Official D3 samples here

If you take a photo of a black cat in a dark room on auto exposure, the cat will appear grey (not black). If you take a photo of snow on a bright, sunny day, the snow will appear grey (not white).

The camera doesn't know you're shooting a black cat, or snow. You either have to use exposure comp (or manual mode) based on the histogram/LCD, or use a grey card or light meter. This isn't rocket science. Just fundamental exposure skills. NO camera in the world will get these right without user intervention.

Jay
Interesting. I've seen LOTS of shots with + - 1/3 stop or so. Just
never seen anyone pull it back 1.5 stops before. That's a lot of
light to manipulate with that tool it seems.


I guess I just don't get the WHY of it.
--
Jay Philip Williams Photographic Design
http://www.jpwphoto.com
 
I have seen hundreds of real real PJ low light samples with the 5D -
it cant compare. The 5D images typically have much more color noise.
The D3 samples here are another level.
I agree. I had a 5D in addition to my D2x for one year and can only confirm this. The 3200 and 6400 ISO samples are taken in artificial light, the 5D is no match under similar conditions. I did a bit sharpening and printed them in A3 - they look stunning.

-Thilo
 
This is a very good observation.

My initial reaction when I saw the 6400 sample -- which was before reading this thread -- was amazement. It's not often that a camera manufacturer does a torture test for their publicity pics but this is it -- a terrible yellow bias to the lighting, an apparently hand-held image, high ISO and in all honesty, slight underexposure. And it has come through with flying colors, a smooth subtle grain reminiscent of low ISO slide film, and absolutely no chroma noise at all. If there's any issue with sharpness, I suspect it is probably camera shake. Yes, at this resolution, even 1/160th of a second will betray questionable technique.

The 5D in my view is not likely to equal this result even at 3200 ISO. Go take a look at Phil Askey's test shots at the Imperial War Museum in his 5D review.

Even if the sharpness is on the low side, which I agree it is, these images will still produce beautiful 12x18 enlargements. Put sharpening on, put the camera on a tripod (or use better handholding technique) and the image will be better still.
 
I just got done shipping off a bunch of jazz concert performance shots to a record label - shot at ISO 800 and 1600 with a d100 and a d2x. What a comparison - the ISO 6400 shot shown here is lower noise and better color than any of the shots from the d100 or d2x at ISO 800. Similar shot of spotlit trumpeter against dark background, even - while the D2x ISO 800 shot is usable for images inside the CD booklet (because they are so small, the noise doesn't show at all) at poster size the noise is bad enough to make me think, convert to black and white and go for that 1950's pushed Tri-X effect...

Looking at the skin tones here, I am completely sold on the D3 for low light.
I down loaded all images into PS CS3 and took a careful look. My
impressions:

1- At 6400 and poor light the D3 is BETTER at ISO 6400 than my D2x
was at 800. Without a direct comparison it is hard to say, but my
guess is that ISO 640 on a D2x would be close to the D3 at 6400.
That's one heck of a jump. And . . . the noise had a nicer quality to
it compared to my D2x.

2- At ISO 3200 the quality was amazing. The lighting (on the lady)
was very harsh, but similar to what a PJ or event photographer has to
deal with.

3- At ISO 400 I uprezed the file to 240 dpi at 48 inches wide. Did a
slight USM at 0, .4, and 150% . . . and made a 10" print from a
cropped section (around the eye). Pretty impressive! Looks very sharp
from a the distance you would read a magazine from. Again, it's a
guess, but I would say that the 12 megs in the D3 resolves better
than the 12 megs in my old D2x . . even at ISO 400 vs 100 with D2x.
Two extra stops AND better resolution - it looks like. We shall see,
but that's my take.

--
Steve Bingham
http://www.dustylens.com
 
What worries me is that it is possible that Nikon and sony don't have
the on-chip noise reduction as used by canon (some version of it
obviously) but instead uses software anti aliasing and noise
reduction...
Does Nikon have to do it the way that Canon does in order for it to be good? Rather than 'on-chip' Nikon appears to have used large, light-thirsty pixels with reduced distance to the photosites from the microlenses and several optical innovations to capture any lost light to get as much of the 'real thing' (light) as possible rather than resorting to processing prowess.

From these samples and my own limited experience with the 5D, there's no contest between the D3 and the 5D at these elevated ISO's. I haven't seen or heard anything to lead me to think there's anything negative about the images at lower ISO's either.

But alas, I would also like to see some RAW file results as well.

I think about the only negative thing that we'll see with this camera is the wait that we'll endure to get one! The line will be long enough with Nikon-native shooters ordering theirs, not to mention the group of Canon shooters who find it necessary to switch (in their own minds)...

Sean
 
Really ground breaking image quality.
The iso 400 samples are my favourites. They really pop out of the screen!
Well done Nikon.

--
A dpreview browser.
 
What worries me is that it is possible that Nikon and sony don't have
the on-chip noise reduction as used by canon (some version of it
obviously) but instead uses software anti aliasing and noise
reduction...
Does Nikon have to do it the way that Canon does in order for it to
be good? Rather than 'on-chip' Nikon appears to have used large,
light-thirsty pixels with reduced distance to the photosites from the
microlenses and several optical innovations to capture any lost light
to get as much of the 'real thing' (light) as possible rather than
resorting to processing prowess.
No, they don't have to do it the way Canon does it.

But if it's not done on-chip raw files will be noisy, especially since cmos is noisier than ccd by nature albeit not by much. Sure, the photosites are large, but not THAT large...
From these samples and my own limited experience with the 5D, there's
no contest between the D3 and the 5D at these elevated ISO's. I
haven't seen or heard anything to lead me to think there's anything
negative about the images at lower ISO's either.

But alas, I would also like to see some RAW file results as well.

I think about the only negative thing that we'll see with this camera
is the wait that we'll endure to get one! The line will be long
enough with Nikon-native shooters ordering theirs, not to mention the
group of Canon shooters who find it necessary to switch (in their own
minds)...
If Nikon operates as fast as usual, D3 will start to ship here in sweden about two weeks after the D4 is introduced :D

They are a bit like apple that way,good products, lousy marketing (Nikon) and horrible distribution and planning (nikon and Apple).
--
Anders

Some of my pictures can be seen at;
http://teamexcalibur.se/US/usindex.html

event photography and photo journalism
 
I love reading this thread because I never learned to analyze things this way, I'm more of a hacker type, I'm gaining alot of new perspective from you guys.

Weeks ago my 1st reaction to the D3 was what people here seems to have found. Nikon setup models at the D3 party, and I shot them with the new wide "N"s and with the 400/2.8 "N" lens that Nikon had setup on Gitzo's. The detail with the 400/2.8 at ISO 6400 holds up pretty well when compared to the normal lens shot. At ISO 3200 the detail is much better.

The reaction that was bantered about by the many shooters at the D3 party, was that the D2X's ISO 800 is now with the D3, ISO 6400. On the following days, I tried pushing 6400 at nite, I.E. I tried to shoot more at 3200, 4000, and 5000, and only experimented with 6400 at the stadium turns where the light falls off and subjects are backlit.
 
Of course it isn't! I would love to have a D3! but, hey - I'm not made of money here. That said, I think my 5D produces better high-ISO, noise-free images than these "official" images posted by Nikon. A few notes below:
The 5D 3200 stage shot you offer for

comparison is nice, ((((Thanks!- Eric)))) but.. the ISO is one stop down, there's twice as
much available light, and there's more noise - especially chroma -
compared to the D3 image.
Acually, compare the Nikon 3200 shot to my 5D 3200 shot - they are about the same light levels, maybe the Nikon was 2/3 stop less (mine was f5.6 at 1/100, D3 was f4 at 1/125). And that Chroma noise is the easiest to reduce with NR - without loosing detail. See here is my image with a little noise reduction- no detail lost, no noise anywhere (and no clumps in the OOF areas)


The 5D image also holds no more detail -
I agree. And this goes for the D2X - I've seen way more detail from the D2X raw files than my 5D can do (at low ISO, of course)
Several things contribute to its sense of detail. Because this was
taken at 400mm, most of the picture comprises the area of focus, and
the image shows obvious signs of sharpening.
There was no sharpening except the in-camera JPEG sharpening set at defaults. If I could have gone to F4 ro F2.8 at 400mm, I would have - 1/100 was too slow and I lost a few shots from motion blur. Guess I need to buy that $7000 lens! Anyway, I don't buy it that having less DOF makes an image show more detail. If anything, it pulls your eyes to the areas of sharpness.

And one disturbing element of the Nikon trumpet picture is the OOF hand. It has way too much posterization - the colors shades don't have smooth transitions at all. Isn't this something the 14 bit thing is supposed to help with?

--



Bossier City, Louisiana
http://www.pbase.com/ericsorensen
 
... the ISO 6400 shot.

I'm lucky to have a friend living nearby, and we just did a 12 x18 print of the ISO 3200 and 6400 pics without any PP.

I swear, if no one told me, I would have thought they were ISO 200 shots. The impression is formed looking at the prints held from the eye against the light with elbows bent, and then peering cross-eyed at OOF areas (and admiring the "grain" in the process), enjoying the rich, natural colors with no nasty greenish-thingomagic muck in the shadows and black areas are 100% black, etc.

Even "DPI"-peeping the prints, they're far, far cleaner than my D70 (anyone remembers that camera?) ISO 400, out-of-camera, chroma-riddled JPEG images printed at half that size.

The magic is in the prints, people - if you're fortunate to have a printer close at hand - print these pics!

--
Regards, David Chin
http://nikond3.dpnotes.com/
 
Is this ONLY for auto modes? I shoot manual 95% of the time, with a mix of the others thrown in when I'm goofin around shooting surfing or something. I understand the fundamentals of photography at this point just fine.

Thus far I only use ISO , Shutter Speed, Aperture, and ocassionally a filter to control my exposures.

This is completely fascinating to me right now and has led me on a journey (one I probably didn't need to take but it's too late, I'll loose sleep over this if I don't figure it out)

SO... I grab my D2x and sart looking through my menus (which I haven't spent much time with since I updated the firmware)

Exposure Comp (On / Off / On Auto Reset)
vs
Fine Tune Exposure (Per metering mode)
vs
Simply dialing it in via the button next to shutter release

Ran a quick test, program auto mode, auto wb,
1st shot 3D metering - WOW is this thing OFF - way under exposed
2nd shot dial in with button up to +1.0 - STILL not a single blown highlight.

3rd shot dial in with button up to +1.3 - finally one small blown highlight which would definitely be recoverable.

Same result with "Fine Tune Exposure"

Can't figure out what "Exposure Comp (On / Off / On Auto Reset) does though?
 
Terrible vignetting, especially on those first two images!

Just kidding - had to say it. I've been hearing Nikon folks say that for years. FF is the way to go - the bigger the sensor the better - same with film.
--



Bossier City, Louisiana
http://www.pbase.com/ericsorensen
 
I am a doctor and I dont see any hint of yellow in the eyes.

I do feel like giving the model a hamburger though :-)
--
Jake
 
Clearly Nikon wasn't kidding when they called the D3 revolutionary. It's not just the lack of noise at high ISO, the noise quality is also superb. It makes a direct comparison to the D2x or to Canon's cameras impossible.

--
Fabian
 
That shot at ISO 400 w/ the 85mm f/1.4 is incredible... I've shot thousands of pictures with the 85mm f/1.4D and never seen it look that good at f/2 (f/2.2)... Either they used a magic sample of the 85mm or the D3 is doing something special on the back end... It's also incredibly sharp (which I'm used to) but man...

All I want the D3 for is to use with the 85mm f/1.4D and if the real life examples (i.e. my shots) are as good as this one, I'm happy. :)
 

Keyboard shortcuts

Back
Top