Regarding the instructions of Kinovea 0.9.1 - setup I would like to make some comments due to the behavior that I found during testing the new version:
- Options/Preferences/Capture/General
Setting the „Display framerate (fps)“ will NOT change the video acquisition rate. It exclusively influences the refresh-rate of the display (not the recorded video). This value normally should not be higher than 25. To reduce CPU-load it can set to i.e. 5 or even 1.
Setting the video acquisition frame rate can only be done clicking the „configure camera“ - wrench-icon (bottom left for left camera) and changing the value in the configuration-menu. The resulting framerate that the camera will deliver can be found on the title of the capture-window under „Signal: XX fps“ - Options/Preferences/Capture/Memory
If you allocate ALL available memory to Kinovea, there will be no more memory for the system (all other programs) to work. It may cause a crash of the system. Especially if you are using multiple Kinovea-instances: 1 for recording (it may take ALL memory), 1 for replay (no memory left).
This memory value is exclusively necessary for the ring-buffer to store the acquired video frames. The necessary value to set here depends on the format and length of video you are recording. To find an adequate value it is the best, to record a video of the desired length using the actual settings and then look on the size of the video plus some „safe“ MB. If the video is i.e. 80MB, a value of 150MB should be far enough. If AVI-files are recorded, 1000MB may be necessary due to the large files that are created. If you use uncompressed files (MKV/AVI), a fast SSD is mandatory. - Using the function „open replay folder observer“
During testing this function, I found that it may be critical to a crash of Kinovea. It is more save to use 1 Kinovea with 1 Playback screen for EACH video that is created. Using this configuration, the system works very stable. - Furthermore, I found that using 1 Kinovea for 1 recording camera is better than using 2 cameras inside 1 Kinovea. The noise-trigger can be used on multiple Kinovea at the same time, so it is possible to create the trigger on multiple Kinovea that starts at the same moment. Testing has shown a delay between 2 Kinovea of only some milliseconds.
The reason it it more stable may be that the new version uses separate „threads“ on each function that can be distributed to separate cores of the CPU, working independently at the same time. - There is a new function in Kinovea that helps to differentiate each running instance. If Kinovea is called using command line parameters, a specified environment can be used.
The format is: path…\kinovea.exe -name CONFIGNAME
If kinovea starts using this command-line, it will use the configuration of CONFIGNAME
If Kinovea is recalled later using this command, it will reload the same configuration (Camera settings and ALL Options).
An options to load 2 Kinovea Instances using 1 Camera each and 2 Kinovea instances using „replay folders“ each would be to load the following sequence:
...\Kinovea.exe -name Cam1
...\Kinovea.exe -name Cam2
...\Kinovea.exe -name Replay1
...\Kinovea.exe -name Replay2
You could create 4 Links on the desktop to achieve this. After calling each, select the appropriate camera to Cam1 and Cam2 and select the replay folders to Replay1 and Replay2. In the Kinovea-Title, you could see the related setting in brackets after „Kinovea“.
More advanced you could add the command-line option -workspace WORKSPACE
for the Kinovea-instances doing the replay. Then, the replay function is recalled directly after startup. However, I haven’t tested it so far

Comment