D2x and D200 resolution

[OleThorsen]

I have a stupid question regarding resolution.

With the different tests and comparisons between D2x and D200 showing up now more and more, you can see the D2x has a little more resolving power showing little more details.

When many people say, that's quite obvious that the D2x has more resolution per pixel. What does people mean by that???

In my mind, and perhaps I'm slow here, this statement should mean

D2x absolute resolution/12.4 MP is higher than
D200 absolute resolution/10.2 MP

and I really doubt this is true, or at least you have to measure the absolute resolution in test pictures, before you can state this, and I haven't seen this done yet.

But as an example, if the D200 in a given test showed the same detail level as a D2x sample, it would be obvious that the D200 would have more resolution per pixel than the D2x.

Can someone please clarify this for me?

Regards
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

 

[Norwen]

yeah... more resolution per-pixel is obviously impossible. However, better quality pixels is possible. That's probably a bayer filter thing, smoothening each pixels (each group?).

As an example, the D2H is sharper and better then the D100 altough it has less pixels.

From the examples I've seen, the D2X is truly magnificient against the D200. With the possible D2X price drops this body will be a must-own, no matter the higher price.

 

[OleThorsen]

upon if you also include

Absolute Resolution/$ and
Dynamic range/$
ISO Noise/$

in the equation, doesn't it?

Regards
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

 

[Norwen]

And Grip $
And Construction $
And durability $
And Status $
And Shutter construction $
And shutter lag $
...

The market will reform itself, leveling the D2X where it should be. If it is 2500$, then so be it.

All I know is I really don't care But the D200 is a modern breakthrough camera. And that's great for us.

 

[OleThorsen]

I didn't ask which preference people has to either D2x or D200, did I?

It was actually you, that brought this issue up, and to be frank, I do not care what your preference is, because it should be quite obvious reading your postings.

I asked a simple question about resolution per pixel, so have you anything to add to that?

Regards
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

 

[nik1024]

It's a simple mathematical thing.

IF both cameras (or any cameras) are capable to record directly analog image from real life into pixels then more pixels would automatically mean better resolution.

In reality, conversion of analog light signals into pixels is a very complicated matter which includes approximation and interpolation, so a camera with less pixels can produce much sharper image in less pixels than a camera with more pixels if its recording method is inferior in some ways.

Which can be clearly seen in many cases D70 produces VISIBLY sharper images than 20D, at least when seen at 100%. In prints it's all different matter but I would expect even 8"x11" prints from d2x to be sharper LOOKING than d200. It's not only sharpness but also contrast which has been mentioned by others d200 looks washed out compare to d2x. I'm only repeating.

 

[Norwen]

Well, if you read MY answer, it is there: Resolution per pixel is the same for all cameras. 1 pixel only records one data. Pixels per area (or size of each pixel) is different. It's all in the bayer filter strenght.

In reality, your question is nil as it's just impossible as-is.

 

[OleThorsen]

the intension with my question was obvious not to ask if it is possible to measure the resolution in one pixel.

In my mind you can talk about resolution per pixel, just as a mathematical measure to compare cameras with different Mpixels.

But those people saying the D2x has more resolution per pixel than the D200, then also says that the D2x must have more than 22% (12,4/10,2) more resolution than the D200, or what?

Regards
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

 

[AAK]

It's not only sharpness but also

contrast which has been mentioned by others d200 looks washed out
compare to d2x. I'm only repeating.

That kind of comment always baffled me. Is that "at default settings?". Is that .jpg or NEF?

Same goes for sharpness to a great extent. It's mostly a matter of what the defaults are for the camera. Of course, you also have huge latitude to customize the contrast, etc. using curves and many other adjustments, so I don't know how you compare the native capabilities of any two comparable (and well-equipped) DSLRs.

--
AAK - http://www.aakatz.com

 

[Kaj E]

The resolution of a digital camera depends of the resolving power of the lens. the strength of the anti-aliasing filter, the amount of pixels and the processing algorithms of the cam. Assuming that a sufficiently good lens is used the it can be ruled out in the case of D200 and D2x. (Compact digital cameras frequently do not require an anti aliasing filter because the quality of the lens has less resolution than the sensor).

If the resolution of the lens is higher than the resolution (pixel amount) of the sensor the result would be moire if no anti-aliasing filter was used. A anti-aliasing filter acts as a softening filter in front of the sensor to bring the resolving power of the lens and the sensor more in line and actually when successful makes the softening effect of the AA-filter just a little stronger than the resolution of the sensor. This should eliminate the risk for moire but still give good resolution for the sensor. The AA-filter is always a compromise between avoidance of artifacts and sharpness.

Sharpness is not a simple concept it includes also contrast. According to the so far published tests the D200 has a slightly larger dynamic range (about 1/2 stops) than the D2x. When JPEGs are shot and depending on the curves built into the cam the contrast in a higher dynamic range camera may appear flat and less sharp. I do not know if this explains the apparent softer pictures of the test images posted. My guess is that the effect is due to a stronger AA-filter in the D200 than in the D2x.

Some of the resolution lost in the AA-filter can seemingly be brought back with sharpening in post-production (and also in cam sharpening which has been turned off in the tests). PP gives more control of the sharpening and allows better control to avoid artifacts. The sharpening just increases contrast between adjacent pixels.

The last step determining the apparent sharpness in the camera is the algorithm applied to convert the RAW file to JPEG when shooting JPEG. To have full control of this step it is advisable to shoot RAW and see what can be accomplished in post processing. The consensus seems to be that the D200 and D2x have the same in cam RAW to JPEG algorithms.

To summarize the resolution depends on the lens, the AA-filter, the pixel amount, and the in cam algorithms (and also on apparent contrast which can be changed in PP).

Hope I made myself clear.
--
Kind regards
Kaj
C P 5 7 0 0
http://www.pbase.com/kaj_e
WSSA member

 

[Kaj E]

Should have changed the title in my post above above.
--
Kind regards
Kaj
C P 5 7 0 0
http://www.pbase.com/kaj_e
WSSA member

 

[nik1024]

I can only agree with everything you said and I was only replying to the question in the original post about 10mp vs 12mp. Resolution and visible sharpness involves many things, yes. What I meant is that the entire process of "mapping" the image into sensor and then to the pixels in the image on CF card may produce images of 6mp to be much sharper than, say, 24mp theoretically speaking, while 6mp has 4 times less pixels.

The resolution of a digital camera depends of the resolving power
of the lens. the strength of the anti-aliasing filter, the amount
of pixels and the processing algorithms of the cam. Assuming that a
sufficiently good lens is used the it can be ruled out in the case
of D200 and D2x. (Compact digital cameras frequently do not require
an anti aliasing filter because the quality of the lens has less
resolution than the sensor).

If the resolution of the lens is higher than the resolution (pixel
amount) of the sensor the result would be moire if no anti-aliasing
filter was used. A anti-aliasing filter acts as a softening filter
in front of the sensor to bring the resolving power of the lens and
the sensor more in line and actually when successful makes the
softening effect of the AA-filter just a little stronger than the
resolution of the sensor. This should eliminate the risk for moire
but still give good resolution for the sensor. The AA-filter is
always a compromise between avoidance of artifacts and sharpness.

Sharpness is not a simple concept it includes also contrast.

I always thought contrast of d70 (due to weak AA?) was the reason it produced images tad more sharp than 20d, although I may be wrong and there are probably other factors as well.

According to the so far published tests the D200 has a slightly
larger dynamic range (about 1/2 stops) than the D2x. When JPEGs are
shot and depending on the curves built into the cam the contrast in
a higher dynamic range camera may appear flat and less sharp. I do
not know if this explains the apparent softer pictures of the test
images posted. My guess is that the effect is due to a stronger
AA-filter in the D200 than in the D2x.

It may be AA and also that's why may be d200 images look tiny bit "too clean". It's not a bad thing, just different.

Some of the resolution lost in the AA-filter can seemingly be
brought back with sharpening in post-production (and also in cam
sharpening which has been turned off in the tests). PP gives more
control of the sharpening and allows better control to avoid
artifacts. The sharpening just increases contrast between adjacent
pixels.

Yes although most agree than sharpening can only bring back part of what was lost due to AA, not all of it.

The last step determining the apparent sharpness in the camera is
the algorithm applied to convert the RAW file to JPEG when shooting
JPEG. To have full control of this step it is advisable to shoot
RAW and see what can be accomplished in post processing. The
consensus seems to be that the D200 and D2x have the same in cam
RAW to JPEG algorithms.

I personally don't care about JPEG, I always do RAW and then run RAW-> JPEG batch job, it's no big deal.

To summarize the resolution depends on the lens, the AA-filter, the
pixel amount, and the in cam algorithms (and also on apparent
contrast which can be changed in PP).
Hope I made myself clear.

More than.

Thanks for detailed reply,
nik1024

-

Kind regards
Kaj
C P 5 7 0 0
http://www.pbase.com/kaj_e
WSSA member

 

[OleThorsen]

Hi Kaj

Thank you for your very detailled explanation about perceived sharpness.

I actually believe you in another thread said, that in your opinion the D2x has more resolution per pixel than D200. This means the D2x must have more than 22% (12,4/10,2) more resolution than the D200, or what?

Regards
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

 

[DavidMillier]

Also have an effect on detail

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[Bob GB]

The D2X has a linear pixel count of 10% more than the D200 (square root of 12,4/10,2). If everything else are equal, this will lead to a 10% higher resolution, which is barely detectable by the eye.

But the D2X has several other qualities distinguishing it from the D200.

• regards -

I actually believe you in another thread said, that in your opinion
the D2x has more resolution per pixel than D200. This means the D2x
must have more than 22% (12,4/10,2) more resolution than the D200,
or what?

Ole Thorsen

 

[John Yu]

--
John

 

[Starburst]

Just to add some elements in this debate from somebody that has been working with digital detectors since the 1970s

Seems to me that in this debate an important parameter has not been mention and isconfused with resolution, the parameter is: SAMPLING.

The number of pixels in a CCD gives the SAMPLING. Now, RESOLUTION, i.e. the maximum frequency or number of lines that the detector can record before confusion sets in is a different parameter.

For a perfect detector the maximum frequency or number of lines that can be resolved (the maximum resolution) is related to the sampling via de Nyquist or sampling theorem that indicates that, the maximum resolution is about half the number of samples (pixels).

In fact is closer to 40percent. I other words if you have a detector with 1000 samples or pixels the maximum number of lines the perfect detector will be able to see will be around 400. Exactly the same the same applies to a square detector of lets say 1000x1000pixels.
See: http://en.wikipedia.org/wiki/Nyquist-Shannon_sampling_theorem

Real life detectors are not perfect and on top of that, in the case of CCD recording colour images, they need three pixles to construct a single colour pixel, so we can expect a large departure from the Nyquist value above inn the sense that the resolution will be much less(smaller) than 40percent of the sampling or number of pixels. This was nicely explained by Kaj.

One more thing, the dynamical range of the D1s, D2s and D200s is larger than the JPEG standard. JPEG is only 8 bit per pixel (or 24bit colour) while the Analog-to-Digital converters in these cameras have a true dynamical range that approaches 10bits. In other words, the only way to profit from the extended dynamical range is to record in RAW.

RJT

 

[Kaj E]

Hello Ole,

What I tried to say in my reply above is that there are many factors influencing apparent sharpness. Comparing the pictures on Kamera & Bild it seems to me (subjective observation) that the resolution per pixel is less in the D200 than in the D2x. I suspect the main reason is a stronger AA-filter. This does not have to be a negative because a stronger AA filter gives a smoother more film like picture (without moire and jaggies). And as I said some of the perceived softness can be brought back in post production.

If the resolution per pixel actually is lower on the D200 than that of the D2x, the actual amount can actually be determined by a resolution test (let's see what Phil comes up with).

--
mvh
Kaj
C P 5 7 0 0
http://www.pbase.com/kaj_e
WSSA member

 

[JulesJ]

Why suddenly introduce these fasctors. They had nothing to do with the initial posters question. Why are so many posters on this forum unable to relate to what is acually asked and have to answer their own questions.
Jules

....
You wrote:
And Grip $
And Construction $
And durability $
And Status $
And Shutter construction $
And shutter lag $
...
--
Black holes do not destroy information.