DXOMark review: D850 gets a *100* sensor score

[ahsanford]

I'll just put this here and see what happens:

https://petapixel.com/2017/10/06/nikon-d850-best-dslr-ever-gets-first-full-100-score-dxomark/

- A

 

[BasXcanon]

Hooray for Nikon!
My baby album was made by Nikon gear.
I actually want a Nikon camera, but everytime I have gathered enough money I end up with a Canon product. :-\

 

[unfocused]

I find it confusing that it scored a perfect 100, yet the review discusses several areas where it is less than perfect and the data indicates that other cameras either beat or matched its performance on particular criteria.

Does this mean DXO will soon be turning the amplifier up to "11."

 

[Jopa]

I just smashed all my Canon gear with a sledgehammer!!! It has no future.

 

[neuroanatomist]

I find it confusing that it scored a perfect 100, yet the review discusses several areas where it is less than perfect and the data indicates that other cameras either beat or matched its performance on particular criteria.

Does this mean DXO will soon be turning the amplifier up to "11."

DxO's Sensor Score is an open-ended metric (albeit one determined by a 'black box' algorithm, which IMO makes it a steaming pile of cow feces, aka BS = biased score). Thus, a score of '100' is not perfect, by their own description of the metric it can (and likely will) go higher as sensor technology improves.

 

[ahsanford]

I find it confusing that it scored a perfect 100, yet the review discusses several areas where it is less than perfect and the data indicates that other cameras either beat or matched its performance on particular criteria.

Does this mean DXO will soon be turning the amplifier up to "11."

100 isn't perfect. It's just a number. If memory serves, prior EPIC cameras and a medium format rig have scored over 100 before. This is just the first FF SLR to hit that mark.

- A

 

[Talys]

I posted this on the other thread, but I think it's more appropriate here:

Why does the Low Light score suck?

2995 - Canon 5D Mark IV
2956 - Nikon D750
2862 - Canon 6D Mark II
2853 - Nikon D810
2660 - Nikon D850

After ISO 800 D850 has less DR, Tonal Range, and Color Sensitivity than 5DIV on the Screen measurements (it scores better on a bunch of Print measurements).


So anyways, since it is not a perfect sensor, by DXO's evaluation... does a future sensor that is better than the D850 score > 100?

 

[zim]

Jopa wins a doller ;D

 

[AlanF]

I posted this on the other thread, but I think it's more appropriate here:

Why does the Low Light score suck?

2995 - Canon 5D Mark IV
2956 - Nikon D750
2862 - Canon 6D Mark II
2853 - Nikon D810
2660 - Nikon D850

After ISO 800 D850 has less DR, Tonal Range, and Color Sensitivity than 5DIV on the Screen measurements (it scores better on a bunch of Print measurements).


So anyways, since it is not a perfect sensor, by DXO's evaluation... does a future sensor that is better than the D850 score > 100?

"Why does the Low Light score suck?" The fair comparison with Low Light Score is with a sensor with the same pixel density, namely the 5DSR, which has 2308. The 850 doesn't "suck" against that.

I use both the 5DIV and 5DSR, but I am not going to rubbish the D850 out of an irrational sense of tribal loyalty or to justify my purchases. I use the 5DIV when I want speed and for lower light, and the 5DSR when I want the best resolution. For most of the time I use the 5DSR at iso640 and above, where there is little to choose between the DR of the 5DSR and D850. I would love Canon to be spurred into action to produce an equivalent of the D850 because it is a magnificent piece of engineering. So come on Canon, make a 5DSR Mk II that regains the number 1 spot.

 

[Talys]

I posted this on the other thread, but I think it's more appropriate here:

Why does the Low Light score suck?

2995 - Canon 5D Mark IV
2956 - Nikon D750
2862 - Canon 6D Mark II
2853 - Nikon D810
2660 - Nikon D850

After ISO 800 D850 has less DR, Tonal Range, and Color Sensitivity than 5DIV on the Screen measurements (it scores better on a bunch of Print measurements).


So anyways, since it is not a perfect sensor, by DXO's evaluation... does a future sensor that is better than the D850 score > 100?

"Why does the Low Light score suck?" The fair comparison with Low Light Score is with a sensor with the same pixel density, namely the 5DSR, which has 2308. The 850 doesn't "suck" against that.

I use both the 5DIV and 5DSR, but I am not going to rubbish the D850 out of an irrational sense of tribal loyalty or to justify my purchases. I use the 5DIV when I want speed and for lower light, and the 5DSR when I want the best resolution. For most of the time I use the 5DSR at iso640 and above, where there is little to choose between the DR of the 5DSR and D850. I would love Canon to be spurred into action to produce an equivalent of the D850 because it is a magnificent piece of engineering. So come on Canon, make a 5DSR Mk II that regains the number 1 spot.

I get that the traditional calculus is that more megapixels = worse low light performance, all things being equal -- but all things are not always equal. The Sony A7RII has an amazing low light performance score of 3434 at almost as many megapixels as the D850.

Both are more megapixels than I'd use, but frankly, I'd rather have 42.4 megapixels at 3434 low light score than 2660 at 45.7 megapixels.

The reason that the D850 has such high marks (its 100 score) is the 14.8EVs of dynamic range.

Comparing just the A7RII to the D850 sensors, and excluding everything else (not that I would buy a camera that way), if I had to choose between 1 extra EV of DR at ISO 100 (to 15) and proportionately better ISO performance at 800+, I would pick the better ISO performance every day of the week.

The chances of shooting photos that use the extra 1 EV are probably infinitely small for me compared to the chances I'd shoot a photo at ISO 1000.

Of course, what everyone is looking for is different, and if I had to choose between the Sony and the Nikon, there would be no contest, because one of them is crushingly uncomfortable to use and is missing an optical viewfinder. But still, looking at just sensor scores, and untangling the "why", I still don't understand why the Nikon sensor made by Sony so underperforms the older, but similar Sony sensor made by Sony in low light.

 

[AlanF]

By that reasoning, the 26.2 mpx 6D II with its low light score of 2862 does "suck" compared with the Sony 42.4 mpx with its low light score of 3434. Mind you, I don't take DxOMark's scores seriously because, apart from everything else, noise in practice depends on the lenses available, degree of cropping etc.

 

[mb66energy]

Seems to be a very good camera providing very high IQ, good ergonomics including backlit dials etc.. Finally it seems they have good control of colors with that camera.

Otherwise: I have 4k invested in Canon glass (not es much as maybe most of you but enough) and enjoy the 200D / SL2 for "slow photography" with the excellent EF-S 60 Macro as a light and compact package with great IQ and flexibility. Good enough for 60 x 90 cm² prints - if I made everything right.

 

[Talys]

By that reasoning, the 26.2 mpx 6D II with its low light score of 2862 does "suck" compared with the Sony 42.4 mpx with its low light score of 3434. Mind you, I don't take DxOMark's scores seriously because, apart from everything else, noise in practice depends on the lenses available, degree of cropping etc.

I have no problem conceding that the 6DII's sensor is significantly inferior to the Sony A7RII's sensor. Would I like an A7RII's sensor in a Canon body? Of course!

However, the Canon 6DII is superior in other ways, like: a much lower price, a viewfinder that I like 1000x more, ergonomics that don't kill me, and a vastly superior selection and quality of lenses.

If I were willing to pay the price of a A7RII, there is no question that I would enjoy the D850 or the 5DMkIV more than the A7RII. However, on the basis of scoring sensors, I can't see how nearly a thousand points in low light performance out of 2500-3500 is not a huge deficiency, at least as much as a couple of points of base ISO DR, because while most people rarely or never photograph 15EVs, most people will find a use for ISO 800+.

I would probably not have said anything if DXO gave the D850 an identical overall sensor score to the Sony A7RII.

I did consider buying an A7RII, solely on the basis of its excellent sensor. But the other shortcomings killed it. I would consider switching back to Nikon, but it would really take Nikon being "on top" in a combination of lenses I need or am likely to buy and camera bodies for about 10 years.

At the moment, there's no question that the D750 and D850 are really great bodies. But I still vastly prefer the Canon glass (I vastly prefer Canon's 70-200, 100-400, and dream of Canon's 600/4, while I do not lust for any Nikon Super Telephotos), and that's a much longer term investment, and I have gotten very used to Canon controls. Plus, in the past, Nikon has had QA issues and that's a little hard to get past. There's also the issue that Nikon doesn't service grey market equipment (they won't fix it even for a fee), whereas Canon will service it under warranty -- that makes a difference to me, because I have no problem buying grey market.

 

[AlanF]

There is no Nikon glass I would like for bird photography. I don't go for the 500/4 and 600/4 even for Canon as they are too heavy for me. The Nikom 200-500mm is too heavy, and several reports prefer the Tamron 150-600mm (it zooms in the same rotation as Nikon) and the 80-400mm is not a good lens. I do not know first hand but I have read that Nikon's don't have C1, C2 and C3 on the mode dial that store settings faithfully. Please correct me if I am wrong. I do want a Canon with the D850 specs.

 

[Ryananthony]

There is no Nikon glass I would like for bird photography. I don't go for the 500/4 and 600/4 even for Canon as they are too heavy for me. The Nikom 200-500mm is too heavy, and several reports prefer the Tamron 150-600mm (it zooms in the same rotation as Nikon) and the 80-400mm is not a good lens. I do not know first hand but I have read that Nikon's don't have C1, C2 and C3 on the mode dial that store settings faithfully. Please correct me if I am wrong. I do want a Canon with the D850 specs.

Nikon also doesn't seem to have ''Recall shooting functions'' assignable to a dedicated button like canon and by the looks of it Sony as well. This has become an amazing feature that would make it difficult for me to live with out.

 

[Talys]

There is no Nikon glass I would like for bird photography. I don't go for the 500/4 and 600/4 even for Canon as they are too heavy for me. The Nikom 200-500mm is too heavy, and several reports prefer the Tamron 150-600mm (it zooms in the same rotation as Nikon) and the 80-400mm is not a good lens. I do not know first hand but I have read that Nikon's don't have C1, C2 and C3 on the mode dial that store settings faithfully. Please correct me if I am wrong. I do want a Canon with the D850 specs.

Yes, you've described the logical conclusion that I've drawn as the "intellectual" reason to choose Canon.

Even though I don't own a big f/4, I'd like to own one, one day, and even though I expect I will use a gimbal for that, it's okay; I'd much rather the canon 600 f/4 than the Nikon 600 f/4 (for weight reasons). On lens under $3,000 USD, the 200-500 appeals on paper, but is much less friendly to carry around and shoot stuff with than Canon's 100-400.

It doesn't really matter if the D850 is super duper awesome, if there aren't matching, intermediate telephoto friendly to carry around. At the moment, frankly, the third party 100-400's are the best choice for carry-around, and I far prefer Canon's because of ease of MF, consistent autofocus, and MFD.

And sure, I, too, would love to see D850 sensor specs in a Canon body, though to be frank, I probably wouldn't spring for it, as it may be over the price range I like to spend on bodies. I am much more willing to invest in lens, because of the lifespan where they stay in their "prime" (ie when they aren't superceded by a better lens choice).

The less intellectual reason that I choose Canon is just one of preference. I like Canon cameras. And by the way, I would never buy another camera that didn't have custom dial functions like C1. It's like buying a car without a cupholder.

 

[Hflm]

I find it confusing that it scored a perfect 100, yet the review discusses several areas where it is less than perfect and the data indicates that other cameras either beat or matched its performance on particular criteria.

Does this mean DXO will soon be turning the amplifier up to "11."

100 isn't perfect. It's just a number. If memory serves, prior EPIC cameras and a medium format rig have scored over 100 before. This is just the first FF SLR to hit that mark.

- A

I can't understand, why people are so concerned with an arbitrary score, weighted to base iso DR.

If I look at the A7rii, for example and compare the measurement curves, apart from low iso dr, the A7rii is as good or better over the whole iso range in tonality, color sensitivity, S/N ratio (the D850 is 1db behind the A7rii or 5div). That alone shows how useless a single number is for a complex system, which many here know. But the human strive for simplifying and rating everything doesn't stop many from thinking such a score to be important.

 

[Hector1970]

Well done Nikon on getting a *100* Sensor Score.
They have created a great camera in the D850.
It ticks alot of boxes for users. They pushed the boat out quite a bit with it.
I think it ensures current Nikon users will stay with the brand and they needed a confidence boost.
It's a camera that gives them hope for the future that Nikon can continue to compete and even beat the opposition.
I'm heavily invested in Canon. Their lens are excellent and their cameras are good.
I think alot users like to feel they have the best brand even if they haven't the best version themselves.
It's why companies create flagship models.
They often don't expect to sell that many of them but it gives people something to aspire too and a confidence to stick with the brand as it shows what they are capable of. Canon need to remember this from time to time and not rest on their laurels. Other than the 1DX in recent years you don't feel they are making the best camera they could. The 5D IV made sure it wasn't too far behind anything else but Canon certainly didn't push the boat out. Still a fine camera but not a dream upgrade from a 5DIII.
The better Nikon and Sony cameras are the better Canon cameras will need to be to attract new users to the brand. This in the end is good for a user like me.
While people here feel DXO ratings are not valid they are at least a method to judge. It does influence buyers new to the market and its a great marketing tool for Nikon. So for me well done again Nikon.

 

[AlanF]

There is no Nikon glass I would like for bird photography. I don't go for the 500/4 and 600/4 even for Canon as they are too heavy for me. The Nikom 200-500mm is too heavy, and several reports prefer the Tamron 150-600mm (it zooms in the same rotation as Nikon) and the 80-400mm is not a good lens. I do not know first hand but I have read that Nikon's don't have C1, C2 and C3 on the mode dial that store settings faithfully. Please correct me if I am wrong. I do want a Canon with the D850 specs.

Yes, you've described the logical conclusion that I've drawn as the "intellectual" reason to choose Canon.

Even though I don't own a big f/4, I'd like to own one, one day, and even though I expect I will use a gimbal for that, it's okay; I'd much rather the canon 600 f/4 than the Nikon 600 f/4 (for weight reasons). On lens under $3,000 USD, the 200-500 appeals on paper, but is much less friendly to carry around and shoot stuff with than Canon's 100-400.

It doesn't really matter if the D850 is super duper awesome, if there aren't matching, intermediate telephoto friendly to carry around. At the moment, frankly, the third party 100-400's are the best choice for carry-around, and I far prefer Canon's because of ease of MF, consistent autofocus, and MFD.

And sure, I, too, would love to see D850 sensor specs in a Canon body, though to be frank, I probably wouldn't spring for it, as it may be over the price range I like to spend on bodies. I am much more willing to invest in lens, because of the lifespan where they stay in their "prime" (ie when they aren't superceded by a better lens choice).

The less intellectual reason that I choose Canon is just one of preference. I like Canon cameras. And by the way, I would never buy another camera that didn't have custom dial functions like C1. It's like buying a car without a cupholder.

The Nikon AF-S NIKKOR 600mm f/4E FL ED VR is actually lighter than the Canon EF 600mm f/4 II; 3,810g vs 3,924g. You might have got the weight information from TDP, which has just the older Nikon 600mm f/4G AF-S VR and hasn't updated it.

 

[Talys]

The Nikon AF-S NIKKOR 600mm f/4E FL ED VR is actually lighter than the Canon EF 600mm f/4 II; 3,810g vs 3,924g. You might have got the weight information from TDP, which has just the older Nikon 600mm f/4G AF-S VR and hasn't updated it.

You're right -- I wasn't even aware that Nikon had refreshed their 600mm f/4. The last time I looked at reviews of Nikon 600mm f4's was around the time I switched (from Nikon) to Canon, a very long time ago, I guess, now that I think about it. I've only occasionally flirted with the idea of buying a $10,000 lens, so I don't exactly follow those.