Full frame lenses resolution on smaller sensor question.

I have a bunch of top quality 35mm AF fixed-aperture Nikkor lenses
which I use on my F4. They vary from a 20mm f2.8 to a 600mm f4. I
would like to purchase a DSLR and would prefer a full-frame digital
Nikon for lens compatibility but until that time, I am seriously
considering the new D200.

Question 1:

Would using the smaller sensor on the D200 limit the potential
practical resolution obtainable by these pro lenses even though we are
using the central portion of the field where the resolution is best?
Of course, the actual resolution of a lens is independent of the sensor
size, but a full frame would squeeze in the most line-pairs of
resolution into a frame. Cutting the sensor size would reduce the
number of line-pairs of resolution you can pack into your image. Do
you get what I am saying?

Question 2:

Is the low light sensitivity better on full frame sensors like those of
Canon or is the D200 noise levels at corresponding higher ASA, say 400,
comparable? Is this a real factor or can software image processing
correct for these differences between these models?

Question 3:

Should I wait and pray Nikon will come up with a full frame DSLR???

Thanks,

Tien

 

Probably you are already aware that switching from film to digital
means a lot of changes in your workflow, and any DSLR body will to
some extent make you change the way you work with respect to a SLR
body. So there are a lot of other factors involved in this switch than
lense performance. Also, do you prefer to shoot pictures or count
line pairs? On to your questions:

I don't know about your specific lenses, but the D200 sensor probably
allows a resolution close to that provided by good Nikkors (especially
long tele). On extreme wideangle Nikkors, resolution should be limited
by the lense itself. Of course, a smaller sensor area gives you a
smaller number of line pairs across the image. However, most DSLRs
have a lowest sensitivity of 200 ISO, while high resolution colour
film is around 50 ISO. So, if you compare the resolution and
noise/grain of a D200 with 200 ISO colour film, the D200 wins hands
down. Diffraction at small apertures is more of a problem with a small
sensor, but this is offset by the apparent increase in FL and DOF of
the lenses. So it all becomes a matter of changing the way you work,
rather than a clear better/worse judgement. You might even find that
you need to keep both film and digital, and choose between these
depending on the job at hand.

A full frame sensor receives twice the amount of light, compared to
current Nikon DSLR bodies. This difference does matter, but probably
its effects are within the range of variation of different sensor
types and models. In-sensor and in-camera post-processing also plays a
role. So a precise answer is possible only if you compare specific
models in a lab. Whether the lab results matter in practice is also
open to question.

My recommendation: if you wait for the perfect DSLR body, you will
wait forever. The D200 is a very good body. If it gives you what you
need today and in the near future, go for it. The only extra thing to
get is an extreme videangle prime or zoom, because all your present
lenses in practice will increase their focal lengths 1.5 times,
compared to film (which is good for teles).

 

It is my opinion that if you desire the maximum depth of field in your
photography, then the smaller sensor sized digital cameras appear to
have an advantage in low light. Many folks may disagree with my
opinion but after owning several digital cameras and wondering why the
smaller sensored cameras perform better in low light, I have come to
the conclusion that there is a tradeoff between depth of field and low
light photography. If you are going to be shooting landscapes, then
this would not be a problem but if you are trying to shoot subjects at
close range in low lighting conditions, then I suggest you make a few
tests for yourself. Purchase a memory card and test the various
cameras before you purchase them. Memory is getting cheaper but the
cameras can put a serious dent in your wallet. Once again, this is my
observation. Test it for yourself.
---------------------------------


Question 2:

Is the low light sensitivity better on full frame sensors like those of
Canon or is the D200 noise levels at corresponding higher ASA, say 400,
comparable? Is this a real factor or can software image processing
correct for these differences between these models?

 

Here is a detailed review showing that the D200 has no significant
extra noise over the 5D up to ISO 800:
http://www.cameralabs.com/reviews/NikonD200/index.shtml

I was in a similar position and bought a D200 together with a 12-24DX
zoom to regain the wide end of the range and have not looked back. You
will also gain an effective 900mm f/4 lens. I would not wait for a FF
Nikon as there will not be one for a while and the only difference is
in >ISO800.

 

That would be up to ISO 400. There is significantly more noise at 800 (and
the text notes that), but I also notice that either they shot a different
target with the Nikons, they underexposed the Nikon shots or they
overexposed the Canon shots, since the Canon shots are significantly
lighter, thus showing more noise. And the 5D, although the captions show it
to have only a 100 lpph advantage over the D200, seems much, much sharper in
the ubiquitous "ruler test."

 

In message <>,

That's not how I read that test at all.

Did you notice that the Nikon noise samples are all darker than the
Canon ones? That hides their noise. The average *DISPLAYED* levels
have to be the same, with the same manual exposure settings, to
accurately compare noise between two cameras.

It seems they are JPEGs out of the camera as well; the Nikons have a
noise reduction option, and they don't tell us if it is on or not.
--

 

In message <[email protected]>,

Yeah; I don't know how they made those decisions; I drew the line at
very different points than they did. I draw the line where the lines
and lightness between them show any aliasing or weighting problems at
all.
--

 

The text says they used default sharpening. I don't know how the
default compares between the 5D and D200. 100 lpph is consistent with
the difference in the number of pixels.

Popular Photography described the D2X saying "no other DSLR we've
tested controls its noise at higher ISO's as well as the D2X" and this
test says the D200 is better than the D2X up to ISO 800. I maintain the
the noise benefits of FF sensors are limited unless you do a lot of low
light shooting.

The original question was whether someone with a range of pro Nikkor
glass should get a D200. My answer is a solid "yes".

 

I reviewed this website and thank you all for your input. Having
invested a small fortune on quality glass, it offends my sensitivities
that the APS size sensor cannot take full advantage of that investment.
The review above demonstrating the obvious superiority of the 5D just
makes me long for a full-frame Nikon even more. It appears to me that
in the end, the number of photons is important and, as my wife keeps
telling me, size does matter. (

Tien

 

IIRC the 5D sharpens less, but needs it less because of a weaker AA filter.

This was also the issue that named the Canon 5D "Camera of the Year," wasn't
it? But, I agree, the noise benefits are limited at ISO 100-200, but 400
isn't all that unusual, nor is 800 where the difference starts to show.

And I can't argue with that, either... ;-)

 

Not here. Looks like they have to refine their methods.
http://www.cameralabs.com/reviews/Canon350D/page4.shtml

Also, what kind of noise reduction option does Nikon have?
Is it the kind of overall noise reduction at high ISO, or is
it for noise reduction with exposures longer than 1 second?

 

In message <>,

It has both, AFAIK; the former, an option for JPEGs; the latter, an
option for all files.
--

 

Q1...yes, in history of photography, smaller frame size always puts
more demand on higher resolution per mm to be provided by the lens.
But at the same time, larger format frames provided better total image
performance because the subject occupied more space on the film (for
better tonality) and the film image had to be magnified less to achieve
final size,
so that larger format lenses could beat 35mm on ultimate detail and
tonality.

Q2...a smaller pixel is inherently more noise prone than a larger
pixel. But software plays a real role in noise, which is why Canon
images are so superb in low noise at high ISO compared to other brands.

Q3...your guess is as good as ours about how soon Nikon can find a
source of large sensors. There are only a handful of manufacturers who
can make them larger than crop format sizes, and some of them are power
consuming technologies vs. the relatively low power requirement of
Canon CMOS sensors.

 

In message <>,

Where did you find out about this alleged software (firmware?) noise
reduction?

Don't you think they'd have removed the banding if the were cleaning the
images in firmware?
--

 

Yes and don't sweat it. Unless you regularly make to the limit blowups
of your images, this won't matter much. If you intend to go no further
than about 21.5 x 14.5 (180 dpi printed or upsampled to 300), then it
won't matter at all.

Look at the graphs at dpreview regarding noise for the various sensors
of the cameras. Generally, in a given generation of sensors (a 1 or 2
year cycle), the least dense sensor arrays have the least noise.

One definition of God is that he is the guy who ignores your prayers.
(All indications are that Nikon will not go FF).

Cheers,
Alan.