This code was written to extend the functionality of two Imagenex 881A sonars; an "imager" and a "profiler". The packaged software did not record the full (base) data stream. So I wrote a complete acquisition and visualization program. I separated the software into a client and server side because the instruments were deployed at a remote cabled observatory ( Kilo Nalu -> HIOOS).
The software worked great. I remember sitting on my father's couch late at night, viewing the real-time feed and modifying the acquisition settings. I was in Montana and the sonar was 10 m deep in Hawaii. It ran for several months, at which point the cable connector corroded past the water housing and one of the instruments was completely ruined. Mixed metals shouldn't be used in seawater.
The resulting data was disappointing. The independent azimuthal drive wasn't stepping in regular intervals. Even with correction, this created some bad artifacts in the data. This could be ironed out easily enough with some testing. Lesson learned; field tests before deployment! Too much pressure to get the instruments out and not enough time budgeted for development. Of course, sometimes the time constraints are outside our control.
A small amount of algorithm work was required to extract the reflected surface from the echo amplitudes. This was fairly easy since the data was relatively noise free.
These sonars require considerable knowledge on theory of operation. There are ~10 operation settings that must be correctly configured for optimal data. Of course, all scientific instruments work this way. Solid understanding plus manual adjustment (almost always) works better than automatic control. Still, I'd love to write an automagic calibration function... or even a "setup wizard" to simplify the setup process.
| 
