OPTINAV

1414 127th Place NE #106

Bellevue, WA 98005

Tel (425) 891-4883 

© 2016

Bob D
Oct 9, 2018

Version 4.202

0 comments

This version fixes two bugs. One, introduced in Version 4.200, prevented frequency adjustments using the scroll bar when BeamformX was paused. The other caused a null pointed exception sometimes when clicking inside the Spectrum window.

New Posts
  • Bob D
    Sep 18

    The site www.optinav.info has been developed as a repository for OptiNav resources. It has some videos and download links, including the latest version of BeamformX. The page for BeamformX version history, www.optinav.info/ftp.html , has been been redirected to optinav.info. You may want to update any links to remove the /ftp.html part. The version history is still accessible from optinav.info.
  • Bob D
    Sep 4

    A new version of BeamformX has been posted on http://www.optinav.info/ftp.html . This version is slightly different from 5.100, posted yesterday, but more detailed information that applies to both is given here: For some users, the most important change is that some equalization has been applied to attempt to compensate for frequency response of the microphones in the ACAM 100 and ACAM 120. I based the corrections on the published datasheets for the two microphone types. Significant changes occur below 200 Hz and above 5 kHz. Some extrapolation of the published curves was required.  The array shading and the overload light thresholds have changed slightly. When there is an ROI in the display, the near-real-time narrowband spectrum shows the result of beamforming to the point in the center of the ROI. The method for this beamforming has changed from delay and sum beamforming to conventional, frequency domain, beamforming (FDBF).  This gives much better beamformed spectra. Note that the limitations of FDBF apply: poor dynamic range and low frequency resolution and treatment of coherent sources. Better spectra can be obtained by doing a frequency sweep with Quantitative Beamforming, but not yet in near real time. The old delay and sum calculation is still used for the audio output. The near-real-time spectra creates the opportunity to compute a near-real-time dBA value for the center of the ROI.  If Settings/A-weighting is selected, then this is shown in the upper right corner of the Display as the "20 Hz - 20 kHz" value.  A number was shown there starting from Version 5.003, but it always gave the median of the microphones.  Now it can be focused by creating the ROI. The old "20 Hz - 20 kHz" dBA was found to be badly wrong in some cases. Several contributors this problem have been identified and corrected, including the microphone frequency response and some software bugs. Even with the corrections, the ACAM/BeamformX system is likely to give different results from those of a quality sound level meter. Some of the reasons are good. A version of BeamformX called DroneHound is being developed to identify drones.  A new version has been posted, but it is not ready for real use. Known bugs/requests that are not yet fixed: Sometimes clicking on the spectrum does not adjust the tuning frequency.  This happens so rarely that it is difficult to troubleshoot. The Spectrogram shrinks in between runs. There is no scripting language or way to load new data without exiting BeamformX. The language strings need updating. There is no way to reopen the Spectrogram or Spectrum windows after closing them. Higher resolution optical images are possible. BeamformX does not look like a Windows program.
  • Bob D
    Jul 6

    A variation on BeamformX called DroneHound has been posted www.optinav.info/ftp.html Sounds are shown in the display as open circles if they are not classified as drones or angry red drone shapes if they are. The tabular results give the bearing, overall sound pressure level, and "drone metric" of sightings that are classified as drones. The classification threshold for the drone metric can adjusted in the Settings dialog. Most of the controls are simplified relative to BeamformX. Fisheye mode in the Camera dialog is recommended. The frequency should usually be set to about 4 kHz. The posted version, 0.001, is untested and should be viewed as a rough draft.

See our Privacy Policy here