Re: Pixel size

Posted by Mark Cannell on
URL: http://confocal-microscopy-list.275.s1.nabble.com/Acceptor-Photobleaching-vs-Sensitized-Emission-FRET-results-tp4828946p5632775.html

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi All,

I think one should pay attention to the fact that this is a color  
camera implying a Bayer filter/mask. This effectively increases  pixel  
_pitch_  by a factor of 2.  Please note that it is pixel pitch that is  
critical for sampling/resolution issues _not_ pixel size.  RGB output  
for each pixel is calculated from the Bayer mask by the support  
electronics and this can fool you as to the actual sampling that is  
taking place. Fortunately, in monochrome cameras the pitch is often  
the same as the pixel size, but this is not always the case -you need  
to read the chip specification carefully.

In any case, I completely agree with other posts here: One should  
calibrate/check the overall magnification with a known test object  
because overall system magnification may not be solely determined by  
the nominal magnifications of the objective and (possible) relay lens.

My 2c

Cheers Mark


On 14/10/2010, at 9:53 AM, P. Johannes Helm wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Bonsoir, Louis,
>
> that's one of the frequent questions I am being asked by "my" users.
>
> The thing is:
> Your microscope objective and condenser will provide a certain  
> resolution,
> officially dependent solely on numerical apertures. In case of a
> transmitted light image, it will be dependent on the numerical  
> apertures
> of both, the condenser and the objective (proper Koehler illumination
> assumed). The LATERAL microscope resolution in this case is
>
> d = (1.2 * lambda) / (NA obj + NA cond.)
>
> (at which, in case of day light filter operation, lambda is a somewhat
> undefined thing, assume 500nm and / or use a panchromatic green  
> filter to
> reduce wavelength bandwidth).
>
> In case of epi-illumination, the objective will also be the  
> condenser, so
> that the denominator reduces to (2*NA obj) though, in some special  
> cases,
> you might nevertheless end up with different ray paths for the
> illumination and the detection, anyway, since the illumination ray  
> path in
> very special cases is a light cylinder around the "real" objective"
> mirrored onto the preparation (annular illumination).
>
> ("Axial resolution" resp. "depth of field" in the wide field case is a
> complicated and sometimes "debated" issue.)
>
> You might call "d" the size of a "resel" (resolution element). Which  
> is
> different from a "pixel" (picture element).
>
> Multiply the size of that resel with the magnification factor of all  
> the
> optics between the object and the chip of the camera. If you are  
> lucky,
> you have an adjustable zoom optics which allows you to adjust the
> magnfication so that the image on the chip is magnified by a factor,  
> which
> makes sure the Nyquist theorem is fulfilled, at least. So: One resel
> imaged onto the chip covers at least 2 pixels of the chip, preferrably
> more (but not too much so that you do not oversample too much,  
> loosing a
> lot of light).
>
> There is, in other words, to my mind not a straight forward answer  
> to your
> question. Unless you simply like to divide the dimensions of your  
> camera
> ship, which you can find in the manual of that camera, by the total
> magnification factor. Then, you have the pixel size in the image,  
> although
> this might not really help you unless you compare it to the resel  
> size as
> mentioned above.
>
> Best wishes,
>
> Johannes
>
>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Bonjour à tous,
>>
>> I have a tiff image, 1392 x 1040 - 4.65 um square pixels acquired  
>> with a
>> camera interline Sony , 1.4 megapixel, color(7.6mm x 6.2mm array).  
>> I use
>> a 40 x objective mounted on  a table microscope Leica DME.
>>
>> Can I find the pixel size in hte image?
>>
>> Thanks ,
>>
>> Louis
>>
> --
> P. Johannes Helm, M.Sc. PhD
> Seniorengineer
> CMBN
> University of Oslo
> Institute of Basic Medical Science
> Department of Anatomy
> Postboks 1105 - Blindern
> NO-0317 Oslo
>
> Voice: +47 228 51159
> Fax: +47 228 51499
>
> WWW: folk.uio.no/jhelm