Testing the Response of the Watec WAT-910BD
and WAT-910BD-LM (Linear Modified)

Tony George reported on an anomaly in the response of the WAT-910HX and Bob Anderson tested the camera on his ArtStar aparatus.
If you are a member of the Yahoo eGroup IOTAoccultations you can view Tony's message here.
However if you can not view the message, a very brief summary of the findings were;
  • The camera's full-scale response, from saturation to extinction of light, is non-linear.
  • However the camera is linear in the lower brightness range, i.e. less than 65% of full brightness.
    The former may well be a problem when observing events that, by their nature require a full-scale linear response - like Lunar Occultation events of double stars, while the latter does not present a problem when observing events like asteroid occultations. The report goes on to state that Watec can change the full-range performance of the camera, for a fee.

    It is my understanding that to date, nobody has had a WAT-910HX modified, let alone had it's response tested.

    Now, the WAT-910HX and the WAT-910BD are very different beasties physically as can be seen below...

    ... and Watec states that the performance of the two cameras are identical.
    So in my mind the questions that need answering are;
  • Is the WAT-910BD response similar to the WAT-910HX as described by Tony and Bob?
  • Does the full-range modification offered by Watec result in a linear response?
  • Is it a good idea to get a Linear Modified WAT-910BD or not?
  • And Why or Why not?

    In May 2016 I invested in a linear modified camera (henceforth known as WAT-910BD-LM), so let's find out.
    Note: I don't own a WAT-910HX, so I cant test a DB and a HX side-by-side.
    The first task was to further develop our Star-Chamber, and this took time and effort to sort out (thanks Tony Barry).
  • Star Chamber consists of a cardboard tube with the control electronics mounted on the tube itself.
  • Early experiments used fibre optics to deliver the light to the right hand tube plate. Later experiments mounted the LEDs in a bush in the tube plate with the light shining through a 0.5mm aperture
  • A Pentax 50mmf4 macro lens is mounted in the left hand tube plate and focused the light onto the test camera’s CCD chip.
  • The electronics was designed to vary the intensity of the LED over time in synchronisation with the V-Sync pulses inherent in CVBS video.

  • A video (.AVI) recording was made using;
  • AVerMedia DVD EZ Maker USB2.0 grabber
  • Virtual Dub
  • HuffYUV (lossless) compression
  • Dell Inspiron 5150 laptop
  • Tangra3 was used for analysis.

    Four Watec cameras were tested; WAT-902-Ultimate, WAT-120N(non+), WAT-910BD and WAT910BD-LM. The settings were;
  • Gamma = off (or 1.0)
  • Gain = low (identical in the case of the two BDs)
  • Exposure = set to frame rate

    As a reality check, I first tested the response of the WAT-902 Ultimate and the WAT-120N, and their responses are fairly linear.
    Note the "toe" at the bottom of the 120N's curve. This may well be unique to my old camera.

    Next is the full-range response curve of the standard WAT-910BD. This is very similar to the curve produced by Bob and TonyG, except theirs is brightening while ours is dimming.

    Next is the full-range response curve of the linear modified WAT-910BD-LM produced a linear curve similar to the 902 and 120N.
    But note: for the same settings the peak value is much less than the standard 910-BD. i.e. 3550 versus 3100... Hmm!

    To confirm the perceived loss of sensitivity of the linear modified camera compared to the standard camera, I took two sets of images of a famous open cluster - The Jewel Box - NGC 4755. Exposed at x2 and x16. All other settings were identical.

  • Here is a .zip containing a composite image comparing the WAT-120N, the WAT-910BD-LM and the WAT-910BD.
  • The old saying "You can't have your cake, and eat it too" is true.
  • Can a standard WAT-910BD be used for high quality photometry? Hristo Pavlov had demonstrated that it can. In this paper (Section 5 starting from Page 19) he used a standard WAT-910BD to do spectroscopy of Neptune. During the process he did an absolute calibration of the obtained spectra using CalSpec stars (assuming the camera response was linear). As stated in the last paragraph of page 20 the observed error in the calibration process i.e. combined differences between observed spectra of the standard stars and their spectra observed by HST was 5%.
  • Due to the superior sensitivity the standard WAT-910BD (or HX) is possibly "the king of the analogue video cameras" for observing asteroid occultation events.
  • However analysis of observations of Double-Star Lunar Occultations may be a challenge for a non-linear camera, e.g. see this typical image here. Where the field is far from flat due to glare from the nearby sunlit bright limb and from internal reflections in the telescope, and the measuring tool has a dose of unlit-limb for good measure (pun), and finaly the almost always complete lack of comparison stars makes determination of absolute stellar magnitudes all but impossible. All we can do is observe the height of the step and determine the component magnitudes by simple calculations. Easy to do with a camera with a linear response thanks to LiMovie.

    The TACOS-BD System works with either the standard WAT-910BD or the linear modified WAT-910BD-LM. The choice is yours.

    Return to TACOS BD System last updated: 1st October 2016 email Dave: d a v e 4 g e e @ y a h o o . c o m . a u