34
Detectors
Three types of detectors are generally employed in image microscopy; (a) photomultipliers
for point sources such as CLSM, or 2-dimensional cameras such as (b) charge-coupled array
devices (CCD) and charge induction devices (CID), and (c) video cameras (Wang and Taylor,
1989; Jovin and Arndt-Jovin, 1989; Inoue, 1986). Photomultipliers of high sensitivity and
good linearity exist and the optical properties of the point source microscopes ensure a high
spatial resolution. However, the scanning mechanisms of all of the available commercial
microscope systems are very slow compared to 2-dimensional detection systems. Assuming
that many measurements are low-light level applications (fluorescence) a variety of
2-dimensional cameras have been assessed under these circumstances. Their resolution,
linearity, dynamic range, sensitivity and speed have been compared by Tsay et ai. and
Bookman (in ref 8). A summary of this and other data available in the literature shows that
slow scan CCD cameras are superior in resolution, linearity and sensitivity but cannot be used
for rapid kinetic studies. The best choice for the latter is the multichannel plate intensified
CCD although some fast kinetic studies can be performed on subarrays of the slow-scan CCD
cameras as well (Jovin et al. in 8). The newest slow-scan CCDs contain 7 micron square
pixels in arrays of 10 6 or more elements providing spatial resolution in the X-Y plane
equivalent to the optical components of visible microscopes. New CID cameras are now
available which will probably overcome the disadvantages of the slow read-out CCD arrays
since they provide nondestructive and direct subarray read-out capability (Sims and Denton,
1987; Kaplan, 1990) and may take over the detector field in the near future.
BIOLOGICAL PROBLEMS
The identification of objects in oceanography encompasses a broader spectrum of size and
diversity than any other biological discipline. One need only to reflect on the interdependence
of the world's largest creatures, the cetaceans, and their food source, plankton, to grasp the
problem of identification and classification. It has been recently estimated (Giovannoni et ai.,
1990) that < 1 % of all microbial species in the world have been characterized so far. Thus,
identification alone, not to speak of description of whole ecosystems, requires complex
analysis not afforded by simple DNA or macromolecular dyes, or even by enzymatic tests and
Detectors
Three types of detectors are generally employed in image microscopy; (a) photomultipliers
for point sources such as CLSM, or 2-dimensional cameras such as (b) charge-coupled array
devices (CCD) and charge induction devices (CID), and (c) video cameras (Wang and Taylor,
1989; Jovin and Arndt-Jovin, 1989; Inoue, 1986). Photomultipliers of high sensitivity and
good linearity exist and the optical properties of the point source microscopes ensure a high
spatial resolution. However, the scanning mechanisms of all of the available commercial
microscope systems are very slow compared to 2-dimensional detection systems. Assuming
that many measurements are low-light level applications (fluorescence) a variety of
2-dimensional cameras have been assessed under these circumstances. Their resolution,
linearity, dynamic range, sensitivity and speed have been compared by Tsay et ai. and
Bookman (in ref 8). A summary of this and other data available in the literature shows that
slow scan CCD cameras are superior in resolution, linearity and sensitivity but cannot be used
for rapid kinetic studies. The best choice for the latter is the multichannel plate intensified
CCD although some fast kinetic studies can be performed on subarrays of the slow-scan CCD
cameras as well (Jovin et al. in 8). The newest slow-scan CCDs contain 7 micron square
pixels in arrays of 10 6 or more elements providing spatial resolution in the X-Y plane
equivalent to the optical components of visible microscopes. New CID cameras are now
available which will probably overcome the disadvantages of the slow read-out CCD arrays
since they provide nondestructive and direct subarray read-out capability (Sims and Denton,
1987; Kaplan, 1990) and may take over the detector field in the near future.
BIOLOGICAL PROBLEMS
The identification of objects in oceanography encompasses a broader spectrum of size and
diversity than any other biological discipline. One need only to reflect on the interdependence
of the world's largest creatures, the cetaceans, and their food source, plankton, to grasp the
problem of identification and classification. It has been recently estimated (Giovannoni et ai.,
1990) that < 1 % of all microbial species in the world have been characterized so far. Thus,
identification alone, not to speak of description of whole ecosystems, requires complex
analysis not afforded by simple DNA or macromolecular dyes, or even by enzymatic tests and