Page 1 of 1

Yawcam BETA 2012-09-08

Posted: Sun Sep 09, 2012 9:12 am
by malun
Get the new beta release:
yawcam_BETA_2012-09-08.exe (4,42 MB)

What's new since last official release?
- Changed limit behavior in motion detection's file action.
- Fixed motion detection's problem running .bat files.
- Added text file hack: ipcheckurl.txt
- Added workaround for unstable ftp hosts.


Posted: Sun Sep 09, 2012 9:21 am
by malun
The "text file hack" ipcheckurl.txt provides an option to use an alternative url when detecting the public ip address of the computer running Yawcam.
The file must contain a url that you supply. The service of the url must return ip address of the client sending the request. The response must only contain the ip adress in text format without any html or xml formatting.

More info on the rest of the text file hacks here:


Posted: Sat Oct 13, 2012 7:44 pm
by keymapper
Hi, I just installed this beta and tested it. It does continue to record, which is great! Thanks for making the improvement.

A couple of things I found during my testing. First, it records anywhere for 20 to 27 frames per second. That in itself isn't a big issue. The problem is that most of the frames within the second is exactly the same.

A second issue, possibly similar to the current non-beta version, is that during the sequence of motion, the subject (me) moved into frame over a couple of seconds, stayed mostly steady in the frame for a few seconds in the middle of the sequence, and then I continued my movement. During the time I was steady in frame, recording stopped, even though I was still in frame.

It seems like the app determines motion by comparing the current frame with the last frame. Is it possible to compare with a baseline frame, or the frame right before motion is detected, and if there a difference, recording should continue until there is no more motion detected between the current frame and the previous frame or the baseline frame.


Posted: Sun Oct 14, 2012 9:53 pm
by z3r0c00l12
I'm sure it could be done, but then further complications would occur, for instance, in a light-changing environment, the motion would trigger and run for almost 12 hours before the light level comes back to normal. Although it would seem like a nice feature, it would bring a lot of other complications.

I'm not saying no, nor yes, as I am not the author of the software, but from my point of view, I don't see it likely the feature will see the light of day.

Posted: Mon Oct 15, 2012 7:51 pm
by keymapper
Thanks for the reply, z3r0c00l12.

To avoid the complications you mentioned, you would simply set a max time limit for recording (e.g. max 10 minutes or some user-adjustable value), and/or you would only us a baseline frame that is maximum x number of seconds/minutes old.

One great example of how motion detection is done properly is iCamSource, the PC/Mac companion software for iCam for Android and iPhone. Its motion detection has 2 settings, frequency and sensitivity. Both settings are just sliders, less often <-> more often, less sensitive <-> more sensitive, respectively. It is simple to configure and it works as expected. When there is motion, it continues to record until the subject is moved out of frame, even if the subject comes into the frame and stays steady for a short time.