This page last changed on Nov 27, 2007 by brian.

Questions

Answers

Q: Where does VARS get the camera data from?

The camera data is retrieved through the EXPD database on Solstice. Rich Schramm set up a process to load the camera-logs. Here's a note from Rich:

Normal access should be to query through the views...either select * from VentanaCamlogData or select * from TiburonCamlogData

This view is going against base tables that are partitioned by vehicle and year. I recommenf you write software against the view rather than trying to go to the base tables. Access through the view should be very fast due to how Ive partitioned and indexed the tables however that remains to be verified once we get the several million rows added.

The status of loads can be viewed by select * from CamlogLoad, you can select by vehicle, year yearday etc....

Q: How does VARS handle frame-capture?

Overview
VARS provides the ability to capture images from a video input. We are currently using Decklink cards from Blackmagic. In the video lab, we are using Decklink. On the ships we use Decklink Extreme cards, which support multiple video sources. VARS communicates with the Decklink cards using Quicktime for Java.

Where are the images initially stored?
Images captured from video are initially stored on the workstation that is running VARS. The image location is <USER_HOME>/VARS/data/<platform>/images/<divenumber>/<timecode>.jpg. For Example, Saving an image for Tiburon dive # 789 for timecode 02:43:25:28 when you are logged in as user 'Tiburon' on Windows would create an image at C:\Documents and Settings\Tiburon\VARS\data\Tiburon\images\0789\02_43_25_28.jpg. A reference to the image is inserted into the database as a local URL (i.e file:/path/to/file/filename.jpg). The image is then immediatly accessible to the annotator but will not yet be available to any VARS applications that are not running on the same workstation.

How do the images get onto the web?
Annotations made on the ships and annotations made on shore are handled differently. However, all images will eventually make it onto //Tornado/framegrabs which can be browsed from the web at http://search.mbari.org/ARCHIVE/frameGrabs/

When the ship returns to dock, an rysnc is scheduled at night to copy the images to //Tornado/framegrabs. The rsync job looks for images only under specific users, which are 'Tiburon' on the Western Flyer and 'Ventana' on the Point Lobos. If you are logged on the workstation as a different user when you annotate on the ships the images will NOT be copied to shore. At 5:01 AM, the computer Seaspray runs a process that updates file URL's to HTTP URLs in the database. Any framegrabs that are made during a cruise will not be available through the VARS query until at least 5:02 AM of the day after the dive.

On shore, assuming you are on the internal network, VARS will attempt to copy the framegrabs over to //Tornado/framegrabs whenever a Video-archive is closed. The copy process also updates the URLs in the database. On shore, it does not matter who you are logged in as. Framegrabs that you make on shore will be available to everyone as soon as you close the Video-archive that you are annotating.

Q: How do I upgrade Java?

VARS should run on Java 5.0 and Java 6.0. If you are interested in upgrading your Java installation, you can get a more recent version of Java at http://java.com. If you are upgrading Java on a Windows computer that you use to annotate, you will must include QuickTime for Java in the new Java installation. There are 2 options to accomplish this:

Option 1: Reinstall QuickTime

Option 2: Copy QTJava.zip from a previous installation of Java to the new one.

  1. Locate your previous installation of QuickTime for Java. On Windows, the file will be located on a path like C:\Program Files\Java\jre1.5.0_06\lib\ext\QTJava.zip.
  2. Copy the zip file to your new Java installations ext directory (e.g. C:\Program Files\Java\jre1.6.0_01\lib\ext)

Frame_Capture.dp.gif (image/gif)
Document generated by Confluence on Feb 03, 2026 12:59