Nikon D5 Sensor Score from DXOMark

[ahsanford]

Apparently the D5 review is up at DXO... but it's not.

This is the link: http://www.dxomark.com/Reviews/Nikon-D5-sensor-review-A-worthy-successor

But it's a dead end.

Here it is in my newsreader, so I suppose it will be reposted shortly, FYI.

- A

 

[neuroanatomist]

A worthy successor...with less low ISO DR than the APS-C sensor in Canon's 80D.

 

[d]

A worthy successor...with less low ISO DR than the APS-C sensor in Canon's 80D.

DR is so last season....this time around it'll be all about the high ISO.

d.

 

[ahsanford]

Results now posted: http://www.dxomark.com/Reviews/Nikon-D5-sensor-review-A-worthy-successor

Get a load of this. Compared to the prior D4S and D810:

* Color depth: D810 > D5 > D4S
* Dynamic range: D810 >> D4S > D5
* High ISO: D4S > D810 > D5 (this is a bit misleading -- see the plots for this)

So:

• The D810's Sony 36 MP sensor is outperforming Nikon's 20-21 MP D5 sensor on all three of DXO's sensor metrics. Let that one soak in. I don't know if that's more of a slow clap for the D810 sensor or a Price is Right failure sound effect for the D5 sensor.
  
  
• The camera that the D5 is replacing arguably has a better sensor in it. Chasing those 4 extra MP with the D5 apparently came at a cost.
  
  
• Neuro is correct. The Canon 80D has more DR than the D5.

...And yet they entitled this article "Nikon D5 sensor review: A worthy successor". Wow.

...And they called the 'predecessor' to the D5 as being the D4 and not the D4S to possibly soften the blow that the D5 didn't get much better. Double wow.

#dxo #fairandbalanced

- A

 

[neuroanatomist]

...And yet they entitled this article "Nikon D5 sensor review: A worthy successor". Wow.

But, you know, they're not biased or anything. :

 

[neuroanatomist]

Re: Review - D5 by DXO

DPR claimed that with the D5, Nikon sacrificed some low ISO DR for better high ISO performance. Yet...DxO's measurements show the while the D5 has less DR than the D4s at low ISO, it has no more DR at high ISO...and more noise at high ISO.

DPR uses DxO's scores in their 'reviews' – I wonder how they'll handwave around these results... :

 

[Chaitanya]

Re: Review - D5 by DXO

DPR claimed that with the D5, Nikon sacrificed some low ISO DR for better high ISO performance. Yet...DxO's measurements show the while the D5 has less DR than the D4s at low ISO, it has no more DR at high ISO...and more noise at high ISO.

DPR uses DxO's scores in their 'reviews' – I wonder how they'll handwave around these results... :

Also according to DPreview statement they use the data from DXO that DXO shares out to general public. I cant find the exact page on which DPreview staff made that statement but it was regarding some cellphone review.

Also on a side note: Since DXO has compared Leica SL as well in their D5 review, two weeks back when I was watching the Formula 1 Spanish GP, during the prerace show I saw some photographers present on the grid who were using that camera.

 

[ahsanford]

Re: Review - D5 by DXO

DPR claimed that with the D5, Nikon sacrificed some low ISO DR for better high ISO performance. Yet...DxO's measurements show the while the D5 has less DR than the D4s at low ISO, it has no more DR at high ISO...and more noise at high ISO.

DPR uses DxO's scores in their 'reviews' – I wonder how they'll handwave around these results... :

This (below) is what they are talking about, I believe. Much like how the 24-70 f/2.8 VR they put out seems to deprioritize center resolution to do better away from center than you'd expect, the sensor seems to have been tweaked to not lose a DR as quickly from ISO 1600 - 12800. It's an atypical plot shape for a sensor's DR.

But, as you noted, the noise is higher even if they saved some higher ISO DR from falling off. That's a really nuanced line to walk.

- A

 

[Mt Spokane Photography]

The DXO sensor score of 88 for the D5 tell us that their measurements favor a sensor with high DR and High MP by a wide margin. Their score gives weighting to the different parameters and those weightings are assigned by people who may not be actual money earning or award winning photographers. They could have experts help assign those values and make them public, but they keep them secret and the end results puzzle us. Their values may not match mine.

The same is true of the tests. They have devised their own test methods, ones that are not accepted by the photographic industry, believing that their tests are better. This means that the individual tests they post have little or no correlation with results from other testers. Any similarities are accidental. Don't expect this to change, its somewhat of a cultural thing that I saw in my job of evaluating components and tests from products around the world as part of my previous job. I had the ability to force would be suppliers to use standardized industry approved tests, or their would not be approved. Those who wanted to invent their own tests that favored their products or way of thinking hated that. They came from that same country (Good and smart people, but wanting to do things differently.)

Fortunately, Nikon has ignored their scoring, because the D5 is not aimed at those who need high DR or high MP. The camera sensor just barely beats out the 4 year old 1D X. That makes the 1DX somewhat of a bargain.

Of course, people think that DXO is rating a camera. They are not, they are only rating the sensor. A rated camera may have poor autofocus, or lots of shortcomings, but get a top score. I think that as a overall camera, the D5 is very good, as are Canon's top of the line products. I'd have no issues using either Canon or Nikon top of the line cameras.

 

[ahsanford]

Of course, people think that DXO is rating a camera. They are not, they are only rating the sensor....

100% understood. But if we drink the Kool-Aid that DXO is 'directionally accurate' (even if methods are somewhat flawed, disregard the batsh-- overall score and focus on the individual tests) -- just for a hypothetical argument -- I think the following statements could roughly be made:

1) Gripped top-end rigs no longer are the 'best of the best' sensors. Lower cost + higher MP rigs seem to be outperforming the IQ of the top end rigs right now. I appreciate that a D5 or 1DX can do a ton more than a 5D3 or D810 can, but at a pure IQ level they are not necessarily the top dog like they used to be. This undermines the gripped-rig value proposition slightly.

2) Nikon users might be correct in grilling Nikon corporate for who is making the sensor in any new offerings. EXMOR seems to clearly outperform over Nikon sensors yet again here.

Fair statements? Disagree?

- A

 

[zim]

The DXO sensor score of 88 for the D5 tell us that their measurements favor a sensor with high DR and High MP by a wide margin. Their score gives weighting to the different parameters and those weightings are assigned by people who may not be actual money earning or award winning photographers. They could have experts help assign those values and make them public, but they keep them secret and the end results puzzle us. Their values may not match mine.

The same is true of the tests. They have devised their own test methods, ones that are not accepted by the photographic industry, believing that their tests are better. This means that the individual tests they post have little or no correlation with results from other testers. Any similarities are accidental. Don't expect this to change, its somewhat of a cultural thing that I saw in my job of evaluating components and tests from products around the world as part of my previous job. I had the ability to force would be suppliers to use standardized industry approved tests, or their would not be approved. Those who wanted to invent their own tests that favored their products or way of thinking hated that. They came from that same country (Good and smart people, but wanting to do things differently.)

Fortunately, Nikon has ignored their scoring, because the D5 is not aimed at those who need high DR or high MP. The camera sensor just barely beats out the 4 year old 1D X. That makes the 1DX somewhat of a bargain.

Of course, people think that DXO is rating a camera. They are not, they are only rating the sensor. A rated camera may have poor autofocus, or lots of shortcomings, but get a top score. I think that as a overall camera, the D5 is very good, as are Canon's top of the line products. I'd have no issues using either Canon or Nikon top of the line cameras.

very well said

 

[East Wind Photography]

Of course, people think that DXO is rating a camera. They are not, they are only rating the sensor....

100% understood. But if we drink the Kool-Aid that DXO is 'directionally accurate' (even if methods are somewhat flawed, disregard the batsh-- overall score and focus on the individual tests) -- just for a hypothetical argument -- I think the following statements could roughly be made:

1) Gripped top-end rigs no longer are the 'best of the best' sensors. Lower cost + higher MP rigs seem to be outperforming the IQ of the top end rigs right now. I appreciate that a D5 or 1DX can do a ton more than a 5D3 or D810 can, but at a pure IQ level they are not necessarily the top dog like they used to be. This undermines the gripped-rig value proposition slightly.

2) Nikon users might be correct in grilling Nikon corporate for who is making the sensor in any new offerings. EXMOR seems to clearly outperform over Nikon sensors yet again here.

Fair statements? Disagree?

- A

Don't agree with the undermining of the gripped rig value proposition. If you don't need a gripped rig system but only prefer better sensor performance then buy something for that purpose. I think you guys are splitting hairs on the performance issue. Valuation is better weighed on the system and not just the sensor.

I can't shoot sports with an electronically cooled CCD sensor fixed to a manually focused lens....though it might be fun to try.

 

[unfocused]

I quit paying much attention to DXO testing years ago.

I look at the graphs and consistently see only tiny differences in the lines. I know people on both sides get worked up about their tests, but frankly, I would never make a buying decision based on the insignificant differences their tests show.

 

[Mt Spokane Photography]

I quit paying much attention to DXO testing years ago.

I look at the graphs and consistently see only tiny differences in the lines. I know people on both sides get worked up about their tests, but frankly, I would never make a buying decision based on the insignificant differences their tests show.

That is certainly my way of thinking. I've made the mistake of purchasing cameras based on reviews that were flawed more than once (you'd think I'd learn )

I tether my camera to a computer to photograph my products on a light table. I focus and set the aperture to get the depth of field I want, or put the feature I want in sharp focus. I can change the angle or take other measures to eliminate unwanted reflections, all in near real time with every Canon XD or XXD since the 40D, and they all work great.

I bought a D800, it was horrible to tether and remote control, reduced resolution so you could not focus on fine details, it was so slow as to be unusable. I also had a D300S, it was worse. I don't know if current Nikon models work any better, I've tried to find good reviews, but they seem to be lacking. I'd think the A7R II being a mirrorless would tether well, but the few reviews I've seen hinted that it is not good at tethering for remote operation.

I stick with Canon now because I know the remote tethering just works.

 

[3kramd5]

Re: Review - D5 by DXO

DPR claimed that with the D5, Nikon sacrificed some low ISO DR for better high ISO performance.

I still don't get how that works.

I suspect that other trades (e.g. framerate) impacted read noise, and at the same time improvements were made to high ISO.

Coincidence, not causation (sacrifice).

 

[MigueEsteves]

I still have my 300D working, better than the D5 everyday ;D

 

[Yiannis A - Greece]

DXOMark D5 crappy results...

Dear friends,

i just finished observing DXOMark review of the Nikon D5 and it's finally "official" that, D5 scores worse than it's Nikon predecessors (D4, D4s) in every single test apart from bit depth! 1+ stop worse dynamic range, 18-25% less high ISO image quality (in fact, the worst FF sensor since D3x) and a total score of only...88 (for a heavily biased Nikon site)! Even Canons' "humble", beloved, "good old" 1Dx Mk1 (i personally own a pair of 5D Mk3s among others) achieves a 2786 ISO score!

So, how the hell do they call the D5 "A worthy successor"??? Why should somebody with a D4 or d4s should even think of buying one? well, maybe, for 3 minutes of heavily cropped 4K footage with lousy continuous AF!!!

I'm pretty sure that 1Dx Mk2 is going to s**t all over D5 in an upcoming review; although i'm more than sure that DXO will give Mk2 an overall score of...87 ;D

Here is the link to the review, have fun:

http://www.dxomark.com/Reviews/Nikon-D5-sensor-review-A-worthy-successor

As always, all my best wishes for you and your beloved. Have a nice evening wherever you are.

Yours
Yiannis - Athens, Greece.

 

[StudentOfLight]

I wonder if the ISO score took a knock because they overstated their ISO more than on previous models. According to DXO, the D5 is only producing an image of brightness equivalent to ISO 66, as opposed to ISO 80 on the 1DX, when the cameras are both set to ISO 100. (p.s. The D4s was ISO 75)

The D5 Sports ISO score appears to be failing the SNR threshold first. This is in contrast to what DPR studio samples presented. Does this mean that:
a) the D5 RAW are cooked with some added NR spices at higher ISOs or
b) the RAW converter is helping the D5 more than older cameras or
c) Other (explain)

 

[Skatol]

Found the conclusion most humorous.
Did not realize the 1DX MII is considered old now.

"Conclusion

Moving from a 16-Mpix to a 20.8-Mpix sensor is quite significant for professionals, and the improvements in high ISO DR are useful in a camera like this. In terms of sensor resolution, the Nikon D5 now matches the admittedly aging Canon EOS 1Ds Mk II, a camera still widely used for magazine features (think double-page spreads) as well as commercial photography. While it’s true that models such as the Nikon D810 (and the Canon EOS 5DS and SR) are arguably more suited now for that kind of work, the Nikon D5 nonetheless represents a tipping point in all-around capability. With its excellent sensor performance, phenomenal AF system, impressive burst rate and durable build, the Nikon D5 is clearly optimized to appeal to those targeted by the company’s marketing department, yet its capabilities should also entice other kinds of photographers, whether they own a Nikon D810 or not."

 

[neuroanatomist]

Found the conclusion most humorous.
Did not realize the 1DX MII is considered old now.

"Conclusion

Moving from a 16-Mpix to a 20.8-Mpix sensor is quite significant for professionals, and the improvements in high ISO DR are useful in a camera like this. In terms of sensor resolution, the Nikon D5 now matches the admittedly aging Canon EOS 1Ds Mk II, a camera still widely used for magazine features (think double-page spreads) as well as commercial photography. While it’s true that models such as the Nikon D810 (and the Canon EOS 5DS and SR) are arguably more suited now for that kind of work, the Nikon D5 nonetheless represents a tipping point in all-around capability. With its excellent sensor performance, phenomenal AF system, impressive burst rate and durable build, the Nikon D5 is clearly optimized to appeal to those targeted by the company’s marketing department, yet its capabilities should also entice other kinds of photographers, whether they own a Nikon D810 or not."

I expect they meant the 1Ds Mk III, which has a 21.1 MP sensor and was launched in 2008.

(Because, you know, comparing the new 20 MP FF D5 to an 8 year old 21 MP FF 1-series camera makes so much more sense than comparing it to a the new 20 MP FF 1-series body...)