Resources
Digital Cameras
Accessories
Tutorials
News
Forums
Search
About Us
Could you spare a few dollars to help someone who needs it? Support our adopted charity House of Hope. Click for more info.
|
Digital Camera Home > Digital Camera Reviews > Nikon Digital Cameras > Nikon D1
|
We've begun including links in our reviews to a Thumber-generated index page for our test shots. The
Thumber data includes a host of information on the images, including
shutter speed, ISO setting, compression setting, etc. Rather than clutter
the page below with *all* that detail, we're posting the Thumber index so
only those interested in the information need wade through it!
;) |
 |
 |
Outdoor
portrait: (1229k)
This is a tough shot for
many digicams, due to the extreme tonal range (which is why we set it up
this way!). The trick is to hold highlight and shadow detail without
producing a "flat" picture with muddy colors and the D1 did an excellent
job. We shot this image using the automatic
(1231k) and manual (1229k) white balance settings, with the manual setting
producing the most accurate results in both skin tone and white values.
(We actually did shoot a series with the daylight setting, but as it was
nearly identical to the automatic, we didn't save the series). The D1 did
an excellent job with colors in the flowers and the model's pants,
particularly the often-difficult blues. (Many digicams have a tendency to
reproduce these with a purplish hue). We did feel though, that the skin
tones were a little problematic for a $5,000+ camera: They're a little
"hot" overall, with a yellowish cast in the highlights that doesn't appear
in the lower tonal values. If this were a prosumer-level digicam, we'd let
this level of color inaccuracy pass and give the camera a "very good to
excellent" rating on color, but feel that a camera of this price and
sophistication needs to be held to a bit higher standard. (In fact, we'd
rate the D1's color as more accurately saturated overall than that of
Nikon's excellent prosumer Coolpix 990, which we commented on more
favorably in this regard) - Fortunately, there appears to be an excellent
third-party solution to color problems of this sort, in the form of Mike
Chaney's Qimage Pro program: It can work from the "raw" files the D1
takes, producing very natural skin tones and more accurate color overall.
(Unfortunately, in all our tests, we only saved one image in "raw" format,
none of the outdoor shots!) Do check Mike's site though, for $30, Qimage
looks like a "must have" for the D1 shooter. Addendum: A number of
readers have written it to tell us of Eric Hyman's excellent Bibble software,
which appears to do even better than Qimage on the color-correction front
(although it lacks Qimage's high-ISO noise filtering). At only $50 for
Bibble, we'd also put it on the absolute must-have list for every
D1 owner! (Which one should you get? - Why not both? Each
offers capabilities the other lacks, and for $80 for the pair of programs,
you really can't go wrong! - Note that we have no affiliation with either
Mike Chaney or Eric Hyman.)
Resolution and detail look
extremely good, particularly in the small green leaves next to the model's
shirt and the strands of her hair. (Again, making the obvious comparison
to the Coolpix 990, we feel that the D1 did a better job on the finest
details, despite having a somewhat lower resolution at the CCD level than
the 990. One possible reason might be that the lens used on this shot (the
Nikon 17-35 f/2.8 D) costs more than the entire 990 camera!) The shadow
areas also look great, showing a lot of detail and virtually no
noise! We required +0.3 EV of exposure compensation on this shot for our
main image (less than most cameras require, perhaps due to the accuracy of
Nikon's matrix metering system), which managed to get the best exposure
overall, albeit at some cost to the highlights in the shirt and the white
flowers. The table below shows the results of a range of exposure settings
from -0.3 to +1.0 EV in both automatic and manual white balance
settings.
Exposure Compensation Series, Auto White
Balance
-0.33 EV
Size: 2000x1312
Shutter: 1/500
Aperture:
F11
(1205k) |
0.00 EV
Size: 2000x1312
Shutter: 1/400
Aperture:
F10
(1235k) |
+0.33 EV
Size: 2000x1312
Shutter: 1/350
Aperture:
F9.5
(1231k) |
+0.67 EV
Size: 2000x1312
Shutter: 1/320
Aperture:
F9
(1179k) |
+1.00 EV
Size: 2000x1312
Shutter: 1/250
Aperture:
F8.5
(1174k) |
Exposure Compensation Series, Manual White
Balance
-0.33
EV
Size: 2000x1312
Shutter:
1/500
Aperture: F11
(1221k) |
0.0
EV
Size: 2000x1312
Shutter:
1/400
Aperture: F10
(1239k) |
+0.33
EV
Size: 2000x1312
Shutter:
1/350
Aperture: F9.5
(1229k) |
+0.67
EV
Size: 2000x1312
Shutter:
1/320
Aperture: F9
(1176k) |
+1.00
EV
Size: 2000x1312
Shutter:
1/320
Aperture: F8.5
(1169k) |
|
|
| |
 |
Closer portrait: (1167k)
The D1 also
turns in an excellent performance with this "portrait" shot, although we
would have done better to use a longer-focal length lens. Shorter focal
length lenses tend to distort facial features in close-up shots like this
and the availability of longer focal lengths is a key feature if you're
going to be shooting close-up people shots. (This image was again shot
with the 17-35mm zoom, this time set at the 35mm position, which resulted
in an "effective" focal length of 52.5 mm. Our ultra-sharp 105mm Nikkor
Micro lens would have done well on this, producing an effective focal
length of about 160mm, but we were pressed for time, wanting to catch the
sun at the right angle, and didn't have that lens with us.) Our main
shot (1167k) required no
exposure compensation away from the settings chosen by the metering
system. Sharpness and detail are again exceptional, again with almost
imperceptible noise in the shadow areas. We're still not crazy about the
skin tones, but the detail in the image is exceptional. The table below
shows the results of a range of exposure settings from -0.3 to +1.0 EV. We
chose the manual white balance setting again as it produced the most
accurate color balance.
Exposure Compensation Series
-.33
EV
Size: 2000x1312
Shutter:
1/400
Aperture: F10
(1191k) |
0.0
EV
Size: 2000x1312
Shutter:
1/350
Aperture: F10
(1167k) |
+.33
EV
Size: 2000x1312
Shutter:
1/320
Aperture: F9.5
(1132k) |
+0.67
EV
Size: 2000x1312
Shutter:
1/320
Aperture: F9
(1143k) |
+1.00
EV
Size: 2000x1312
Shutter:
1/250
Aperture: F8.5
(1115k) |
|
|
|
 |
Indoor portrait, flash: (1199k)
The D1 performs
very well in this category and has a tremendous range of exposure and
flash options to play with, thanks to the flexibility of the optional external SB-28DX flash unit. (NOTE that the D1 does not have any internal flash:
This test was performed with an external unit, available at extra cost.
Some shots (including our "main" one for the category) were taken with TWO
flash units... - The lack of onboard flash is normal and expected on a
professional SLR, but we wanted to point out that these shots were taken
with an external flash, so nobody would accuse us of stacking the deck in
favor of the D1 relative to the prosumer cameras with onboard flash.) We
took several shots with the main flash and then tried a series with two
flash heads, daisy-chaining the flash units together with an accessory
cable. In the first series (using just the main flash) we achieved the
best results overall when we bounced the flash and boosted the exposure
compensation to +0.7 EV while using the flash white balance
(1172k) setting, which lit the
foreground and background nicely without blowing out the highlights or
causing a major color shift. Both the highlight and shadow areas look very
good, as does color balance, although the strong room light gave things a
rather warm cast. Staying in the flash white balance setting, we snapped
the same shot as before but with a direct instead of a bounced flash at
+0.7 EV, producing this (1192k) image. Using the direct flash makes the
lighting a little harsh, darkening the background slightly and producing
more noticeable shadows. We took the same shot using the automatic white balance (1202k), which produced similar results. We
also tried a little photographer's trick of balancing the flash to the
ambient lighting, and then using the ambient setting for the white
balance: Here's (1194k) a shot taken with the flash on, but
using the manual white balance setting for the room lighting. Yuck, really
blue! Check this out though: Here's (1122k) a shot taken with the same white
balance setting, but an orange gel taped over the flash head. - Pretty
good, but it'd need a little tweaking for professional work. (A very weak
yellow gel added to the flash head would probably have done the trick.)
Still, this illustrates a very useful technique for balancing room light
and flash illumination. (And for it to really work well, you need a camera
with an accurate manual white balance capability, like the D1.)
Now, on to the multiple flash
tests. In the first shot, we bounced one flash from the side at full
strength and pointed the main flash directly at the subject at -2.7 EV
(just the barest fill), producing this (1196k) shot, which resulted in subtle, moody
lighting and a nice color balance (although a touch warm). Using the same
setup, we kept the side bounce at full strength and increased the main,
direct flash to -0.7 EV, producing a brighter image with better lighting
and color balance. We chose this image as our main shot (1199k), since it produced the best exposure
overall with good lighting and subtle highlights. Keeping the side bounce
at full strength, we boosted the main, direct flash to 0.0 EV, producing
this (1182k) image which still looks nice but the
main flash begins to dominate the image slightly. Finally, we decreased
the side bounce to half strength and left the main, direct flash at 0.0
EV, producing this (1196k) image, which is much closer to the
effect of the on-camera flash alone, but still has some fill to eliminate
some of the harshness and shadows. Although the overall exposure still
looks pretty good, the main flash definitely dominates the image and the
highlights become more prominent while the background gets darker.
Again, an excellent
performance on this test shot, as we achieved good results in both of our
series and the D1's many exposure options give you a lot to play with. For
practicing pros (or well-heeled amateurs), the ability of the SB-28DX
flash units to work in concert with each other is a huge plus. The
lighting control you can achieve in this manner is unprecedented, going
way beyond anything we've been able to achieve with the prosumer cameras
we've tested in the past. (To preempt the likely comments though: You're
right, we haven't tried working with any of the special slave units for
prosumer digicams, with which we might be able to achieve the same
results. Part of our point here though, is that the Nikon system offers a
single-vendor solution for multiple-head flash that's nicely integrated
with the camera itself.)
|
|
|
|
Indoor portrait, no flash: (1043k)
This shot is a
very tough test of a camera's white balance capability, given the strong,
yellowish color cast of the household incandescent bulbs used for the
lighting. No problem for the D1, which gave us plenty of exposure options
to try. For our main EV test, the D1 did by far the best job in the manual (1043k) white balance setting (which we chose
for our main shot (1043k)), producing nice color balance and good
detail with little noise. We also shot in the tungsten (1067k) and automatic (1077k) white balance settings, which both
produced very warm results (auto a little more so than tungsten). The
table below shows our standard series, a range of exposure compensation
settings from zero to +2.0 EV in the manual white balance setting.
Exposure Compensation
Series
0.00
EV
Size: 2000x1312
Shutter:
1/40
Aperture: F3
(1079 k) |
+0.33
EV
Size: 2000x1312
Shutter:
1/30
Aperture: F3
(1084k) |
+0.67
EV
Size: 2000x1312
Shutter:
1/25
Aperture: F2.8
(1112k) |
+1.00
EV
Size: 2000x1312
Shutter:
1/20
Aperture: F2.8
(1043k) |
+1.33
EV
Size: 2000x1312
Shutter:
1/15
Aperture: F2.8
(1044k) |
+1.67
EV
Size: 2000x1312
Shutter:
1/13
Aperture: F2.8
(1040k) |
+2.00
EV
Size: 2000x1312
Shutter:
1/10
Aperture: F2.8
(1016k) |
As an interesting aside, discerning readers may note
the much shallower depth of field in these shots than is commonly the case
with the prosumer digicams we test. This is due to the much longer focal
length of the lenses we're using on the D1, and the resulting larger
physical iris openings for corresponding f-stop values. These shots were
taken with the lens set to a focal length of 24mm, and an aperture of
f/2.8. By comparison, a typical prosumer digicam has a lens focal length
of only 8-9mm on this shot, or roughly 1/3 of the focal length. (This due
to the smaller size of the CCD sensor in those cameras.) As a result, the
prosumer cameras will have much greater depth of field than pro units like
the D1, under similar shooting conditions. This can be a two-edged sword:
It's tough to get good subject isolation (shallow depth of field) with
prosumer digicams, but likewise tougher to get great depth of field with
the pro models.
An
interesting feature on the D1 is the ability to adjust the intensity of
the white balance presets. We shot a series in the tungsten white balance
setting from -3 to +3 adjustment levels with the results below. (The
actual increments for these adjustments in degrees Kelvin are listed in
the main review. Very nice feature we wish other digicam manufacturers
would adopt - how difficult would this be to add to most any digicam?)
Finally, we shot a series with each ISO setting. We
found that the noise level only increased minimally from 200 (1055k) to 400 (1090k) and just slightly more when we went to
800 (1240k). We saw the biggest increase in noise
at the 1600 (1232k) setting (as you'd expect). While
shooting at the 1600 sensitivity setting, we took one shot with the
aperture set at f/6.3 (1219k) to gauge the variance in the depth of
field. (Compare this with the ISO 200 shot (1055k), which was taken with an aperture of
f/2.8.) As noted above, we observed that the D1's depth of field is less
than the prosumer digicams because its lenses are made to cover a 35mm
frame and they have much longer (though ineffective) focal lengths. As a
result, f/6.3 on the D1 is more like the f/2.8 setting on a typical
digicam lens. Compared to the prosumer cameras we've tested, the D1's
noise performance is exceptionally good: Noise at ISO 1600 is less than
most lower-end units at ISO 400. (We haven't tested any other pro cameras
yet, so can't comment on the noise levels relative to other pro units.) We
did notice a noise characteristic at high ISO that's been commented on by
some other writers: There's a "banding" in the noise patterns at ISO 1600.
This is most noticeable in the wider-aperture ISO 1600 (1232k) shot, and can be seen in the blue
channel, across the model's face and hair. As noted, the overall noise
level is quite low, but the non-random nature of the noise pattern at ISO
1600 makes it more visible than it would be otherwise.
|
|
|
 |
House shot: (1353k)
NOTE that this
is the "new" house shot, a much higher-resolution poster than we first
used in our tests. To compare the D1 with previously-tested cameras,
here's a shot of the original house poster at the manual (1344k) white balance setting.
First of all, let us say that
the D1 did an excellent job on this shot. We preferred the manual (1353k) white balance setting on this one, as
the color balance seemed the most accurate. The automatic (1332k) setting produced just slightly warmer
results and the daylight (1355k) setting resulted in a noticeably warmer
image. Resolution and detail look great all over. You can see the tiniest
details in the tree limbs and shrubbery, as well as in the often hard to
distinguish brick and shingle areas. As far as we can tell, the only noise
in the image is coming from the actual house poster itself. We picked up
just the tiniest halo (about a pixel or so) around the dark and light
edges, giving away the in-camera sharpening, but the default sharpening is
some of the best and most subtle we've seen on a digicam. The D1 performed
brilliantly and we really don't have any complaints! The table below shows
our standard quality series.
Image Quality Series
Uncompressed
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Fine
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1355k) |
Normal
Size: 2000x1312
Shutter:
1/30
Aperture: F8
(664k) |
Basic
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(331k) |
Sharpness Series
The D1 provides four options for in-camera sharpening:
Auto (default), "Soft", "Hard", and "None". As noted above, the default
sharpening does a really excellent job on most images. You may want to
choose soft or hard depending on the subject matter and type of printout
you'll use. The No Sharpening option would be the best choice if you
intend heavy editing of the images, with sharpening applied in
Photoshop(tm) or other program at the end of the process. We shot tests
with the camera's variable sharpness settings, with the results below.
Overall, the camera did a good job of not over-sharpening in the hard
setting and doesn't get too soft at the other end.
Sharpness Series
Default
Value
Size: 2000x1312
Shutter:
1/40
Aperture: F8
(1355k) |
Soft
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1351k) |
Hard
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1354k) |
No
Sharpening
Size: 2000x1312
Shutter:
1/30
Aperture: F8
(1374k) |
|
|
| |
 |
Far-Field test (1318k)
This image is
shot at infinity to test far-field lens performance. NOTE that this image
cannot be directly compared to the other "house" shot, which is a poster,
shot in the studio. The rendering of detail in the poster will be very
different than in this shot, and color values (and even the presence or
absence of leaves on the trees!) will vary in this subject as the seasons
progress. In general though, you can evaluate detail in the bricks,
shingles and window detail, and in the tree branches against the sky.
Compression artifacts are most likely to show in the trim along the edge
of the roof, in the bricks, or in the relatively "flat" areas in the
windows.
Again, the D1
performed beautifully with this shot. This subject is the strongest test
of detail of any we do, and the bright white of the central bay window
often tricks digicams into losing detail in that area. The D1 somewhat
fell prey to this as well, but you can still see the outline of some of
the bay window's detail. Color balance and saturation look very good and
noise is extremely low, showing just slightly in the roof shingles.
Resolution looks great all around, especially in the bricks and trees. The
table below shows our full quality series.
Image Quality Series
Uncompressed
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Fine
Size: 2000x1312
Shutter: 1/1000
Aperture:
F5
(1318k) |
Normal
Size: 2000x1312
Shutter:
1/1250
Aperture: F4.5
(658k) |
Basic
Size: 2000x1312
Shutter: 1/1600
Aperture:
F4
(328k) |
As with the House poster, we
shot with the camera's variable sharpness settings, with the results
below.
Sharpness Series
Default
Value
Size: 2000x1312
Shutter:
1/320
Aperture: F9
(1323k) |
Soft
Size: 2000x1312
Shutter: 1/320
Aperture:
F9
(1328k) |
Hard
Size: 2000x1312
Shutter: 1/320
Aperture:
F9
(1336k) |
No
Sharpening
Size: 2000x1312
Shutter:
1/320
Aperture: F9
(1324k) |
|
|
|
|
Lens Zoom Range
(new)
We've received a number of requests from
readers to take shots showing the lens focal length range of various
cameras having zoom lenses. The D1's ability to accept a wide range of
lenses means you can get pretty much anything you want, from full-frame
fisheyes to 1500mm equivalent telephoto lenses. Given enough money for
lenses ;-), the D1 performs very well in this category. The following
series of shots shows something of what you can expect with typical 35mm
focal lengths, showing the field of view with two different lenses: the
Nikon 17 to 35mm f/2.8 zoom and the Tamron 28 to 300mm. (Apologies for the tilted subject on the Nikkor
shots!) The focal length numbers in parenthesis are the effective
focal lengths, given the 1.5x multiplier resulting from the size
difference between the D1's CCD and the normal 35mm film frame. These
shots also show the difference between a really high-end lens and a
perhaps slightly better-than-average consumer one: Although impossible to
compare scene elements directly, given the vast range of subject sizes,
the difference in sharpness at the 300 mm end of the Tamron lens relative
to the Nikkor across its range is fairly dramatic. The telephoto Tamron
shot also shows a slight amount of chromatic aberration that isn't
apparent in the Nikkor lens. (To head off the inevitable comments yes,
it's possible there could be some camera shake on the 300mm shot, but we
doubt it: The shot was taken from a pretty rugged tripod, the shutter
speed was 1/320 of a second, we released the shutter with the self-timer,
and the camera was operating in its "vibration reduction" mode, in which
it waits for a fraction of a second after flipping the mirror up, before
releasing the shutter. Thus, there really
shouldn't be any noticeable vibration affecting that image.)
|
|
|
 |
"Musicians" poster (1347k)
As with the
House shot, we shot samples of this using auto (1365k), daylight (1364k) and manual (1347k) white balance options. We chose the
manual setting as it produced the best skin tones and overall color
balance, although we felt it was a bit cool in hue. The daylight setting
came out just a hint warm (mostly visible in the background, which takes
on a greenish hue) and the automatic setting produced very warm results as
well. (In hindsight, this is a shot that would have been a good one to use
the manual "tweak" adjustments on the preset white balance options:
Pulling the daylight white balance setting a bit towards cooler tones
would have produced a near-perfect result.) Color saturation looks good,
as the model's blue robe looks just about right, as do the reds and
yellows of the other costumes. (The shadows of the blue robe show just a
hint of the dreaded blue/purple problem we've seen so often in digicam
images though.) Resolution and detail look great, as the D1 picks up the
subtle detail of the bird’s wings and the tiny silver threads on the
Oriental model’s robe. (Really, the D1 is pulling out all the detail
that's to be found in this lower-resolution poster.) The only noise in the
image comes from the poster itself, so we're very pleased with the D1's
performance in this category. Below is our standard image-quality
series.
Image Quality Series
Uncompressed
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Fine
Size: 2000x1312
Shutter: 1/25
Aperture:
F8
(1347k) |
Normal
Size: 2000x1312
Shutter:
1/20
Aperture: F8
(670k)
|
Basic
Size: 2000x1312
Shutter: 1/20
Aperture:
F8
(313k) |
We again took shots showing the effect of the camera's
variable sharpness settings, with the results below.
Sharpness Series
Default Value
Size: 2000x1312
Shutter: 1/25
Aperture:
F8
(1347k) |
Soft
Size: 2000x1312
Shutter: 1/20
Aperture:
F8
(1347k) |
Hard
Size: 2000x1312
Shutter: 1/20
Aperture:
F8
(1288k) |
No
Sharpening
Size: 2000x1312
Shutter:
1/20
Aperture: F8
(1159k) |
|
|
|
 |
Macro
shot (1373k)
Here again, we're really more testing a
particular lens than the camera itself. (As with the wide/tele examples
above, the message in this test is simply that you can put whatever lens on this camera you want.) These
shots were taken with our 105mm Micro Nikkor F/2.8 D lens, an
exceptionally sharp 1:1 (on 35mm film) macro lens. Given the stellar
performance of the lens, it's no surprise that we were very pleased with
the D1's performance in the macro category: It captured a minimum area of
only 0.95 x 0.62 inches (24.10 x 15.81 mm). The D1 produced the usual
superb detail, sharpness and color.
|
|
|
 |
"Davebox" test target (1176k)
The D1 does a
very nice job in this category as well. We shot with the daylight (1184k), auto (1174k) and manual (1176k) white balance settings, choosing manual
as the most accurate. Daylight produced the warmest cast overall while the
automatic setting looked slightly cool. All the color blocks look very
vibrant and accurate, particularly the bright yellow one, which many
cameras under-saturate. The usually difficult red and magenta color blocks
on the middle, horizontal color chart look nice as well, with the D1 able
to distinguish between the two nicely (many digicams have trouble here and
try to blend the colors into one). The subtle tonal variations in the Q60
chart look very nice also, with the "B" range in the pastels completely
visible and distinguishable. The shadow area of the briquettes shows a lot
of detail and only a minimal amount of noise. The D1's excellent tonal
range is perhaps best seen in the vertical gray scale though: If you play
with the image a bit in a photo editing program, you can see that the D1
actually manages to discriminate between blocks 18 and 19 at the bottom of
this gray scale, the first time we've seen this on a digicam we've tested.
(March, 2000) A job well done!
Image
Quality Series
Uncompressed
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Fine
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1176k) |
Normal
Size: 2000x1312
Shutter:
1/30
Aperture: F8
(599k) |
Basic
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(309k) |
We again took shots showing the effect of the camera's
variable sharpness settings, with the results below.
Sharpness Series
Default
Value
Size: 2000x1312
Shutter:
1/30
Aperture: F8
(1176k) |
Soft
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1111k) |
Hard
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1246k) |
No
Sharpening
Size: 2000x1312
Shutter:
1/30
Aperture: F8
(943k) |
|
|
| |
|
Low Light Series
Well,
there were two tests we really went a little overboard with, even by
Imaging Resource standards, and Low Light was one of them. In the pro
digicam arena, ISO ratings are one parameter that's hotly contested. We
shot with elevated ISO settings in several of our other tests, but this
test under controlled low-light conditions would clearly reveal more of
what the camera and sensor could really do when pushed in this regard than
any other test we'd conduct. Accordingly, we shot a massive matrix of
images, at light levels ranging from 8 foot-candles (88 lux) down to
1/16th of a foot-candle (0.0625 fc or 0.69 lux), with ISO settings ranging
from 200 to 6400. The results were very interesting. First, we found that
overall, the D1 performed exceptionally well at low light levels,
capturing very usable images all the way down to the limits of our test. -
And that's very low light indeed: At 1/16 of a
foot-candle, there was barely enough light to see our way around the
studio, without waiting for our eyes to acclimate a fair bit. First off,
we found that the automatic white balance had a hard time handling the
slightly warm cast of the 4800K photoflood bulb we use to illuminate these
tests. (A fairly typical occurrence.) We also found that long time
exposures (we went out to about 8 seconds) produced modest amounts of
single-pixel noise artifacts at the low ISO settings, while the higher ISO
settings produced more noise all the time, regardless of the shutter speed
involved. One thing that interested/excited us was how easily we could
clean up even the longest exposures shot at ISO 200 in Photoshop(TM), with
an auto levels operation, followed by a dust & scratches filter (to
kill some of the single-pixel noise), followed by an unsharp mask filter
to counteract some of the softening effect of the dust & scratches
filter. OK, that's a little convoluted, but the results are amazing, as
seen in this image (961k), shot at 1/16 of a foot-candle, ISO 200,
f/2.8, 8 seconds exposure time. (You'd swear it came from a prosumer
digicam, shooting at a light level a good 100x brighter!) I mean, that is
*amazing*!
At higher ISO
settings, the noise level increases significantly, regardless of exposure
time. Comparing performance to the prosumer cameras we've tested in the
past, we'd say that the D1's performance at ISO 1600 is better than most
consumer cameras at ISO 400, a full two stops slower. On the other hand,
the "Gain Up" options for ISO 3200 and 6400 available through the CSM
menu. are really of questionable value IOHO. Perhaps if there's a shot
that you just have NO hope of getting at a lower ISO, you might be willing
to put up with the noise levels in the "Gain Up" options, but it's hard to
imagine many photographers or situations that would justify that
(surveillance?). The noise levels are truly horrendous, particularly at
ISO 6400, and to the point that we really question their usefulness.
As noted earlier, we also
observed a nonrandom component in the D1's image noise at elevated ISO
settings: There are very noticeable horizontal striations in the image,
beginning at about ISO 1600. (At ISO 800, the effect is almost invisible,
and by ISO 3200, it's quite severe.) Interestingly, this horizontally
oriented noise appears to be independent of the individual color channels,
expressed almost entirely as a luminance artifact. (Our guess is that it
has to do with data readout, either from charge-transfer efficiency
variations, or by noise sources in the readout amplifiers.) Much has been
made of this effect in some internet discussions, and its nonrandom
character does indeed make it much more noticeable than a purely random
noise distribution would be. Not having had other ISO 1600+ capable
digicams through our labs to test, we can't comment on how the D1 fares
relative to its high-end competition, but we remain quite impressed with
its low-light operation relative to the consumer-level cameras we've seen
in the past.
OK, enough
analytical blather: The proof is in the pudding (printing?) - Herewith the
mondo low-light matrix:
Low-light/Variable ISO exposure series
8fc, ISO 200
(1220k)
Size:
2000x1312
Shutter: 1/8
Aperture: F2.8 |
8fc, ISO 400
(1226k)
Size: 2000x1312
Shutter:
1/15
Aperture: F2.8 |
8fc, ISO 800
(1100k)
Size: 2000x1312
Shutter:
1/30
Aperture: F2.8 |
8fc, ISO 1600
(1289k)
Size: 2000x1312
Shutter:
1/60
Aperture: F2.8 |
|
|
4fc, ISO 200
(1166k)
Size: 2000x1312
Shutter:
1/125
Aperture: F2.8 |
4fc, ISO 400
(1238k)
Size: 2000x1312
Shutter:
1/10
Aperture: F2.8 |
4fc, ISO 800
(1117k)
Size: 2000x1312
Shutter:
1/20
Aperture: F2.8 |
4fc, ISO 1600
(1278k)
Size: 2000x1312
Shutter:
1/40
Aperture: F2.8 |
4fc, ISO 3200
(1095k)
Size: 2000x1312
Shutter:
1/80
Aperture: F2.8 |
4fc, ISO 6400
(1136k)
Size: 2000x1312
Shutter:
1/160
Aperture: F2.8 |
2fc, ISO 200
(1178k)
Size: 2000x1312
Shutter:
1.3
Aperture: F2.8 |
2fc, ISO 400
(1238k)
Size: 2000x1312
Shutter:
1/6
Aperture: F2.8 |
2fc, ISO 800
(1142k)
Size: 2000x1312
Shutter:
1/13
Aperture: F2.8 |
2fc, ISO 1600
(1061k)
Size: 2000x1312
Shutter:
1/30
Aperture: F2.8 |
2fc, ISO 3200
(1126k)
Size: 2000x1312
Shutter:
1/50
Aperture: F2.8 |
2fc, ISO 6400
(1219k)
Size: 2000x1312
Shutter:
1/5
Aperture: F2.8 |
1fc, ISO 200
(1244k)
Size: 2000x1312
Shutter:
1/3
Aperture: F2.8 |
1fc, ISO 400
(1199k)
Size: 2000x1312
Shutter:
2.5
Aperture: F2.8 |
1fc, ISO 800
(1148k)
Size: 2000x1312
Shutter:
1/5
Aperture: F2.8 |
1fc, ISO 1600
(1061k)
Size: 2000x1312
Shutter:
1/10
Aperture: F2.8 |
1fc, ISO 3200
(1122k)
Size: 2000x1312
Shutter:
1/20
Aperture: F2.8 |
1fc, ISO 6400
(1163k)
Size: 2000x1312
Shutter:
1/40
Aperture: F2.8 |
1/2fc, ISO 200
(1228k)
Size: 2000x1312
Shutter:
3
Aperture: F2.8 |
1/2fc, ISO 400
(1245k)
Size: 2000x1312
Shutter:
1.3
Aperture: F2.8 |
1/2fc, ISO 800
(1137k)
Size: 2000x1312
Shutter:
1/2
Aperture: F2.8 |
1/2fc, ISO 1600
(1047k)
Size: 2000x1312
Shutter:
1/3
Aperture: F2.8 |
1/2fc, ISO 3200
(1105k)
Size: 2000x1312
Shutter:
1/6
Aperture: F2.8 |
1/2fc, ISO 6400
(1144k)
Size: 2000x1312
Shutter:
1/10
Aperture: F2.8 |
1/4fc, ISO 200
(1039k)
Size: 2000x1312
Shutter:
6
Aperture: F2.8 |
1/4fc, ISO 400
(1195k)
Size: 2000x1312
Shutter:
3
Aperture: F2.8 |
1/4fc, ISO 800
(1167k)
Size: 2000x1312
Shutter:
1.5
Aperture: F2.8 |
1/4fc, ISO 1600
(1070k)
Size: 2000x1312
Shutter:
1/1
Aperture: F2.8 |
1/4fc, ISO 3200
(1092k)
Size: 2000x1312
Shutter:
1/3
Aperture: F2.8 |
1/4fc, ISO 6400
(1154k)
Size: 2000x1312
Shutter:
1/4
Aperture: F2.8 |
1/8fc, ISO 200
(1113k)
Size: 2000x1312
Shutter:
8
Aperture: F2.8 |
1/8fc, ISO 400
(1040k)
Size: 2000x1312
Shutter:
4
Aperture: F2.8 |
1/8fc, ISO 800
(1197k)
Size: 2000x1312
Shutter:
2
Aperture: F2.8 |
1/8fc, ISO 1600
(1084k)
Size: 2000x1312
Shutter:
1
Aperture: F2.8 |
1/8fc, ISO 3200
(1136k)
Size: 2000x1312
Shutter:
1/2
Aperture: F2.8 |
1/8fc, ISO 6400
(1166k)
Size: 2000x1312
Shutter:
1/3
Aperture: F2.8 |
1/16fc, ISO 200
(1124k)
Size: 2000x1312
Shutter:
8
Aperture: F2.8 |
1/16fc, ISO 400
(1284k)
Size: 2000x1312
Shutter:
5
Aperture: F2.8 |
1/16fc, ISO 800
(1140k)
Size: 2000x1312
Shutter:
2.5
Aperture: F2.8 |
1/16fc, ISO 1600
(1103k)
Size: 2000x1312
Shutter:
1.5
Aperture: F2.8 |
1/16fc, ISO 3200
(1097k)
Size: 2000x1312
Shutter:
1/2
Aperture: F2.8 |
1/16fc, ISO 6400
(1138k)
Size: 2000x1312
Shutter:
1/3
Aperture:
F2.8 |
Love high ISO photography?
Hate noise? Check out Fred Miranda's ISO-R noise-reducing
actions for Photoshop. Incredible noise reduction, with *no*
loss of subject detail. (Pretty amazing, IMHO.) Check
it out! |
|
|
|
|
|
Flash
Range Series
As with most pro SLRs, the D1
doesn't have a built-in flash unit. We worked extensively with the SB-28DX
dedicated external speedlight, but there was really little point in
running a flash range series in our studio, as it's capable of
illuminating subjects at distances much greater than the 15 feet or so we
can manage in our studio...
|
|
|
 |
ISO-12233 (WG-18) resolution test (1100k)
This is the
second test we really went hog-wild with. We did a *very* rough "shootout"
at the Spring PMA 2000, and were roundly criticized for undertaking such a
poorly-controlled test. Questions were raised (and rightfully so) about
lens type on the different cameras being tested, subject positioning &
lighting, etc., etc. With that in mind, we wanted to prepare for a really conclusive set of comparative resolution
tests on professional digicams. With that in mind, we used the D1 as a
test platform for evaluating the performance of various lenses in our
personal collection, as well as those provided to us by Nikon for the
purposes of our tests. The objective was to both check out how the D1
behaved with a fairly wide range of optics on it, and also to settle on a
"standard" lens to use for future tests. As it turns out, we chose our own
Micro Nikkor 105mm f/2.8 D macro lens as the basis for future comparisons:
We'd heard that this lens was tack-sharp across the full field of view,
and full range of focusing distances, and our testing certainly seemed to
confirm this advance billing. The 105mm test images were razor-sharp, to
the point that we were comfortable declaring it the winner amongst the
variety of lenses we tested. For the sake of sharing as many of our
results with the community as possible, we're including an unusually large
set of resolution test images here from the D1. Our "official" resolution
numbers for the D1 are based on the results from the Micro Nikkor 105mm
lens, but we include links to tests from all the other lenses for all the
glass fanatics out there (like us), who are interested in knowing how some
of the other lenses performed.
As to the specifics of the test, we "called" the D1's
resolution as being 800 lines per picture height in both vertical and
horizontal directions, although there was a slight amount of aliasing
before that point was reached. Significant detail was visible all the way
out to 900 lines per picture height, but aliasing became progressively
greater. The D1's in-camera sharpening is fairly subtle, so the res target
images had a slightly soft appearance relative to some other cameras. (We
were in particular comparing images to those from Nikon's own Coolpix 990,
which has a somewhat heavy-handed sharpening applied internally.) We tried
our usual trick of applying fairly strong unsharp masking in Photoshop
(300% at 0.5 pixel radius), with this (747k) result: We felt it created a much
crisper-looking image, without introducing any halos or significant
artifacts.
105mm Quality series, default
sharpness
Uncompressed
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Fine
Size: 2000x1312
Shutter: 1/40
Aperture:
F6.7
(1101k) |
Normal
Size: 2000x1312
Shutter:
1/40
Aperture: F6.7
(658k) |
Basic
Size: 2000x1312
Shutter: 1/40
Aperture:
F6.7
(341k) |
105mm Sharpness
series
Default
Size: 2000x1312
Shutter: 1/40
Aperture:
F6.7
(1101k) |
Soft
Size: 2000x1312
Shutter: 1/40
Aperture:
F6.7
(1069k) |
Hard
Size: 2000x1312
Shutter: 1/40
Aperture:
F6.7
(1145k) |
None
Size: 2000x1312
Shutter: 1/40
Aperture:
F6.7
(966k) |
Default
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Soft
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Hard
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
None
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
One of the lenses we received from Nikon to test
with the camera was the new 17-35mm f/2.8 AF-S lens with the "silent"
internal focusing motor. Very fast focusing, very sharp lens, although
we'd give a very slight sharpness edge to the 105mm above. Here are some
shots from the 17-35mm lens:
17-35mm (Shot at 35mm) Quality series, default
sharpness
Uncompressed
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Fine
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1111k) |
Normal
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(651k) |
Basic
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(333k) |
17-35mm (Shot at 35mm) Sharpness series
Default
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1111k) |
Soft
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(1070k) |
Hard
Size:2000x1312
Shutter: 1/30
Aperture:
F8
(1153k) |
None
Size: 2000x1312
Shutter: 1/30
Aperture:
F8
(956k) |
Default
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Soft
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
Hard
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
None
NOTE: TIFF file!
Download to disk to view in imaging program.
(7747k) |
The other new Nikkor lens we tested was the 85mm
tilt/shift perspective control lens. We shot only a single image of the
res target with it, at f/16, with the default sharpness setting Clearly an
excellent lens, although to our eyes also clearly a slight notch softer
than the 105mm above. Click here (1157k) to see the test image.
The Nikkor 50mm f/1.4 manual
focus lens is widely recognized for its sharpness at medium aperture
settings: Here's (1048k) a shot taken with it, shot with an
aperture of f/8.
Just for
kicks, we shot an image with the cheapie Nikkor 35-85mm f/4-5.6D lens that
came with our Nikon 6006 prosumer SLR, with the lens zoomed to its 35mm
setting. While still much better than the typical prosumer digicam lens,
the 35-85 shows here
(1147k) why it sells for a fifth the cost of the 105mm Micro: Not
only is it softer (actually, looking much like the 85mm tilt/shift at
f/16), but it also begins to show some chromatic aberration in the extreme
corners.
Finally, we stuck
our "vacation" lens on the D1, a Tamron 28-300mm ultra-ratio zoom. This
shot was taken at a focal length of 120mm, again at f/8.0. This shot (1101k) clearly shows why pros look down their
noses at these "all-in-one" zoom lenses: While convenient, this lens at
least definitely doesn't have the critical sharpness of lower-ratio zooms
or single focal-length lenses. (But when using one of these means the
difference between packing along one lens instead of 4 or 5 on a day hike,
we'll opt for the "vacation" lens any time!)
|
|
|
 |
Viewfinder accuracy/flash uniformity
We found the D1's SLR optical viewfinder to be
about 96 percent accurate at both wide angle (247k) (17-35mm lens at the 35mm
setting) and telephoto (151k)(105mm Micro Nikkor lens) focal lengths.
(This is a typical figure for pro SLR cameras, although we're really
partial to full 100% viewfinders, as found in the Nikon F3 film-based
SLR.) As with other lens-related measurements, it doesn't make a lot of
sense to talk too much about lens distortion, given that it's a
characteristic of the lens and not the camera, and in the case of the D1,
you can put any of probably a few hundred different lenses on it. One
thing that became very clear in working with the D1 and various lenses was
just how big a difference in quality there is between the lens on a $999
digicam, and one costing that much or more for just the lens alone! This
will probably surprise no one, but the magnitude of the differences we saw
frankly surprised us: The high-end Nikkor optics showed almost no
geometric distortion, exceptional sharpness corner-to-corner, and no
discernible chromatic aberration. (Now, if only prosumer digicam lenses
could advance at the same pace as the CCDs...) |
|
| |
