The initial decision regarding optical data acquisition 

 relates to the choice between movie film and video tape to record 

 the objects or events of interest. Except for special projects 

 or for situations that require the highest possible resolution, 

 film is far too expensive a medium for recording an entire dive. 

 Video tapes, on the other hand, are inexpensive, available for 

 instant replay, and reusable. Hours of behavioral observations 

 can be permanently recorded or edited immediately. Furthermore, 

 video information is coded electronically and viewed with 

 computer compatible equipment. 



Translation from the NTSC video format to RGB computer 

 signals is far easier than translation of analog 16 mm 

 photographs into the digitized format required for computer 

 analysis. The resolution of electronic images is, of course, 

 limited by the design of the video camera, recorder, and 

 monitor, but the current level of resolution is now acceptable 

 for many purposes and the level of resolution in video systems 

 improves each year. 



3-D Video Optical Systems 



One may record 3-D information underwater via an optical 

 system that is either single or dual band. Single band formats 

 are usually simple add-on mirror or prism devices which can be 

 used with standard cameras and projectors. They are relatively 

 inexpensive and avoid the problems of synchronization and optical 

 calibration, yet some of these simple systems have serious 

 inherent disadvantages. We do not use a single band system (a 

 single camera) because this approach creates a 3-D image by using 

 field dividing devices that produce a peculiar vertical field, 

 that have fixed interaxial distance, and do not work well with 

 wide angle lenses (Lipton, 1982). Because of the optical 

 properties of water we often use wide angle lenses to collect 

 data underwater. Furthermore, "...field-dividing devices 

 employing a prism or mirror optic placed over a single objective, 

 known as frame or field dividers (mistakenly called beam 

 splitters), often produce unwanted asymmetry of illumination 

 (vignetting). Since all lenses have illumination intensity and, 

 generally, correction of aberration - in terms of conventional 

 point symmetry, symmetrical about the center of the frame or 

 field - any superimposition of such a split field must produce 

 fields that are matched in terms of illumination and aberration. 

 This is why it is preferable to use dual objectives in 

 stereoscopic photography and projection systems" (see Lipton, 

 1982). 



We currently use two different three-dimensional optical 

 systems for mid-water research in manned submersibles and for 

 shallow water research by scuba divers. Both are color video, 

 dual band 3-D optical systems. For the self-contained scuba 

 diver operated system, two video cameras are mounted in parallel 



309 



