I'm having a hard time adjusting my motion settings to where they work consistently.
On some days, what I've got works flawlessly, other days I am inundated with emailed pictures with no motion.
Is there a way to print the % difference on each picture (i.e. after the date) to help in determining settings?
If I select the add % difference option it is present in the user text for images uploaded to my website. I use this number as guide to setting this parameter. Its done at the time the image is created. I turn the option off when I am finshed setting up.... which bever seems to happen. See below
I have mentioned the 'no motion' detection problem before. It seems to arise when the reference frame contains some motion. Then when comparing this frame with subsequent frames it generates a %difference for every frame after that ie no motion.
I have been playing with the settings for ages now and this current setup seems to be working for me,,,, at least so far. Also, sometimes the response is so slow to a motion that the motion cause is out of frame by the time a capture occurs.
Set motion sensitive area to be very small central protion of the frame. This allows the motion to move out of detection area but can still be captured outside the area, because its still in the frame, in the case of slow response to motion event.
Do not use fast alternate pixel mode.
Set time delay between detection to be 1s
Set time 2 motion vents required before capture.
I am still not brave enough to set up emailing as I do not have a great deal of confidence in it working reliably yet!!!
Do not start motion capture immediately after starting up the camera. They need some time to settle down.
Use reference frame 'last frame with no motion' setting
I think the real problem is how the software picks a reference frame. Due to using my camera outside I cannot use a fixed reference frame due to varying lighting conditions. The reference frame that is used needs to be one with no motion but there is no control over how a reference frame is picked, excpet for 'last frame with no motion' but this doesn't alwys work. Even with this setting the reference frame used sometimes has motion in it. Typically this seems to happen when the motion is a large area and slow moving. ie It may generate the same % difference for a few frames and thenÂ´the software seems to decide to use one of these frames as a new reference. Maybe it should only select a reference frame if five previous ones are the same, or within a small % the same,,,, or something like that. Or it should wait a longer time after motion occured before selecting a new reference frame.
I find the whole motion detection setup not very user friendly and more a case of trial and error.
Its not easy I guess when the lighting conditions vary quickly eg sun passes in front of a cloud...... but other software seems to able to manage it somehow.
Another idea for Marc ..... make the reference frame a user set number of frames behind the actual frame under test. This should have the effect of also varying the sensitivity to the 'rate of change' in the image and not just the % difference. eg if the ref is one frame behind the actuall frame then the % diff value would only be sensitive to very fast motion.