Nikon D850 First Camera To Score a Perfect 100 on DXO!

kanuck

https://www.dxomark.com/nikon-d850-sensor-review-first-dslr-hit-100-points/

I knew this was a special camera when I started shooting with it. You could tell right away when looking at the images that it wasn't just another camera release and that the pictures looked like something I haven't seen from any camera body before. I have owned a ,F100, D40x, D80, D200, D300, D700, D600, D800E, D4s, D810, and now this D850 and it is definitely magical!!

heartyfisher

hehe .. its not 100 / 100 its probably 100 out of 999 ;-)

PB_PM

Yawn. More useless DXO scores to placitate measurebaters.

WestEndFoto

After owning it for two weeks, I am certainly really liking it. Now if Adobe would only upgrade Lightroom......

donaldejose

I agree with kanuck. The D850 is "magical." I am constantly amazed how much difference there is coming from a D800. Not only does it have more detail than you need it and great dynamic range it also has the clean high ISO ability. I have not shot fast moving subjects like BIF so cannot comment on that ability although I would expect it to be fine with the battery grip and the more powerful battery. Try one out if you have a chance to do so.

tc88

My experience with the D850 sensor is that resolution is good, high ISO noise is good. But pixel peeping at low ISO, on a per pixel basis, I prefer the Toshiba DX 24MP. I found while very fine, the noise is more prominent on D850. I guess that's not really a surprise since D850 sensor is derived from the D500 21MP and may have similar noise profile per pixel. And there is enough debate between the low ISO noise performance between D7100/7200 and D500/7500.

donaldejose

Remember that when pixel peeping at 100% and comparing noise between an FX sensor and a DX sensor you are not looking at the same size image. In order to better reflect what the eye will see in the same size image from each sensor you have to first prepare the images to the same size so you are comparing apples to apples.

tc88

Of course I'm not comparing the overall noise between FX and DX. That will be nuts. That's why I specifically mentioned per pixel wise. Nikon made the trade off to have better high ISO noise for D500, and as projected, it made the same trade off for D850. So I wouldn't call the D850/D500's pixel/sensor design to be best across all ISOs.

donaldejose

Comparative noise should not be judged per pixel wise. It should be judges "per print size wise." People don't view images at 100% on their computer monitor. Nor do they view a 24 x 36 inch print in a gallery from 12 inches away. Why? Because then they cannot see the whole image.

snakebunk

I think it is most relevant to compare crops captured from equal sensor size and then resized to equal with and height.

rmp

Some fools (that means me) blow the images up to a ridiculous size (8 ft x10 ft) then zoom in to check pixels. Yes, I know that is crazy. I do it because, sometimes (1 in 100 images) my wife wants an image printed on metal at 40 to 60 inches and I hate to be embarrassed when the printer (Image Wizards) calls and asks "Do you really want to print this image at that size?" Yes, Image Wizards will actually look at a 40 x 50 image from 10 inches away. And, I like it that way. Of course, it means most of my images never hang on a wall.

tc88

donaldejose, from what you are saying, any print that looks good on 4x6, can be blown up to any size then. Because people are supposed to step back based on the print size, there is no limit regarding how big a print can be then. I don't think most people agree with that.

BTW, I'm not aware of any method that can directly give out noise estimation on a print level. Noise analysis is done on a per pixel level first, then aggregated/averaged/normalized over the entire frame. After all, digital picture stores data per pixel wise.

donaldejose

No tc88, that is not what I am saying. Make a 24 x 36 inch print from a DX image and another one from an FX image each shot at the same ISO. Now you have an equal basis upon which to judge the relative noise from each sensor. Comparing two different size sensors on your computer at 100% isn't an equal viewing comparison.