Strange problems with calibration in Arena
Posted: Sun May 04, 2008 11:53 am
We have a very strange situation with calibration on Arena 1.2.034
It's very hard to make a sucessfull calibration because of "No convergance in 30 iteration" warning.
When we connect cameras to notebook Toshiba P324 (Core 2 Duo 1.7Ghz, 2Gb RAM, USB based on intel ICH7 chip) calibration is usual successful with Excellent and Good results. The annoying message "no convergance..." is never occured.
When we connect cameras to desktop computer what was bought especially for work with Arena (Core 2 Quad 2.4Ghz, 2Gb RAM, USB based on intel ICH9 chip) calibration is failed permanently in all modes (FAST, MEDIUM, SLOW).
We use 12 cameras.
It must be said with Core 2 Quad computer capturing gives better results. Yellow and red zones occure rare essentially, FPS is higher and CPU usage is under 25%. But in calibration wizard marker line looks more segmented and broken.
How could you explain it?
It's very uncomforatble to use notebook for every calibration instead of dedicated hardware.
We didn't understand a cause of the problem and suspect that solving can help to improve capturing more.
It's very hard to make a sucessfull calibration because of "No convergance in 30 iteration" warning.
When we connect cameras to notebook Toshiba P324 (Core 2 Duo 1.7Ghz, 2Gb RAM, USB based on intel ICH7 chip) calibration is usual successful with Excellent and Good results. The annoying message "no convergance..." is never occured.
When we connect cameras to desktop computer what was bought especially for work with Arena (Core 2 Quad 2.4Ghz, 2Gb RAM, USB based on intel ICH9 chip) calibration is failed permanently in all modes (FAST, MEDIUM, SLOW).
We use 12 cameras.
It must be said with Core 2 Quad computer capturing gives better results. Yellow and red zones occure rare essentially, FPS is higher and CPU usage is under 25%. But in calibration wizard marker line looks more segmented and broken.
How could you explain it?
It's very uncomforatble to use notebook for every calibration instead of dedicated hardware.
We didn't understand a cause of the problem and suspect that solving can help to improve capturing more.