Knowledge Base

General Features

• Subjects per Arena: 1 (Base module), up to 16 (Social Interaction module). The limit of 16 subjects is set by the software. It may be difficult to discriminate between 16 different color marks depending on the light conditions. We recommend to only use Tracking from video files when you track multiple subjects and multiple body points.
• Arenas: Multiple arenas with the Multiple Arena Module; one arena with the Base Module. Arenas must not overlap. The number of arenas does not depend on the number of cameras. You can have for example 96 wells in one camera image, or 16 arenas with 16 camera views (one arena per camera). When you combine tracking of multiple subjects (with colors) and multiple arenas, performance may limit the total number of subjects followed simultaneously.
• Trials: With the Trial and Hardware Control Module.
• Behavior Types Scored Manually: With the Trial and Hardware Control Module.
• Hardware Commands and Signals: With the Trial and Hardware Control Module.
• Arena Settings Cache: 50 MB.
• Heatmap Dimensions (Pixels): No restriction specified.
• Rows or Columns in Analysis Results Table: One million.
• Free Intervals / Time Bins: No restriction. If you create more than 1000 time bins, the group charts are not displayed.

Computer Requirements

• Computer Type: Workstation. We recommend the use of Dell workstations especially when using multiple add-on modules (e.g. tracking nose-, center- and tail-base points in multiple subjects simultaneously). When tracking for more than 24 hours, restart the PC before a new trial.
• Video Graphics Card (GPU): NVIDIA, 4 GB or larger.

  Supported GPUs:

  • Quadro P2200
  • T1000
  • GeForce RTX 4060

  If you opt to buy your own PC, it must have a stand-alone graphics card that supports Direct3D acceleration. We recommend NVIDIA video cards with a memory of 4 GB or larger. 5th Generation boards are NOT yet supported. For tracking subjects with Deep Learning, there are further restrictions. See Deep Learning: Requirements.

• Processor and RAM Memory: Intel. We recommend an Intel processor with 32 GB RAM.

Video Resolution, Sample Rate and Trial Duration

• Video Resolution (Camera Image): Depends on camera type. GigE maximum resolution for use in EthoVision is 1280 x 1024.
• Video Resolution (Video File): 1900 x 1200. Keep DDS selected in the Acquisition screen.
• Sample Rate (Live Tracking): 60 frames per second; it depends on the camera. Possible with GigE when using with DanioVision.
• Sample Rate (Tracking from Video File): Not specified. EthoVision XT can handle short video files with a frame rate of 1000, but that also depends on the video format. Test a few sample videos and check for missing samples before running the experiment.
• Trial Duration: 72 hours maximum, but this depends on the tested setup.

  Max duration of 72 hours:

  • 16 GigE (2560x1920) + Contour @30fps
  • Sylics: Contour + Center, 4 arenas
  • DanioVision (live, no recording)

  Max duration of 3 hours:

  • 1 GigE (1280x1024) + DL @60fps
  • 2 GigE (1920x540) + DL @30fps
  • 4 GigE (2560x2048) + DL @25fps
  • 4 GigE (2560x2048) + Contour @60fps

  DanioVision:

  • Recording video: 1 hour
  • Live (no recording): 72 hours

  Deep Learning and/or MST Live (max 4 arenas):

  • 1 animal: 72 hours
  • 2+ animals: 1 hour

Important: When tracking from video files at high resolution and/or frame rate, always select DDS in the Playback Control window. This way EthoVision XT adjusts the analysis speed to the time needed to analyze each frame without skipping frames.

Important: When tracking for more than 24 hours, always reboot the PC after every trial to refresh the PC resources, the RAM specifically. Failure to do so could result in crashing, which can corrupt the experiment or the recorded video.

Applications and Add-Ons

• Multiple Body Point Module (Live Tracking): Four arenas per computer. In principle it is possible to track in more than 4 arenas, but it depends on the apparent size of the subjects in the video image (in pixels). Sample rate must be 25 or 30 FPS.
• Multiple Arena Module: 100 arenas. The number of arenas may be limited by other factors, for example the number of subjects in each arena and the tracking of multiple body points. Contact Noldus if you wish to work with more than 100 arenas.
• Social Interaction Module: 16 subjects per arena. If subjects are unmarked, there may be subject identity swaps.
• CognitionWall (Sylics): Four arenas, only center-point tracking; up to 72 hours recording. In case of recordings longer than 24 hours, we advise to restart the PC after each trial. A workstation is required.
• Behavior Recognition: One subject per arena; one arena (live tracking), up to four arenas (offline tracking). Requires the Multiple Body Point module. Camera must be overhead. Camera frame rate must be at least 25 fps with no missing frames. Cannot work together with Deep Learning-based tracking. The PC must be a workstation.
• Deep Learning for Nose-Tail Base Tracking: One or two subjects per arena (up to 4 arenas). See Deep Learning Requirements below.

Source: EthoVision XT 18 - GigE Install and Setup, Noldus Information Technology