Since NA determines the smallest resolvable object, and the magnification magnifies that object, a pair
of objectives with the same NA, e.g. 20X and 40X, will provide the same resolution. Magnification can
be compensated by use of zoom and box size. In fact, a 40X objective would only provide 1/2 the field
area of a 20X objective. However, use of a 20X lens with a 2X zoom would give the equivalent of a
40X objective in magnification and field size.
6.12.1 Sampling frequency and the Nyquist Theorem.
The Nyquist Theorem holds that any signal must be sampled at minimum of 2.3 times its frequency in
order to record it without loss of information. This means that the resolving power of a given objective
lens will not be recorded unless the sampling interval is at least 2.3 times smaller. If a lens can resolve 2
objects separated by 0.460 µm, then it must be sampled at .460/2.3 or 0.200 µm per sample point (µm
per pixel, in the image). Sampling with a larger pixel size will cause “aliasing” and a smaller pixel size
will result in “oversampling”.
Aliasing will cause a structure to change its appearance depending upon how it is aligned with the
sampling interval. Oversampling may or may not provide additional information but will impose
additional scanning time and laser exposure. However, oversampling by 2X, or a sampling interval at
twice Nyquist (4.6 times the frequency (size) of the structures to be resolved) is necessary for maximal
resolution of structure shapes and is beneficial for deconvolution.
6.12.2 Box size.
This term refers to the number of sampling points in the scanned field. A box size of 512x512 means
that the field will be divided into 512 samples along each of 512 lines. If the lines in the scanned field
are 300 µm long, then the pixel size is about 0.6 µm. If the box size is increased to 1024 x1024, then
that 300 µm is now divided into 1024 samples, each approx. 0.3 µm in size.
6.12.3 Zoom.
Zoom controls the size of the region scanned in the objective’s field of view by limiting the distance
over which the laser beam is deflected by the scan mirrors. The scanned region is divided into a number
of sampling points determined by the box size. If the full field of view at zoom = 1 is 300 µm and
divided into 512 sampling points per line, then each pixel represents 0.6 µm. But, if the zoom is
doubled, then line is only 150 um in length and each pixel represents 0.3 µm. A zoom = 2 combined
with box size increased to 1024 produces pixels of only 0.15 µm.
6.12.4 Considerations for increasing resolution.
1. Smaller pixels require obtaining signal from smaller volumes in the sample which, in turn, possess
fewer fluorophores to generate signal.
2. When zooming, the same amount of laser power is now applied to a smaller region, which is likely
to increase photobleaching.
3. Higher resolution is also dependent upon sample preparation, which may be a limiting factor.
4. Increasing the box size increases the size of the files. A 16-bit image of 512x512 is saved at 0.5 Mb,
whereas an image of 1024x1024 contains is 2.1 Mb.
Olympus Fluoview-1000 User’s Guide
V.M. Bloedel Hearing Research Center, Core for Communication Research
Center on Human Development and Disability, Digital Microscopy Center
May 11, 2011 38
To Next Page
Loading...
