Appendix: Special features for Scion LG-3
This paragraph assume that the reader has read through the
"Imaging" part
of the TI workbench manual.
Video frame grabbers including Scion LG-3 convert video image
frames which are updated 30 times per second. That is, the unit
of exposure time is fixed to 1/30 sec and binning operation is
not available. These features are different from those of digital
CCD cameras: flexible exposure time and variable binning factors.
To adapt these characterisics of the frame grabber to the framework
of TI Workbench, the parameters of exposure time and binning factors
are used differently. These differences and others are described
below.
Exposure time:
- 33 msec is the minimum unit. If exposure time is set to 33 msec,
just one video frame image is stored as an 8-bit depth image. If
exposure time is multiple of 33 msec (user can not set exposure time
other than multiple of 33 msec), image pixels are added for the number
of frames. In this case, bit-depth of pixels are expanded to16 bit
(unsigned) to avoid data overflow. So the image bit-depth is different
between 33 msec exposre and longer exposures. Up to 256 frames
(exposure time 7.5 sec), user do not have to worry about data overflow.
An alternative way may be averaging the frames instead of summing. I
did not take this way because storing 8-bit averaged image could lose a
subtle changes. This sum operation is performed digitally AFTER A/D
conversion of each video frame, which is different from "analog
summention" perfomed on digital CCD chips. The former, "digital
summention" is theoretically worse in S/N than the analog summention
due to the accumuration of digitizing noise.
Binning factor:
- In digital CCD cameras, analog pixel data are added horizontally
and/or vertically before A/D conversion. This is called "binning"
operation which increases sensitivity and frame rate, and, at the same
time, decreases resolution of the image. Frame grabbers have no "analog
binning" capability. In TI Workbench, pixel values are added
hirizontally and/or vertically AFTER A/D conversion. Once the binning
factore is set other than 1x1, bit-depth of the image is expanded to 16
bit due to the same reason explained in the above paragraph. In case
for very dimm and noisy situation, this "digital binning" would be a
similar effect as spacial filtering. However, the resolution is
decreased.
Continuity:
- In most of frame-shift type or inter-line type CCD cameras,
continuous mode of sequential recording or focusing mode guarantee
"continuous" data acquisition. The focusing mode and continuous mode of
sequential recording of frame grabbers do not guarantee "continuous"
acquisition. Host computer (Macintosh) must catch up every frame update
(1/30 sec) to ensure continuous image frames for frame grabbers, and
this was not available even for the simplest "33 msec exposure, 1x1
binning" mode (in case for beige Macintosh G3 233 MHz). However, the
lapse time stamped to each image frame in sequential recording mode is
accurate. So that, the time course plot is also accurate (even after
sent to Igor). User need to aware that the continuous mode does not
necessarily taking "continuous" image frames or even taking constant
interval image frames.
Key operations in focusing mode:
(refer to here)
- up and down arrow keys to change brightness, and left and right
keys for contrast.
TTL output to drive shutter:
- TTL output from LG-3 can be used to control TTL-controlable
shutter devices. Refer to the
wiring section for connection with a Uniblitz shutter.