DXO uh-oh?

[scyrene]

I've no doubt that the D800 and D810 are good for landscape work - having all those extra megapixels, and I don't think anyone here disputes that Nikon is better at base ISO (I only know this from reading this forum). They may well be the best full frame cameras for this purpose, I don't know (I don't do much landscape work).

But to claim Canon is bad at ISO 100 is hyperbole. And to claim the 5D3 is 'painterly' at ISO 800 (or just over) is... perplexing. It's a subjective term of course. But the raw files look good much higher than that - depending on what and how you shoot. I find bird photos can look okay up to ISO 6400-8000 - using the technique of slightly overexposing and then reducing the brightness later (to reduce shadow noise) - and I'm very happy with results at ISO 3200.

As for the shots linked to, I loved the second one, but the one with the jetty was... not to my taste. That's fine, it's stylistic, and I'm happy with people who love their gear. But to trash other people's choices based on that opinion is unwise. After all, if the D800 was so obviously superior, why did it not sell so well (again, something I've seen repeated here often).

 

[unfocused]

Except reproducible double-blind experiments.

As an example of how "experts" can be fooled: http://www.theatlantic.com/health/archive/2011/10/you-are-not-so-smart-why-we-cant-tell-good-wine-from-bad/247240/

These types of experiments have been replicated many times, and the summary is that you can't trust your own perceptions. Show me objective tests, or you've shown me nothing.

Exactly. Sometimes people believe what they want to believe. Your expectations shape reality. Hence the quest to find some definitive source that can tell them what really is good and what's junk. The problem then lies in whether or not you truly trust the source of this information.

Yes, almost all the threads on this forum are filled with endless examples of confirmation bias.

If I spent $12,000 switching camera systems, I would swear on a stack of bibles that I could see a difference in the results. And, I would really believe I could.

Neuro used to argue incessantly that there was no discernible difference between APS-C and Full-Frame. Then be bought a 1D-X and now he argues the exact opposite. It's not hypocrisy. It's not enlightenment. It's just that everyone sees what they want to see.

There's another thread started here referencing a Zach Arias rant on APS-C vs. Full Frame. Amusing, but makes valid points about how little difference there is between the two format sizes http://www.canonrumors.com/forum/index.php?topic=21991.msg418802#msg418802 yet the response that really struck me was that this was just a "Fujifilm commercial."

Point being we all selectively hear and see what we want.

This forum works and generates the traffic that it does because the selective use of facts, confirmation biases, pseudo-scientific testing, etc. etc. that everyone likes to quote or argue about is never truly objective and never completely accurate, so the arguments can just go on and on forever.

 

[Don Haines]

I've no doubt that the D800 and D810 are good for landscape work - having all those extra megapixels, and I don't think anyone here disputes that Nikon is better at base ISO (I only know this from reading this forum). They may well be the best full frame cameras for this purpose, I don't know (I don't do much landscape work).

But to claim Canon is bad at ISO 100 is hyperbole. And to claim the 5D3 is 'painterly' at ISO 800 (or just over) is... perplexing. It's a subjective term of course. But the raw files look good much higher than that - depending on what and how you shoot. I find bird photos can look okay up to ISO 6400-8000 - using the technique of slightly overexposing and then reducing the brightness later (to reduce shadow noise) - and I'm very happy with results at ISO 3200.

Exactly!
Whatever you go for now is way better than just a few years ago... Anyone want to buy my Olympus E-300? ISO 400 was as high as you would shoot at, 800 was tearful, and now a GoPro takes better quality stills.....

I am currently shooting with a 60D... ANY Canon/Nikon/Sony/Panasonic/Olympus DSLR or mirrorless out now beats it for IQ, yet somehow I keep getting great pictures with it......

 

[Aglet]

@zigzagzoe
Thanks for your posts.

There's a few of us here who will heartily agree with most of what you said. I was a Canon fan until the superior raw files from a $400 Nikon consumer body blew my 5D2 into the weeds! Well, at that time I was already a Canon fan in decline and searching for better options as I'd realized the 5D2, so hyped by fanboys, was a bit of a turd with very noisy low ISO raw files and a tendency to underexpose; or at least the one I had was.

Your example of the model, whose arm was in deep shadow, and looking like it has a bad rash because of the read noise, is a good one.
When pushing the limits, sure, there are plenty of "work-arounds" to make Canon cameras able to produce the desired shot. People doing that seem to be forgetting they're compensating for under-performing hardware! Poor hardware IS a good way to improve your skills as a photographer, as you have to be extra creative to overcome the limitations of your tools. Reduce the limitations placed on your shooting by those under-performing (Canon) tools and you're now free to benefit from a range of other advantages like less setup time, less time in post, more freedom and flexibility in lighting and exposure, etc.

Ardent fanboys will continue to argue that technique matters more. It sure does, especially when you chose to use gear with more limitations for certain kinds of shots.

As for the lens argument some make, sure, Canon has some excellent lenses. But there's also plenty of good glass for Nikon and no matter how bad the Nikon lens may be, it still doesn't change the read noise issue.

My D800/e are my favorite bodies for my kind of shooting yet and I only wish I would have not been so prejudiced in the past so I could have benefitted from using the advantages of other camera brands much earlier. I haven't purchased a new Canon camera for years now. But I have purchased Nikon, Pentax, Fuji and Olympus and have enjoyed the benefits they all bring while none of them are as limited by FPN and read-noise issues as Canon.
Despite the sales and marketing success, Canon is now the underdog, and I actually hope they can perform some sort of a comeback miracle; it might give me one more tool to use.

EDIT: BTW, I would not be surprised if Nero isn't already pondering and composing an acrid response to your posts.

 

[Loswr]

I'm a long time reader, first time poster, but I have to address Neuro here.

Welcome to the Forums. I have to say, I find your characterization of me to be rather rude and offensive. You are, of course, welcome to your own opinion...but it's unfortunate that your opinion is apparently based on comments by the likes of dilbert, rather than my own posts.

Can you find a post where I claim that Canon sensors deliver better low ISO dynamic range than Nikon/Sony sensors? No, because the opposite is true, as I've stated more times than I can count. I've also lost count of the times I've stated that if I were primarily a landscape shooter, I'd be using a D800E and 14-24/2.8G.

The issue at hand is that for you, like most people, everything is colored by your personal viewpoint and experience. For example, I say DxO's Scores are biased, but you say they back up your experience. Given your statement, "I am not a high ISO shooter," perhaps you don't see the bias inherent in their Scores...because that bias favors your shooting needs. How is that 'open minded'?

...but please, can we have some debate that actually understands what DR is...

Ok, but you'll have to excuse yourself from that debate. When you make statements like, "With over 14 stops at 100 ISO (14.8 at 32ISO on the D810 it would appear)," you demonstrate that your understanding of the relevant technical issues is quite poor. The D810 has a 14-bit ADC, it is not capable of recording over 14-stops of DR in a RAW image. DxO's 'Landscape Score' of 14.8-stops of DR results from a mathematical simulation of downsampling that 36 MP image to 8 MP. If you go out and meter a scene that shows a 15-stop difference from darkest to brightest, and take one image with your D810, you'll lose 1.25-stops of some combination of shadows/highlights, depending on your exposure. That's at low ISO...once you get above ISO 800, the D800/810 DR advantage evaporates.

I don't believe in the concept of "pure IQ" – I believe in taking pictures. A better sensor coupled with a worse lens does not make for a better picture. A sensor with 20-stops of DR coupled to a 600mm f/4 lens that I cannot handhold would not adequately meet my needs.

Everyone's needs are different. Aglet needs to shoot images (sometimes with the lens cap on) and push the files 4-5 stops in post. It's rare that I need to push an image more than 1 stop, and I don't think I've ever needed to push an image more than 2-3 stops (in those rare cases when I completely screw up the exposure).

The problem I have is when people assume their needs represent the needs of the majority, and what they find to be a limitation is universally applicable.

If the D800/810 meet your needs better than your Canon gear did, then switching was the right decision and good for you. You didn't like the 5DIII? That's fine.

The D800/810 sensors have better low ISO DR than any Canon sensor. But...people don't buy bare silicon sensors, they buy cameras. I believe the 5DIII is a better 'all around' camera, and the sales figures are consistent with that belief. More people chose to buy 5DIII's than D800/E's, just as more people have chosen Canon dSLRs over Nikon dSLRs for at least the past 10 years. That's objective reality. Does it mean Canon is 'better'? No...only that Canon is chosen by more people to better meet their needs.

 

[scyrene]

With over 14 stops at 100 ISO (14.8 at 32ISO on the D810 it would appear),[/i]" you demonstrate that your understanding of the relevant technical issues is quite poor. The D810 has a 14-bit ADC, it is not capable of recording over 14-stops of DR in a RAW image. DxO's 'Landscape Score' of 14.8-stops of DR results from a mathematical simulation of downsampling that 36 MP image to 8 MP. If you go out and meter a scene that shows a 15-stop difference from darkest to brightest, and take one image with your D810, you'll lose 1.25-stops of some combination of shadows/highlights, depending on your exposure. That's at low ISO...once you get above ISO 800, the D800/810 DR advantage evaporates.

I nearly mentioned this, because I thought I'd heard it before, but I'm not confident enough to talk about these things. I'm glad you did As you say, it's a system that matters, not one element.

It's funny how the people accusing folk of being 'fanboys' are the ones making the most extreme sweeping statements like Canon's offerings are 'blown away' by competitors. It's not as if we're even talking about fundamental differences - just fairly minor incremental improvements.

 

[Haydn1971]

Hmm, I smell a troll ;-)

 

[Don Haines]

There are two things that we can't forget. The first is that technology marches on. The steps are not even... sometimes there are incremental improvements that are almost indistinguishable from what came before, sometimes they go in leaps. The Nikon/Sony sensors took a jump ahead in quality a few years ago... now they are the hot ticket. In comparison, Canon has had several years of tiny incremental improvements and are long overdue for a jump. That Jump may very well be the 7D2 with DPAF, on sensor A/D, finer lithography, etc etc... They know what they have to do to jump ahead and they are not idiots.... it will be done at some point and then Canon fanboys will be able to rightfully sneer at Nikon fanboys. And I expect that in another few years it will be Nikon jumping back into the lead.... and then Canon.... and then Nikon.... and so on....

The second thing you can't forget is that a camera is more than the sensor. While Nikon has been fixated on it's sensors, Canon has been fixated on it's focusing system. The end result is that the AF system on Canons is far more capable than on Nikons.... The people at Nikon are not idiots... you can bet that they are working on ways to counter this and that the next few rounds of their cameras will have better and better AF systems.

and getting back to the main topic, DXO...
DXO ratings are biased. period. The camera ratings are biased towards DR. The lens ratings are biased towards T-stops. I doubt that anyone would argue that no such bias exists because it is prominently shown in their numbers. Is it a pro-Nikon bias or an anti-Canon bias? Odds are no, the bias is there because that's what the people who came up with the rating scheme believe in... It is very easy to look at things and believe great conspiracy theories, but it is far more likely that things are the way they are because that is the easiest way to do things.

We all have different likes and different needs and there are various genres of photography. A simple rating scheme ending up with a magic number is not a good solution.... but it is an easy solution. A far better solution would be for the user to be able to enter in weights for the various parameters to reflect their needs... but this solution is far more complicated to set up and has even greater potential to give poor results unless the user has a solid grasp of the subject..... in which case, they don't need that rating number anyway.

 

[Sporgon]

Hmm, I smell a troll ;-)

I think you smell someone who blew £19,000 changing and is determined to believe it was worth it.

I'm a low ISO shooter. I tried the D800 some time ago and I'm about to get my hands on a D810. I can't speak for the D810 but with the D800 there is simply not the 'night and day' difference that these people refer to.

 

[Haydn1971]

I honestly can't get my head around why anyone would chuck in a full kit of either Canon or Nikon to swap to the other, the gains to be had are just not worth the hassle.... Sure if you've a camera and a kit lens, swap away, but I'd need my head seeing to if I sold up and moved wholesale to Nikon, Fuji, Sony, Pentax, Olympus, Samsung or other.

Both Canon and Nikon are great camera ranges, with great bodies and lenses, just a few differences that balance each other out, really, get a grip !

 

[Orangutan]

I honestly can't get my head around why anyone would chuck in a full kit of either Canon or Nikon to swap to the other, the gains to be had are just not worth the hassle.... Sure if you've a camera and a kit lens, swap away, but I'd need my head seeing to if I sold up and moved wholesale to Nikon, Fuji, Sony, Pentax, Olympus, Samsung or other.

Both Canon and Nikon are great camera ranges, with great bodies and lenses, just a few differences that balance each other out, really, get a grip !

Chasing shiny things -- the same reason people dump solid employment for marginally more promising employment, or change spouses once the "magic" is gone (i.e. after hormones have diminished)

At least changing your kit is just money, though more than I'd want to spend.

 

[unfocused]

Hmm, I smell a troll ;-)

I think you smell someone who blew £19,000 changing and is determined to believe it was worth it.

I honestly can't get my head around why anyone would chuck in a full kit of either Canon or Nikon to swap to the other, the gains to be had are just not worth the hassle...

If they wanted to sponsor me, I'd switch.

Seriously though, I absolutely agree. And, while I'm convinced that the refresh rates of DSLRs will be extended with maturing technology, they are still not exactly long-term investments. So switch today and then when the next model comes out, what? Switch back?

When the Canon 5DIII came out with its high ISO performance, I could understand wedding photographers who live in low light and cutthroat competition switching because every little advantage is important. And, I suppose the five or six people who actually can earn a living shooting large scale landscapes might want to move to Nikon, but for most people, I just don't get it.

Chasing shiny things -- the same reason people...change spouses once the "magic" is gone (i.e. after hormones have diminished)

At least changing your kit is just money, though more than I'd want to spend.

I've changed spouses twice in my life (once my choice, once her choice). Talk about money...

I'm hanging on to the current one...third one the best ever and worth the wait. Besides, I don't imagine I have that many years left anyway.

 

[edknuff]

Personally, I'd love to have the Nikon D810, but only if I could put my Canon glass on it!
Nikon doesn't make the high quality glass that I need, for the lenses that I use. For example: Canon TSE 24mm mark2.

 

[Orangutan]

Personally, I'd love to have the Nikon D810, but only if I could put my Canon glass on it!
Nikon doesn't make the high quality glass that I need, for the lenses that I use. For example: Canon TSE 24mm mark2.

I'll take this opportunity to trot out one of my favorite photo wishes: an industry-standard SLR mount so we can freely interchange cameras and lenses across manufacturers.

Ain't gonna happen in the U.S. Hey, EU! We need your regulatory assistance here! 8)

 

[Sporgon]

And, I suppose the five or six people who actually can earn a living shooting large scale landscapes might want to move to Nikon

You're right about five or six people compared with wedding photographers, maybe less, but one of those is a British guy called Colin Prior. He is one of the world's best known landscape photographers and has shot primarily on LF film with same MF film for more inaccessible places. However in mid 2013 he began using digital - FF - and Canon at that, 1Dx and 5DIII. He is on record as saying the files from the 5DIII are the cleanest he's ever come across. So here you have a real landscape photographer using a 5DIII.

As I said earlier, there just isn't the dramatic difference at low ISOs that these people like to believe.

 

[Don Haines]

Hmm, I smell a troll ;-)

I think you smell someone who blew £19,000 changing and is determined to believe it was worth it.

If I had £19,000, I could invest it at a rate of return that would allow me to buy 6 beers per day forever..... My photography wouldn't be any better, but I'd be too drunk to care

 

[jrista]

I can’t give up clean shadows and over 14 stops at 100 ISO, even if my heart is really with Canon due to my history with them.

Only complaint I have, and the reason I persist in my critique of DXO, is the native RAW files actually have 13.2 stops of DR. The 14+ stops of DR ONLY occurs when you downsample your images to a very specific size: 8x12 @ 300ppi. I believe that is a very critical point, as theoretically it is impossible to have more than 14 stops of dynamic range when using a RAW image that uses 14 bits per color channel. It's also critical that we always edit our RAW images at full size...otherwise, we wouldn't be editing RAW, we would be editing exported TIFF or JPEG images produced from RAW, and that amazing editing latitude would disappear into thin air.

You can downsample (or upsample) your images to any range of sizes, from 1x1 pixel to tens of thousands by tens of thousands of pixels. The dynamic range that each individual photographer may get from a D800 is completely arbitrary. Using the 8x12" print @ 300ppi ONLY has relevance when comparing cameras on DXO's site. Referencing the 14.2 stop number (or 14.8 stops in the case of the D810) outside of the context of DXO, such as you have (not to fault you, your post was great) here is actually invalid, and it's THAT, THAT very thing, that gives me persistent cause to call DXO out for producing misleading results: When actually editing a RAW in a RAW editor like Lightroom, the dynamic range of the D800 files is 13.2 stops, no more. (Similarly, the dynamic range of Canon files is somewhere between 10 and 11 stops, so that doesn't give Canon any kind of benefit, the same issues apply to DXO's tests of Canon sensors...in reality, Canon users are still stuck in the realm of ~10.5 stops of dynamic range instead of ~13, so there IS a benefit to using a Sony Exmor sensor if you use low ISO a lot.)

When your in a tool like Lightroom, lifting shadows, then all you have is the 13.2 stops of the only real "measure" of dynamic range that DXO does: Screen DR (it's something you can select when browsing through DXO's results on their site.)

I want them to blow everyone out of the water and deliver a killer high MP, hi DR sensor and give me my well missed rate button back.

But life goes on, and Canon have a following that means they are still the number one as far as sales go due to their reputation from prior to 2012, and I guess that won’t change for 2-3 years yet, so they have time to get it right.

But lets not deny reality.

If they don’t catch up, a good brand name is not going to last forever.

I couldn't agree more with all of this! While for someone like myself, who pretty much always shoots at ISO 400 or above for the vast majority of my photography (birds, wildlife, astrophotography), the difference between Canon cameras and any camera from the competition is negligible. If you always shoot at low ISO, there is no denying that Canon has noise problems, and anything that uses a Sony Exmor (which is now a fairly good number of cameras from a range of brands now) is going to produce superior low ISO results. Canon hasn't really, fundamentally changed their technology in...what...a decade plus? They have made evolutionary improvements every couple years...added microlenses, removed the gaps between microlenses, increased quantum efficiency (although recently, their improvements there have been less), and a few other things, fundamentally the core technology is the same: 500nm process, off-die, high frequency ADC, HIGH read noise.

I do agree, if Canon doesn't do something to catch up to and compete with the competition, the high Canon is still riding will fade, then disappear, and then the bottom will fall out. Canon is the top camera manufacturer in the world...but pretty much every company that just sits and rides on their past success has ultimately failed (i.e. Kodak! They were THE film camera company of the masses for decades...where are they now? Do they even hold any more patents? They are a pitiful shadow of their former shadow, let alone of their former self.) I personally really do not want to see Canon go down that route. I LOVE Canon glass...I think their lenses are second to none in all but a very few cases (less than a handful.) I don't think I could do without Canon ergonomics (that whole package deal, the body shape and size, the button placement, options, and configurability), the menu system.)

Nevertheless, I do photograph landscapes on occasion, and I would really love to have a better full frame camera with phenomenal dynamic range and resolution, along with some improved wide angle glass (a 14-24 would be nice, but I'd settle for a kick-ass 16-35mm f/2.8 III with excellent corner performance, as I've enjoyed my 16-35mm II.) I would love to be able to lift shadows and not have to apply extra work removing banding, crosshatch noise, and the sprinkle of salt and pepper noise that shows up in the really deep shadows of Canon sensors. (Although, it should be noted, it is possible to remove banding, and when Canon's banding noise is removed, the DR of their files increases considerably...maybe not the full 2.2-2.4 stops difference between a Canon file and the Nikon D800/600/810, but enough... The key difference is sometimes you have to take a shadow detail hit when denoising, which you don't have to do with a Nikon file).

I am still a big fan of my Canon equipment, and I'll always be involved in the DXO debates (which is more what this whole thread is, a debate about the validity of many of DXO's results (which apply to ALL cameras tested by DXO, not just Nikon cameras...it's just that the skew is so much greater with Nikon cameras...14+ stops of editing latitude is impossible when the files are only 14-bit) and particularly their scoring, rather than a Canon vs. Nikon thing), but I can't help but think I'll lose some faith in Canon's ability to compete if the 7D II and 5D IV hit the streets AGAIN with...not even eleven stops of dynamic range and their nasty banded read noise. Canon needs to use some of the billions they make every year to bring their still image sensor technology into the 2010s (and out of the late 1990's).

 

[unfocused]

And, I suppose the five or six people who actually can earn a living shooting large scale landscapes might want to move to Nikon

You're right about five or six people compared with wedding photographers, maybe less, but one of those is a British guy called Colin Prior. He is one of the world's best known landscape photographers and ... in mid 2013 he began using digital - FF - and Canon at that, 1Dx and 5DIII. He is on record as saying the files from the 5DIII are the cleanest he's ever come across.

Yeah, well, looking at your website, it's pretty obvious you're no slouch when it comes to landscape/scenics. Makes me think of one of my goals when I retire – contact some of the pros on this forum and ask to be a free assistant for a few weeks.

 

[arcanej]

jrista,

If you wouldn't mind, I'd love to get a walk through of your post processing. You mentioned using DeepSkyStacker and PixInsight to boost the DR of your files. I'd love to get a a step-by-step look at what you do. Thank you!

 

[Loswr]

Well, I never intended it to be offensive...

In that case, let me say that you have a blind spot when it comes to sensor technology, and it's the kind of blind spot typical of a 16 year old high school student flunking out of physics.

Hmmmm...not offensive at all, right?

...DXO scores seem to be on the money, not just with the D800, but with all the cameras I have personal experience.
The GH4 scores mimic what I expected form using it for a few weeks before the score popped up.

Are you suggesting that everyone's usage pattern mimics yours? Speaking as someone who shoots a significant proportion of my images above ISO 1600, with a fair number above ISO 6400, I can tell you that DxO's Scores absolutely do not mimic the cameras with which I have personal experience.

That doesn't particularly bother me, because I understand the general nature of the bias in their scores (even if I don't understand the specifics or the magnitude of that bias, because DxO does not disclose their formulae). But it's unfortunate that many people accept their scores as generally applicable, when that's far from the truth. It's even more unfortunate when people support them without knowing (or caring) that they are equally biased.

If there is any system justification going on, it is on this forum, where no matter what any other manufacturer does, Canon are still the best.

And few see that that's that's going on. It's not a forum, it's a fan club.

Well, it's Canon Rumors, after all...not Nikon or Sony or Photography Rumors. But still, I don't see that particular attitude very often here. What I do see are a handful of people who have decided (long ago or recently) that Canon is not the best system for them, and come here to convince others that means Canon is not the best system for anyone. Where is Nikon's PC-E 17mm? Why don't most Canon bodies (except the 1-series) allow spot metering linked to any AF point? Where is Nikon's handholdable 600mm f/4? Where is Canon's sensor with 13-stops of DR at base ISO? Where is Nikon's 1-5x super-macro lens? Where is Canon's sharp-to-the-corners f/2.8 ultrawide zoom?

A system comprises many components. Having one part of one component (the sensor in a camera, in this case) that is better than the equivalent part in a competitor's system does not make the system using that part 'the best'. Individuals make decisions about what is best for them. As I stated previously, it's an objective fact that more people have decided Canon makes the system that best meets their needs, which is why Canon has been the dSLR market leader for >10 years, and remains so today.

People who argue that Canon 'is behind' and 'needs to catch up, or else,' as you are arguing, don't seem to grasp that simple, objective fact. Does that mean Canon can do nothing and remain on top? Probably not. But consider...Nikon chose to buy Sony sensors with better low ISO DR, an issue with which relatively few people have needs that aren't met by Canon's sensors. Canon chose to develop a groundbreaking new AF system for video/live view (a system subsequently incorporated into their Cinema line)...a system you praised. I think far more people will feel that an improved AF system adds to overall system performance more than improved low ISO DR.

Canon seems to be 'skating to where the puck will be,' whereas Nikon is playing catch up. For example, did you notice that Nikon is now using fluorite elements in their supertele lenses? When only Canon used it, Nikon said fluorite "easily cracks," but now fluorite's "superb anomalous dispersion properties...effectively correct chromatic aberration," and it allows "a more effective lens with less weight," (all of those are quotes from Nikon's lens glossary). I guess Canon knew what they were doing when they first started using fluorite elements in SLR lenses...45 years ago (in fairness, Nikon has used fluorite elements for many years in their microscope objective lenses, since fluorite transmits UV better than glass, an advantage for fluorescence microscopy).

Time will tell, of course, but for the past 10 years Canon has made the right choices to drive sales, and while they're predicting slight a loss this year, Nikon is predicting a substantially bigger loss.