Re: Noise levels as a function of pixel size

Ilya Zakharevich wrote: 
[A complimentary Cc of this posting was sent to
Roger N. Clark (change username to rnclark)
<username@xxxxxxxxx>], who wrote in article <43AA38D4.5080704@xxxxxxxxx>:

focus plane.  The sensor which will collect ALL the information
available in the focal plane should have the step 1/2 of this.  This gives

lambda/(4*sin(phi)*n)

There are several factors that are being ignored in this discussion.

You are quoting someone else, not me. 

Sure.  Ignoring the irrelevant factors, and keeping all the relevant
is the key of the "scientific approach"


Yes.  You stated "For lambda=0.55 micron, and phi close to 90
degrees, one gets the absolute limit for pitch of 0.086 microns."

How does a 0.55 micron photon fit into a 0.086 micron pixel?
Question: what critical part have you ignored?


1) To do color photography, you need to focus the visible spectrum
onto the pixel.  That means from about 0.38 to 0.7 microns.


The estimates are for capturing the information at 0.55 microns.  They
will, or course, work for longer wavelengths than this.  At shorter
wavelength some information is lost; however, note that lenses usually
have a very sharp decrease of performance near the violet (this is the
reason for purple fringe); so maybe the information is not there in
the first place ;-).


Yes, the information "is there."  The simple fact is that designing
lenses for photography (e.g. wide angle to telephoto) over the
whole visible spectrum is not easy.  The perfect lens has not
been made for even high end consumers.  I work with NASA
optical engineers, some of the top ones, and never has one
of them said, we could use the perfect optical system.
It doesn't exist, even in theory (at least if the theory
is complete enough).  E.g. gougle Pantazis Mouroulis


2) You can't use pixel sizes less than the wavelength of light
(0.7 microns for color photography).  The equations for diffraction
spot size you have been using are for conditions resulting in larger
than the wavelength of light for the diffraction disk.  When you
approach the wavelength of light (this means above the wavelength),
more complex equations will apply.


I'm afraid you do not know what you are talking about.  It is the
Maxwell equation; it works for any scale (well, until about 1e-16 cm,
when quantum fields effects appear).

Or do you know some better equations? 

Try reading a book on advanced optics that includes Fresnel diffraction,
rigorously treated.  Your insulting tones show you are grasping at
straws.  How many scientific papers on optical design have you
published?  I have; see my bio on my web site.  I have also designed
and ground my own optical components, then used those optics for a book.

If you think you can make a 0.55 micron photon go completely into
a 0.086 micron pixel, and not also into the adjacent pixel, publish it.
It will be another one of your Nobel prize winning ideas.
Honestly, you have so many, I don't know why you
hang around here.

In microscopy, for resolution below the wavelngths of visible light,
people generally use electron microscopes.

2a) As pixel size becomes smaller, other problems arise in the
sensor, including leakage of electrons into adjacent wells.


Sure.  This should be taken into account.  Actually, this affects MTF,
so it IS already taken into account...


This is an absurd statement with no basis in fact.

Smaller pixels have some technical problems; larger pixels have some
other technical problems. Both are resolvable, 

Great.  Patent it.  Another winning million+ dollar idea.

3) You state that you can get these fast lenses with great performance
in the small form factor.  This may be true for monochromatic
wavelengths, but fast f/ratio lenses in the small form factor
notoriously have poor chromatic aberration.


Chromatic aberration is visible on resolution charts.  Please try to
find some for the lens I investigated (one of KM A200).  [Do not know
about other lenses for small format.]  DPreview shots show MUCH
less visible chromatic aberration with this lens than with 35mm primes.


Resolution charts are generally low contrast, especially compared to
real world imaging.  Chromatic aberration often does not show on
such charts unless it is extremely bad.

See (FYI purple fringing IS chromatic aberration):
http://www.dpreview.com/reviews/konicaminoltaa200/page6.asp
"As noted in our other eight megapixel digital camera reviews
this particular CCD in combination with compact wide angle
lenses does appear to lead to purple fringing. However on the
DiMAGE A200 fringing, while visible in pretty extreme shots
(generally those with overexposed highlights), isn't as strong
as we have seen in other cameras and is eliminated once the
the lens is stopped below F4."

The biased resolution charts at:
http://www.dpreview.com/reviews/konicaminoltaa200/page11.asp
(biased because it is not true MTF as in camera sharpening
modifies the results on all such tests at dpreview for all
cameras) says:
"Disappointingly the fine detail capture is still not that
clean, with visible moire from about 1400 lines."
Note, that on dpreview, they use lines from an old
definition: black+white lines over the picture height.
So 1400 "lines" is 700 line pairs per picture height.
We don't know the MTF of those lines because of the in camera
sharpening algorithms.  The test should be done on raw
output, linear transform with no processing to understand
the true MTF of the system.

Cons:
http://www.dpreview.com/reviews/konicaminoltaa200/page13.asp
"Soft images (not just as a result of poor AF)"
"Lens being stretched past its resolution capabilities?"
"Dynamic range issues with high contrast scenes (better shooting RAW)"

I see nothing that indicates this camera can compete with
a quality DSLR with larger pixels, nor that the lens is an
outstanding design that helps your case that small lenses
coupled to small cameras can equal larger cameras.

Merry Christmas,

Roger
.

Relevant Pages

  • Re: Noise levels as a function of pixel size
    ... but most probably still outside of practical range of lenses. ... You can't use pixel sizes less than the wavelength of light ... The relevant result to this discussion is shown in Figure 21, which concludes a 50% error of correct detection per pixel at a signal-to-noise ratio of 5. ...
    (rec.photo.digital)
  • Re: Measure distances w 1mm precision
    ... For a good camera system you need at least 3000-5000 Euro (camera and lenses). ... With a 20 mm lenses and a sensor with 1280 x 1024 pixel you will have an field of view of 850 mm at a distance of 2 m. Lenses with a smaller focal length are also really expensive and not accurate enough for measurement systems. ...
    (sci.image.processing)
  • Re: Noise levels as a function of pixel size
    ... > onto the pixel. ... wavelength some information is lost; ... > sensor, including leakage of electrons into adjacent wells. ... > 3) You state that you can get these fast lenses with great performance ...
    (rec.photo.digital)
  • Re: A dSLRs fastest shutter-speed is really no faster than its flash-sync speed.
    ... 2)very fast AF in most lighting conditions, ... - the EVF of a small-sensor camera can be brighter, although the DSLR offers finer control of framing. ... ZLR than having to change between lenses. ... E.g. if a small pixel camera, say 6 megapixels with 2-micron pixels ...
    (rec.photo.digital)
  • Re: Nikon and Megapixels
    ... top camera has about half what Canon does. ... Very few lenses - even very expensve pr ones - can exeed the resolution of a 12 million pixel sensor most of the time. ... But double the pixels on even a full frame sensor would stretch the capability of most lenses other than in lab controlled conditions. ... Apart from the noise issue there is no point in having bigger files without a better image quality. ...
    (alt.photography)