning B&W is because it unambiguous. The values are easy to see, as are the transitions.

This is not even remotely comparable to LV, a new feature that is not necessary most of the time.

Apr 23, 2012 at 11:37 PM

cameron12x
Offline
• • • •
Upload & Sell: On

p.10 #5 · Canon 5D Mark III and Nikon D800 Review

jamesf99 wrote:
the issue comes down to the sensor for me. I'd love to shoot more B&W, and with two more stops of DR, I pretty much guarantee you that the images coming from the Nikon are going to be superior; that kind of added range can do wonders for B&W (and color of course).

As I've mentioned multiple times in this thread (and others), the only reason I'd consider using the Nikon D800 would be because of the dynamic range advantage it has, which is very significant in my eyes.

The LV issue may be troublesome for some and not for others. Show-stopper "might" have been the incorrect word to use as a general use case for this issue, but many landscape photographers are adamant about it.

But no one can discount the DR differences between the two systems and that is what is most important to me too. Canon has a lot of catching-up to do in that regard...

Apr 23, 2012 at 11:41 PM

AJSJones
Offline
• • • •
Upload & Sell: Off

p.10 #6 · Canon 5D Mark III and Nikon D800 Review

n0b0 wrote:
I agree, but the problem is, people here often argue based on 100% crop that they see on their monitor, which are most likely very different. Some might not even be calibrated. That in my opinion, renders the argument pointless.

On the other hand, most folks participating in these discussions here will have some idea of how their 100% view on their own screen relates to the prints they make on their printer, so that evens it out. The comparisons are valid and they will have an idea of what that 100% crop will look like when printed - and can give their opinions based on that. Then there's the folks who don't know what calibration is and all their prints look awful, but there's no-one here by that name

Apr 23, 2012 at 11:52 PM

AGeoJO
Offline
• • • • • • •
Upload & Sell: On

p.10 #7 · Canon 5D Mark III and Nikon D800 Review

Talking about prints... I read it somewhere that the DR of prints (paper and your printer) is limited to a value (I don't remember the exact value, sorry

), but for sure it would be even lower than the DR what Canon sensor can produce. With other words, if you want to print and print big, which is the main reason of having a 36MP camera with a wide DR, the limiting factor is the actual output (print, in this case)? Can anyone confirm or dispute this? On the other hand, a good and well calibrated monitor can actucally showcase the extended DR better. Again, I am just throwing it out there....

Apr 24, 2012 at 12:04 AM

stanj
Offline
• • • • • •
Upload & Sell: On

p.10 #8 · Canon 5D Mark III and Nikon D800 Review

You can compress the DR when printing. Even if your input image were to have 14 stops DR, you could compress it in print to 7 or 8 or whatever paper has. The resulting image will be less contrasty, but it will have all the detail of the original. In some situations, that's more important than a "nice contrasty" image.

Also, you can unevenly compress the DR before print. Nobody says that you have to compress it equally - you can choose to recover more highlights and/or shadows, and compress the center.

Apr 24, 2012 at 12:11 AM

AGeoJO
Offline
• • • • • • •
Upload & Sell: On

p.10 #9 · Canon 5D Mark III and Nikon D800 Review

Stan, or anybody else for that matter, correct me if I am wrong here. So, depending on the paper you use, among others, the DR is limited to approximately 7-9 or so while Canon sensor can generate what a DR of 11-12 or so? Granted, I would prefer a sensor with DR of 13-14 or so but if you print, the expanded DR won't show then? I am not discounting a sensor with a wide DR but this puzzles me somewhat, if true. I am strictly referring to DR here and not MP count.

Edited on Apr 24, 2012 at 12:35 AM · View previous versions

Apr 24, 2012 at 12:26 AM

stanj
Offline
• • • • • •
Upload & Sell: On

p.10 #10 · Canon 5D Mark III and Nikon D800 Review

No this is not quite correct. This is because you can _compress_ the DR before printing. Again, you will get a more "flat" image as a result, but it can be done.

Don't know if you ever printed manually in the lab, but for BW there is multi contrast paper, and for color there are different contrast papers. With BW, you would stick a negative into the enlarger and use one filter to give the image a more contrasty look, and another one to give it a flatter look. It depended on what you wanted to achieve. All from the same negative. You could take say a 7 stop negative and make it exceed the paper's DR, or vice versa. With color paper, you could not use color filters but had to use separate papers - portra, supra, ultra for example - but with the same result.

So, the capture DR matters a huge deal. You can choose to use it (or not) with print. It can be done.

Apr 24, 2012 at 12:34 AM

AJSJones
Offline
• • • •
Upload & Sell: Off

p.10 #11 · Canon 5D Mark III and Nikon D800 Review

AGeoJO wrote:
Stan, or anybody else for that matter, correct me if I am wrong here. So, depending on the paper you use, among others, the DR is limited to approximately 7-9 or so while Canon sensor can generate what a DR of 11-12 or so? Granted, I would prefer a sensor with DR of 13-14 or so but if you print, the expanded DR won't show then? I am not discounting a sensor with a wide DR but this puzzles me somewhat, if true. I am referring to strictly DR here and not MP count.
Think of how HDR works - you can start with 10 or 12 or 14 stops of DR and you can compress it down to 8 for your printer limits. You don't have to compress the midtones the same as you compress the shadows or highlights. Let's say you start with 12; you can compress the three top stops down to one and the five midtones to three you've got four left to show the details from all 4 stops in the shadows. Or leave the highlights intact and compress the shadows. Or some smoothed variation on the theme.

Apr 24, 2012 at 12:39 AM

stanj
Offline
• • • • • •
Upload & Sell: On

p.10 #12 · Canon 5D Mark III and Nikon D800 Review

Then there's the whole DR vs. bit depth misconception. You can mathematically capture infinite DR in one bit -- it just won't look very good. More practically, you totally can capture 16 stop DR in 8 bits - but it would look pretty silly. One has nothing directly to do with the other - but it "makes sense" that they are related. But for instance with the modern Nikon sensors you're getting to a point where they have more DR than bits in the default 12-bit capture mode. Or heck with 8-bit JPGs.

Apr 24, 2012 at 12:40 AM

AGeoJO
Offline
• • • • • • •
Upload & Sell: On

p.10 #13 · Canon 5D Mark III and Nikon D800 Review

Stan, I did a fair amount of B&W printing back then during the film days. So, I am quite familiar with that and that's not what I meant. I am now referring to strictly printing of digital images. Again, I understood, I read it somewhere that the printed media right now is limited to a certain DR, which is quite low and I am questioning whether the expanded DR of the D800 could be realized in the current printing media. That's all.

I do have a printer that is gathering dust and several large prints I had made was done by a digital lab. So, my printing experience or lack thereof is prevalent.

Apr 24, 2012 at 12:43 AM

stanj
Offline
• • • • • •
Upload & Sell: On

p.10 #14 · Canon 5D Mark III and Nikon D800 Review

Ok. Well, the simple answer is yes, the high capture DR can be realized, at the expense of overall image contrast.

Apr 24, 2012 at 12:45 AM

AGeoJO
Offline
• • • • • • •
Upload & Sell: On

p.10 #15 · Canon 5D Mark III and Nikon D800 Review

Stan, Andy, thank you!

OK, so, depending on how you want the print to look like, you can "manipulate" certain spectrum to create the look you want to see in the final print. Got it!

The expanded DR can be realized in a print but the print won't look good since it is not contrasty enough then? So, you deliberately compress the DR to a certain degree again to make the print look good. This where I am lost, I guess

. So, in a well printed (landscape) image, the best output would be if the DR is limited to a lower value again it seems?

Apr 24, 2012 at 12:59 AM

stanj
Offline
• • • • • •
Upload & Sell: On

p.10 #16 · Canon 5D Mark III and Nikon D800 Review

What makes a good print is highly subjective

I have been often accused of printing "boring", low contrast images that convey the full DR of the scene, while others would have produced a much more "dynamic" version of it with generous black portions, for instance.

You can either linearly squish the DR into a paper safe range, which will make for an evenly low contrast print, which is the thing I prefer. Or you can apply some sort of an S-curve, or a completely arbitrary curve, where you increase the contrast of say the midtones, to give the image more punch again.

I think the most important thing to realize is that with any modern day printer, one that's properly profiled, you don't need to do _anything_ in order to transfer what you see on screen to paper. The black pixels on your screen will be black on paper, and the white pixels on screen will be white on paper (i.e., no ink laid down - as white as it gets). But because the screen has a higher contrast ratio than paper, the print will look much, much flatter, and you may find yourself applying an S curve to the print to give it some punch on paper.

If I weren't as tired as I am now, I'd draw a sketch

You basically are looking at two projections - one from the raw file to your say LR display on the screen, and then that display to your printer's output. Each projection is lossy and can either preserve the DR at expense of contrast ratio, or preserve the contrast ratio at the expense of DR. You can freely control the compromise - say maintain DR at the top and maintain contrast at the bottom, or vice versa. Statistically, most people prefer maintaining contrast in print, I swim against the current with generally preferring to preserve DR at the expense of contrast.

Apr 24, 2012 at 01:35 AM

Pixel Perfect
Offline
• • • • • • •
Upload & Sell: Off

p.10 #17 · Canon 5D Mark III and Nikon D800 Review

And to add to what stan has said of course this is also highly dependent on the paper you print on. You wouldn't print an ultra punchy colour image on matte paper say (well I wouldn't), you'd tend to use a gloss paper to obtain a larger DMAX.

As long as you have a fully coloured managed workflow and good icc profiles for the paper it's all a seem-less process. You can do softproofing to see what the paper you have chosen will do to your image and you can choose to bump up the contrast or whatever, to improve the DR of the image before doing a real print.

Apr 24, 2012 at 02:14 AM

n0b0
Offline
• • • • •
Upload & Sell: Off

p.10 #18 · Canon 5D Mark III and Nikon D800 Review

stanj wrote:
Ok. Well, the simple answer is yes, the high capture DR can be realized, at the expense of overall image contrast.

That's because monitor technology hasn't caught up with it yet. Well, I think there are a few HDR monitors but they probably cost a fortune.

Apr 24, 2012 at 02:44 AM

WebDog
Offline
• • •
Upload & Sell: Off

p.10 #19 · Canon 5D Mark III and Nikon D800 Review

Ralph Conway wrote:
@ chez:
I am not used to. I did a couple of fashion posters for a customer in 90x60 cm. They looked fantastic. But for me it was at the border. It was 8MP/99 dpi. I could NOT see any bad quality from 1 meter distance. But I "felt" uncomfortable. So I switched to 21MP. That is 156 dpi without cropping (what I like to prevent). What compares to a 12x18 printout in 300 dpi. I guess it will be enough for me for ever.

Your calculation is wrong. Indeed it is 65% more pixels (based on 22) and generates a...Show more →

Given the same resolution (linear)
Not sure how you do your math... 22x1,64=36 right... and to translate this to linear you have to take squareroot(1,64)=1,28 hence 28% larger. The area will be 64% larger though

Apr 24, 2012 at 04:01 AM

Ralph Conway
Offline
• • • •
Upload & Sell: Off

p.10 #20 · Canon 5D Mark III and Nikon D800 Review

So 64% instead of 65 and 31 instead of 28. Sorry, it was late last night and I did a rough estimation in my tired head ...

Apr 24, 2012 at 04:51 AM

1 2 3 … 9 10 11 end
FM Forums \| Canon Forum \| Join Upload & Sell 1 2 3 … 9 10 11 end