Review of D850 from DPReview.com is out, but. . .

Discussion in 'Nikon' started by Landrum Kelly, Oct 18, 2017.

  1. Of course, there is no particular reason why packing more megapixels should give better results at high ISO and low light. Everything else being equal, it won't!

    As for those who do want truly clean results with 45 megapixels, there are always tripods and time exposures--so I am sure that the D850 really is a winner all the way around. As a landscape shooter, I don't need its kind of tracking and autofocus, but others might.

    I will buy one myself when I can afford it--for the resolution at lower ISOs. I still don't understand the hyperbole, though. It's got plenty of strengths without the exaggerations.

    --Lannie
     
  2. Ilkka: I'm happy to hear from people who might have considered the D850 but decide it's not worth the upgrade for them. Some will have different cameras from me, most will have different requirements, so I'm interested to learn where other cameras solve problems well enough. I also like being in a position to advise others on what to buy (be they on this forum or my friends), and the more perspectives I get, the better. As with the D810, the D850 probably won't be the right camera for most people, especially at its current price. I've no interest in buying a D750 replacement, but I'll be interested to know what Nikon release for similar reasons.

    Lannie: While I did read the dpreview article as "buy this camera" (as coloured by my eyes), I don't think they were being all that glowing. The speckles they report did worry me a bit, and they made what I consider to be a fairly nonsense complaint about cameras not recording raw at the nearest amplifier ISO and leaving the rest to the decoder. (Well, it's not that nonsense, I'm sympathetic, but I wouldn't particularly call out the D850 for doing it.) It's worth calling out the slightly strange appearance of the crop shown, and I'm happy to calm down anyone with unreasonable expectations of the D850 - they'd only be disappointed if they got it.

    If we want to criticise, can make a plenty long list of change requests even from just reading the manual. (The same applies to most of the camera industry, possibly starting with a question about whether it's really ergonomically best to have the grip facing forwards given the typical articulation of the human arm...) It's a camera, no more and no less. And Hypnoken will probably claim it's still not as "good" as a Mamiya 7.
     
  3. The camera definitely delivers in ordinary shooting:

    [link]

    I think that it's fair to say that it would be a great and versatile all-around camera. It could handle just about any situation quite well. There's a lot to be said for that.

    Do I contradict myself? Well, then, I contradict myself. It was not quite fair of me to call it a "niche camera," or, if so, it's a pretty sizable niche that it fills.

    --Lannie
     
    Last edited: Oct 19, 2017
  4. I only see worse or at best equal performance; certainly not "almost one stop better".
    Sorry, but you can't state that the differences are tiny and in the same sentence give an advantage to one camera. I care mostly about the 1:1 view (that would be "Screen" in dxomark speak), and there the D810-to-D800 SNR differences are indeed tiny while the D850 numbers are always below those of the D810/D800: Nikon D850 vs Nikon D810 vs Nikon D800 | DxOMark. IMHO the "Print" view shouldn't be discarded outright but it mostly reflects that more MP "collapsed' onto a common size always favors the higher MP camera. As the data almost fall on top of each other, there's no advantage or disadvantage to be claimed for either D800, D810, or D850. I actually suspect (but don't know) that the error margins for each measured data point are actually larger than the differences between cameras.
    Exactly! Certainly a far cry from the control layout differences between the D8x0 and the D750, for example. Or between a Nikon DSLR and a Sony A7 Series mirrorless. Differences between bodies are always somewhat annoying, but I am sure there are also some between the D850 and the D500 (like the missing d9 AF area mode on the D500). I have by now essentially found a mode of operation with my Nikon DSLRs that minimizes the number of changes I have to do and luckily the one button I need most is at the same spot for years now (exposure compensation). I haven't actually checked but maybe it is even possible to program the ISO button on the D500 to function as the MODE button and the REC button to be ISO - which would make the arrangement the same as on the D810.
    Heaven forbid that I may actually learn something! I am NOT in the market for a D850 NOW, which doesn't mean I am not interested to see what it has to offer or am not interested in buying one EVER. Maybe something I learn here will prompt me to reconsider? One major reason I am not currently interested is the simple fact that a Nikon mirrorless is on the horizon and I want to wait and see what it has to offer.
     
    Last edited: Oct 19, 2017
    paul_b.|1 likes this.
  5. I have a D810 for studio and/or tripod work and a D500 for fast sport.

    I'm guessing Nikon didn't try to make a FX blend between the two, although they will now share certain abilities and features..... but with a bulky grip, it gets kinda close on frame rate.

    There's no way I could afford a D5, so to get the frame rate, a Pro DX body was 'good enough'., but i do lose about a stop in ISO. However, I use a Sigma 50-100mm 1.8, which works brilliantly at f2, so get the stop back compared to my Nikon 70-200mm VRII.

    Do I trade in my D810 and my D500 for a D850 and grip? I don't think so.

    Do I trade in my D3S +70-200mm VRII sports combo for some better prime FX glass, probably!

    My D3S has 487000 frames and although it hasn't missed a beat, apart from intermittent odd Auto-ISO behaviour, has gotta be knocking at the shutter failure door.

    A 45MP D850 is as accomplished, IQ wise, as a 12MP D3S. despite smaller pixels and higher density. 5 years has seen a lot of progress. My D3S never made a good studio cam as it just didn't really have the DR (or the pixels) of a more modern sensor in good light.

    All Nikon single digit cams go fast, some very fast, but at the expense of DR, especially at low ISO, and that remains the same between the New D850 and the D5. I clear case of horses for courses.

    However, the sheer data throughput of 45MP at 9 frames/sec is pretty amazing
     
    Last edited: Oct 19, 2017
  6. I appreciate the fact that you pointed out that that Amazon is behind DPReview. I did not know that, I only thought they sponsored the site. To be fair, I now have a harder time understanding one part of the D850 review on DPReview; the XQD card is considered a necessary evil for those wanting speed rather than something the user might already have or be willing to buy anyway in order to maximize transfer speed. The reviewer is not all that positive regarding them and the XQD is listed in the Cons in the summary. I assume Amazon sell both XQD cards and card readers, so why is that novelty not hyped then?

    All in all, I look forward to your upcoming review. I also hope it will include a comparison on the 1.2 and DX crop modes for when you need the added reach, say for wildlife (IQ at both base ISO and higher ISO settings).
     
  7. > A 45MP D850 is as accomplished, IQ wise, as a 12MP D3S.

    I hadn't checked that comparison, but I'd be pleased. I always vaguely lusted after the low-light capabilities of the D3s. (Yes, a D5 is better, and they're both faster than the D850 unless you go to 8MP fixed-focus JPEG mode, but it removes some of my floating NAS.)

    There seems to be a consensus: if you already have a D8x0 and a D500, adding a D850 is quite a tough sell. Having had a D700 alongside my D800 and barely used it, I no longer have a back-up body alongside my D810. I'm less inclined to pick up a D500 at this point - I'd sooner trade in for the D850, but I can see how others make a different trade-off.

    I would quite like a back-up body, once I'm feeling more flush with finance. I'll report back whether that's a D3x00 series, a beater used D800, or whatever I might hope Nikon releases to replace the D610.
     
  8. I'd like to think the DPReview staff retain some journalistic integrity. They may be under some pressure not to hurt their sugar daddy, but I don't think they're entirely obliged to push sales. Or maybe, for once in my life, I'm not being cynical.

    Just to be clear: as with any Nikon body, a crop mode on the D850 doesn't add any reach. All it does is throw away the pixels around the image border, as though you'd cropped the image in post-processing. This increases the buffer size and the number of images you can fit on a storage card (because you're storing fewer pixels the files are smaller and take up less buffer space), and on lesser bodies it may increase the frame rate (there's less data to read off the sensor, although this doesn't seem to be the limiting factor on the D850), but you don't gain any detail when you crop.
     
  9. ShunCheung

    ShunCheung Moderator

    Phil Askey founded DPReview in London, UK back in 1998, and he sold that to Amazon in 2007. In 2010 Amazon relocated DPReview from London to Seattle where Amazon's headquarter is. See this Wikipedia article: Digital Photography Review - Wikipedia

    If you clink on any purchase links on DPReview's own content, it will lead to Amazon. (Obviously people can post other links to their forums.) And you will not find advertisements from Adorama, B&H, etc. on DPReview. Obviously DPReview still has a lot of good information (as well as bad information on their forums), but to me, DPReview is more like a marketing arm for Amazon.

    By the same token, Nikon USA is loaning me equipment for review to promote their products. While I don't get paid by Nikon, I do try to maintain a good relationship with them. Meanwhile, I also would like to maintain a good reputation and provide fair comments on Nikon products.
     
    Landrum Kelly likes this.
  10. Pixel-level comparisons of SNR do not really have a clear meaning in applications. If you shoot a subject from a certain position with D810 and D850 and the same lens and focal length, and compare individual pixels, the image content in those pixels is different so a numerical comparison between SNR at the pixel level is not the correct way to do it. What you need to do is compare areas of equal image content. And that is what the Print comparison does.

    The D5 has better dynamic range than most FX cameras at high ISO. Dynamic range is to me more important at high ISO than low ISO since in high ISO applications ”good enough” image quality is not a given and often lighting indoors is highly contrasty (light bulbs and then lots of darkness). Color correction is often needed by boosting the blue and as that channel has the lowest innate sensitivity, the noise can easily become offensive especially after color correction. Having excellent DR at high ISO does help in such situations. In landscape and real estate photography one can often take two or more exposures and combine the images with appropriate masking, something that won’t work for documentary shots indoors at night.
     
    Last edited: Oct 19, 2017
  11. I am fully aware of what the DX crop mode does and its usability for wildlife and other small subjects (macro, to name one). I do not believe than many reading this thread thinks the DX mode adds detail, just that it saves you time in post - and memory cards in the field. The 24 Mpix FX cameras are limited to about 10 Mpix in DX crop mode and photos show quite a bit of grain, too much to really be useful to me. The D800/810 DX crop roughly correspond to a D7000 and it starts to become usable. As the DX crop on the D850 is similar to the resolution of the D500, I strongly believe that more photographers than before will use it every now and then for birds and other small subjects.

    I know I can download RAW test files from various D850 tests, crop them myself to DX and have a look, but I as I know Shun likes to shoot birds I wanted to ask him for such a comparison.
     
  12. For anyone coming back to this after a post edit (apologies for posting screen grabs from dpreview - I wasn't sure if that was okay, and apparently it isn't)... I see substantially more low-frequency chroma noise at the same ISO from the D810 and D800e than the D850 at the same ISO in the dpreview image, particularly in relatively monochrome regions, for example around the lock of hair. I don't think the D850 is as good at ISO 6400 as the D800e is at ISO 3200 (at least at "screen" size) but it's close. Other areas of the image show different behaviour; that's not unusual, and we saw similar variations when comparing images from the D7000 and D700.

    As Ilkka says, I'd expect the "print" (normalised image size) comparison to be more meaningful in most situations unless you're actually trying to determine the merits of increased resolution. By the measure of "screen" (1:1 pixel comparison), the D700 is highly competitive with any modern Nikon on image quality, and I'm afraid I don't buy that. I'm curious to know why Dieter values this style of comparison. With the DxO comparison I was really trying to argue that the D850 is no worse (in the print view), which I'd read some other comments as indicating, not that a negligible improvement was significant. I'll take the extra dynamic range at higher ISO though, even if it's not as big a benefit as a D5 or A7R2. I want the low ISO benefits too, so the D850 is the right trade-off for me currently, although I won't turn down a D5 if offered.

    Sorry, Heimbrandt - just making sure, and clarifying for any less-experienced photographers reading this thread. Your reference to "added reach" (rather than, say, "reduced field of view") made me think it was worth checking. I agree that the DX crop of the D850 is a lot more appealing than that of the D810 (and, especially, D700); there are extra sensor electronics in the way on the FX body (similarly if you compare a D7000 with a D800), so I wouldn't necessarily expect the D850 to keep up at the same ISO, but it should be close; I, too, am curious about the comparison.
     
  13. Printing bigger when the camera offers more MP? Cropping to a certain dimension? Both real-life application where pixel-level performance matters. Not to mention that it serves to keep those who claim pixel-level improvements honest ;)

    The image comparisons done at dpreview are at pixel level (which maybe the reason Andrew sees "significantly less" chroma noise). Comparing dpreview images between the D850 and the D500 shows no difference in chroma noise to me (which is supported by the dxomark results when viewed in Screen mode (which is the correct one for comparing results with dpreview images)). In Print mode, the expected 1-stop difference surfaces - identifying it as merely an artifact of forcing an DX and an FX image to print in the same size. Shouldn't a D500 image and a DX-crop from the D850 be virtually identical?

    It's simply lacking the resolution. And the strong AA filter doesn't do it any favors in that regard either.


    I do have to walk my argument back a bit though. The "Print" view at dxomark is the more useful one as it reflects the more common application of printing in the same print size or reducing images to the same number of pixels, for example, to publish on the web. It doesn't make the "Screen" view totally useless though.
     
  14. This whole Amazon thing explains a lot now. Thanks Shun. I have never really paid too much stock in their reviews. In the beginning though I thought they were great.
    But I thought they lost their way a long time ago.
     
  15. ShunCheung

    ShunCheung Moderator

    I have been reading DPReview since around year 2000 when Phil Askey was in charge. Traditionally their reviews are very thorough with lots and lots of details, but I simply don't have the patience to read any one of them from beginning to end. Usually I check the conclusions and maybe take a look at some of the details I am interested in.

    Since essentially nobody else advertises on DPReview, they must generate enough revenue for Amazon to justify their a dozen or so staff members, which is not exactly a tiny payroll.

    In a broader sense, in the last few years, I think social media have been used to promote photo gears as well as to trash rival brands. Whenever Nikon has a new camera, there seems to be an army of people trying to find every little fault and then the issues are greatly exaggerated in various forums. Clearly the D600 indeed had some oil-on-shutter problems and then the D800 was declared "no recommended" by Thom Hogan: Thom Hogan Declares D800 "Not Recommended" .... At least so far people haven't found any serious problems with the D5, D500, and D850. But it is not limited to Nikon; I see various new Canon cameras are trashed in the last couple of years. Meanwhile there is plenty of hype on various Sony mirrorless cameras and most recently, we see one article after another promoting the D850.

    None of these should be surprising. The social media have been used to influence elections. They have also been used to both promote and trash various cameras for several years.
     

Share This Page