-
Semper Vaporo
-
- Posts: 331
- Joined: Sat Oct 09, 2010 11:51 am
by Semper Vaporo » Thu Apr 09, 2015 8:58 am
I don't work for Sighthound but maybe I can offer some minor help as I am a long time user of the program and have experienced some things similar to what you describe.
I am presently running just 1 IP cam (the other 5 are USB) and I am accessing it with both SightHound and a browser (IE10 via the camera's web page). Technically, they are separate streams of video from the camera. With both systems I get a "disconnect" every 5 minutes and 55 seconds, (355 seconds), but it comes back after only 1 second. This "disconnect" occurs simultaneously on both displays of the video so it seems to be a Network problem, not IE or SH. I recognize the "disconnect" because there is a hanging birdfeeder in the scene that gently swings in the wind and it will 'freeze' in mid swing for a second and then jump to a new position in the swing and birds will 'jump ahead' in mid flight. I also can detect the "disconnect" by watching the "Network" tab of "Task Manager". With both cameras running I see about 8% of network usage, but every 355 seconds there is a spike down to zero and right back up to 8% (the actual percentage varies as to the contrast of the scene, large areas of different contrast take more bandwidth than a whole scene of uniform brightness). This is how I have determined the period of this type of disconnect.
Additionally, I get a random disconnect (anywhere from a few minutes to several hours apart) that can take from 8 to 30 seconds to come back. Both views of the camera experience this disconnect, but they exhibit it slightly differently and not necessarily at the same time. The browser window will just freeze for a few seconds (birds freeze in mid flight, the birdfeeder stops swinging, etc.), then it briefly shows a white screen, then continues showing the scene (and in that pause, lots of things can change in the scene!). The SightHound window will first show the blue screen of "Could not connect" for several seconds, then show the blue screen of "Connecting" for a second or two, then revert to showing the scene (sometimes showing a black screen for one second before the scene returns). I have seen each view (IE and SH) of the scene do this "disconnect" at separate times and once I saw it occur at nearly the same time. When they do it at separate times, the Task Manager display will drop to about 4% usage for the duration of the "disconnect". When it occurred to both at the same time, it dropped to 4% for 1 second, then dropped to 0% and then both cameras re-connected at the same time as the usage returned to 8%. (Actually, the IE view seems to take slightly higher bandwidth because when it disconnects, the drop in percentage is slightly more than when SH experiences the disconnect, but it is very difficult to get exact numbers because the grid behind the graph is not very well defined and the graph data line is quite squiggly).
In the past, I'd get 'antsy' and start clicking away at things to try to get the camera to "reconnect" and only recently discovered that it seems to correct itself if I can sit on my hands and wait.
To simplify your "cure", have you tried to turn off the camera (via SightHound, not actual power to the camera) before you terminate SH? (Click the green dot lower left corner of the display.) This might cure the tedium of having to close the individual instances of the SightHound Agents. If any cameras are left "On", then the agent for that camera and the other Agents for the other processes are kept running, but if no cameras are "On" then they all get stopped when you terminate the user interface program. By turning the camera "Off" you might not need to do all that Task Manager work. (Not that you should have to be doing so in the first place.)
Some other things that I have found that had some effect on how often the random disconnects occur... I have found that if I physically turn off both the camera and the router it can reduce the frequency of the random disconnects (does not affect the 355 second type). I have also found that completely erasing the SH "Archive" folder has helped in the past, but didn't do a thing for me the last time I tried it as they started occurring right away that time.
I will put the SH tech's here on the spot and say that I am sure they can provide access to older copies of the program.
Semper Vaporo,