Nikon D3 and D300 NEF raw files now supported by Dave Coffin'sdcraw

I just discovered that yesterday Dave Coffin released
version 8.79 of dcraw.c, which among a number of other
updates includes code to handle Nikon's soon to arrive
D300 and D3 model DSLR's (announced for release in
November, and apparently rumored to be available on or
about the 27th).

http://www.cybercom.net/~dcoffin/dcraw/dcraw.c

It is safe to say that UFRAW, the graphical interface
that uses dcraw.c code and provides a plugin for The
GIMP, will soon have an upgrade to include the new dcraw
code, and will also likely be available before the
actual cameras are on the shelf.

http://ufraw.sourceforge.net/

 

I'll be interested in hearing some more objective evaluations of the
noise performance on these models, perhaps.

 

I tried it with some NEF files I downloaded from one
of the demos a few weeks ago, and found the results
quite interesting. It's mighty hard to adjust things
precisely using a command line utility, but I did get
fairly close output from dcraw compared to what the
embedded JPEG image was. The surprizing thing was
that the noise on ISO 25600 images was better from
dcraw. At ISO 6400 it was about the same.

It will have to wait until UFRAW upgrades for it to be
easy enough to really experiment much.

 

UFRAW is pretty cool. Between UFRAW and The GIMP you can do an awful
lot with your Nikon DSLR and Free Software.

 

There's some interesting stuff coming out on DPreview's
forums re the D300:
http://forums.dpreview.com/forums/read.asp?forum=1039&message=25515548

I note his comments about how well Noise Ninja works on D300
high ISO noise. David Kilpatrick also posted samples of
ISO6400 raw converted with no NR. I played with those using
Noiseware, and got much better results than the in-camera NR.

There's a consistent pattern appearing where it seems that
the D300 IQ is excellent - sharp, some moire with lenses
like the 24-70, at least 2 stops better high iso noise
performance than D2xs, more than a stop improvement over the
D200/D80.
The impression of good low noise performance might be
further boosted by a consensus of opinion that the D300
exposes less conservatively than the D200, but more
accurately than D80.

The ISO 200 shots I've seen have been extremely good -
seeming to have noticeably better shadow detail and low
noise when playing with levels.

 

Udi Fuchs just release UFRAW 0.13, based on Dave
Coffin's dcraw version 8.80. There is a significantly
long list of changes adding new features. Included are
a few bug fixes and of course hidden in it all is that
the dcraw code includes support for the soon to be
released Nikon D300 and D3 cameras.

 

I've made a preliminary analysis of the Nikon D3's dynamic range,
using the Nikon D3 raw samples posted here:

http://aaronlinsdau.com/gear/articles/d3.html

The latest build of the raw converter dcraw can indeed understand D3
raw files though it seems there are some bugs with some of the
conversion options, the -D option doesn't seem to work; but I was able
to get a linear 16-bit conversion without bayer interpolation that I
could analyze. I should stress that it is important to take the raw
converter out of the equation as much as possible or else you are
testing the raw conversion software as much as the camera. The
linear conversion w/o bayer interpolation does that; to be specific
the options used were
dcraw -v -h -4 -T -k 0 -H 1 -o 0 -r 1 1 1 1 filename.nef

Looking at the ISO 200 sample, I analyzed a few regions of the image
in the red and blue channels, measured the noise in IRIS and
extrapolated the result to vanishing luminance. The dynamic range
seems to be about 11.7 stops (read noise is about 4.9 in 14-bit raw
levels, and the highlights clip at 16383 in those units).

I had to extrapolate down to zero raw level from a bit above since
Nikon clips their blackpoint and so the noise is not a Gaussian
distribution but a half-gaussian at the bottom end, and fluctuations
are thereby clipped as well; using the absolute lowest exposure levels
underestimates the noise.

This is a *very* rough, preliminary measurement. A proper measurement
would use a blackframe image, shot at a high shutter speed with a lens
cap on the body. That will take a lot of sloppiness out of the above
analysis, but I would be surprised if the dynamic range is more than
12 stops. A similar analysis at ISO 1600 yields about 10 stops DR.
Thus it would appear that the D3's dynamic range is about the same as
Canon 1 series DSLR's.

Unfortunately the sample images are not appropriate for determining
the quantum efficiency of the sensor, which to my thinking is one of
the determining factors in high ISO performance.

The raw was shot in 12bit mode rather than 14-bit mode (all raw levels
were multiples of four in 14-bit units). It will be interesting to see
whether an image shot in 14-bit mode has more dynamic range, since the
12-bit sample here has a quantization step which just about equals the
noise. Theoretically that could get you a very tiny fraction of a stop
more, but I don't think quantization error is the limiting factor
here. Anyway, DR appears to roughly equal the Canon Mark 3 at ISO 200.

BTW, pattern noise (banding) seems fairly well controlled (though
visible in the deepest shadows if you boost them enough).