Amcrest 841's on Blue Iris suffer bandwidth

Joined
Feb 2, 2018
Messages
16
Reaction score
1
I've been playing around with a Blue Iris install for about a month. Still learning the software.

Right now my setup is 10 Amcrest 841B cameras with 9 of them on Wifi. My Wifi network is rather robust, using Unifi gear and lots of access points. I don't tend to have more than 2 cameras connected to any one access point and the AP's are very well spaced out.

The problem I'm having is that after running for some time, I come back to find one or more cameras have gotten themselves into some kind of bandwidth trap. They're set to stream at 4000kbps, and tend to read in at a little over 500kB/s in the Blue Iris status window when working properly. When they're failing, they'll read ~200kB/s and the fps will drop from either 30 or 15 (depending on the camera) to anywhere below that. Live view and recording suffers to the point of being worthless. Resetting the cameras either manually or reconnecting them to the network through my Unifi network controller usually fixes them.

Some of the cameras seem to never exhibit this problem, and some have it daily. Since I can simply reconnect the camera to Wifi and the problem goes away, I don't think it is an issue of interference or Wifi congestion. So I think it is a problem with my network or the cameras or Blue Iris or the way they interact.

Any suggestions?
 

jason0

n3wb
Joined
Jan 10, 2018
Messages
5
Reaction score
1
When this happens, have you tried to connect to the cameras directly through their web interface, or using another application on your smartphone to connect directly?
Does restarting blueiris fix the issue?
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
When this happens, have you tried to connect to the cameras directly through their web interface, or using another application on your smartphone to connect directly?
Does restarting blueiris fix the issue?
For some reason I can't get liveview through the camera web interface anymore. The only way to do it before was using a chrome app and it seemed to just stop working a month or two ago (maybe coinciding with upgrade to win10?).

I'll try restarting BI next time I see the problem.

I'll try the phone app too.
 

jason0

n3wb
Joined
Jan 10, 2018
Messages
5
Reaction score
1
Well I will echo @IDLE... I've had PoE cameras and WiFi. Never had any trouble with my PoE cameras, and always trouble with the WiFi ones, unexplained drops in FPS, and occasionally completely falling off the grid.
Even if you have an excellent WiFi backbone, the Amcrest WiFi units themselves could be junk.

I am now 100% PoE.
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
Well I will echo @IDLE... I've had PoE cameras and WiFi. Never had any trouble with my PoE cameras, and always trouble with the WiFi ones, unexplained drops in FPS, and occasionally completely falling off the grid.
Even if you have an excellent WiFi backbone, the Amcrest WiFi units themselves could be junk.

I am now 100% PoE.
Don't get me wrong, I'm big on wiring everything that can be wired. I also have 22 Ubiquiti POE cameras.
However I want the ability to drop cameras in places without having the spend any amount of time. That is how I've been using the wifi cameras
 

bp2008

Staff member
Joined
Mar 10, 2014
Messages
12,666
Reaction score
14,006
Location
USA
You could try reducing the bit rates on wifi cameras, starting with those which are furthest away.
 

jason0

n3wb
Joined
Jan 10, 2018
Messages
5
Reaction score
1
Are they all running the same firmware version?
If you conclude its the camera's themselves, you may be able to setup a script to reboot them automatically when they exhibit problems. Or just automatically reboot them every X hours.

Terrible solution but you might not have another choice.
 

bp2008

Staff member
Joined
Mar 10, 2014
Messages
12,666
Reaction score
14,006
Location
USA
I wouldn't call that a solution. The bandwidth trap doesn't seem to relate to distance in any way.
Larger distance would typically result in more re-transmissions and/or higher output power, so reducing the bandwidth of those could have a slightly greater effect (reducing interference for the other cameras).
 

awsum140

Known around here
Joined
Nov 14, 2017
Messages
1,254
Reaction score
1,128
Location
Southern NJ
Further distance means reduced signal levels in any case. WiFi is nice for a phone but not for a camera IF you want consistent, reliable performance. There are too many variable with RF transmission and reception at the extremely low power level permitted for WiFi. The higher the bandwidth, the higher the signal quality, level, has to be to achieve reliable communications. That applies to every RF device. Add in marginal, or low gain, high noise, components, and things get worse yet. Just because your laptop sees a -45dbm signal, for example, doesn't mean the camera sees that same signal or has the same receiver capabilities. Since BI only process the signal it receives from your router I don't think the problem is there.
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
The bandwidth trap happens with different cameras at different times, including 2 cameras that are 8 and 10ft from an access point. The 8ft one is line of sight, and I have no other AP's in the house operating on that channel, and would have to go outside to see other AP's on that channel.
Reconnecting the cameras fixes the problem immediately. I've also learned that simply reassigning a camera to a different user group in my Unifi controller also fixes the problem immediately.

I hear what you guys are saying, but all things indicate that this is not an interference or distance issue. It has something to do with the wifi handshaking/protocol/whatever where the clients get stuck in a bandwidth limited state.
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
Oh, and concerning the two cameras that I mentioned that are right near their AP. I've observed either or neither or both of them get stuck in this bandwidth trap. One can operate perfectly and the other sinks to a few fps.

It is just really hard to imagine that this is happening because of a crappy rf environment.
 

awsum140

Known around here
Joined
Nov 14, 2017
Messages
1,254
Reaction score
1,128
Location
Southern NJ
Two cameras, close to each other, will create their own interference. Think about what happens in your car when you're listening to a radio and drive by a radio station transmitting site. That's called intermodulation and desensing the receiver front end. The same thing happens with WiFi, transmitters and receivers on adjacent, or the same, frequencies.
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
Two cameras, close to each other, will create their own interference. Think about what happens in your car when you're listening to a radio and drive by a radio station transmitting site. That's called intermodulation and desensing the receiver front end. The same thing happens with WiFi, transmitters and receivers on adjacent, or the same, frequencies.
Ok, but Wifi is designed to share airspace amongst clients, right? If what you're describing is pervasive, you'd never get a wifi network to work with more than one client.

For the hours that the cameras are working perfectly, is that just pure luck that they're not interferring with each other?

I have another two cameras in a garage, connected to one AP, and they haven't had a problem in days.
 

awsum140

Known around here
Joined
Nov 14, 2017
Messages
1,254
Reaction score
1,128
Location
Southern NJ
WiFi may be designed, but RF isn't aware of it. Two transmitters and two receivers on adjacent channel and physically close to eah other withh create desense and intermod. It doesn't take much to make it "go south". At the frequencies involved multi-path is also a factor and the fact that it is almost a full duplex operation only makes it worse. Video requires perfect signal conditions and WiFi sure ain't perfect.
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
WiFi may be designed, but RF isn't aware of it. Two transmitters and two receivers on adjacent channel and physically close to eah other withh create desense and intermod. It doesn't take much to make it "go south". At the frequencies involved multi-path is also a factor and the fact that it is almost a full duplex operation only makes it worse. Video requires perfect signal conditions and WiFi sure ain't perfect.
The way you describe it, it is such a wonder that it can work for many hours straight. If it errs, a tiny reset fixes it. Feels a lot more like a networking bug than a pervasive RF problem. If it would an RF problem it would come and go, right?
 

awsum140

Known around here
Joined
Nov 14, 2017
Messages
1,254
Reaction score
1,128
Location
Southern NJ
To me, and I've had some experience in RF, it's a wonder it ever works at all. It does come and go, you describe that. The problem is it only takes a tiny change in the RF path/signal to make it drop. A branch moves in the wind, someone or something walks or flies by, a fly farts. Just about anything can do it and you probably won't even notice when it happens or see it happen. Levels are so low and the hardware, especially compared to real transmitters and receivers, it pretty marginal so the tiniest variation can cause problems. The "capture" effect also comes in to play, the receiver latches on to a signal and hangs on to it, which is why a reset cures it. It's latching on to a marginal signal and stays there. The reset makes it start over and it "sees" the better signal path again.

I was using WiFi to set up a camera in the front yard, using BI and UI3, say about 130 feet, and a number of walls, from my router. The tablet, a Samsung, wouldn't hold a connection to the router for more than a few seconds, then signal loss. The craptop, an Asus, would hold it for a few minutes, then I'd move or something else would, and it would drop. Went back into the house, rotated the WiFi router 90 degrees reorienting the internal antenna, and was able to maintain a connection long enough to get the camera zoom and focus set. Minor changes in the RF path made all the difference. How long it would hold I never found 'out, didn't want to spend the afternoon standing on a ladder under a tree mounted camera. Like I said, WiFi, being as marginal as it is, is not a good solution for video transmission.

Ever notice how TV stations do real remotes? Fancy, high gain, microwave dishes on telescoping masts and high power, 100 watt or so transmitters. Even with that they drop signals all the time. The new trend of cell phone video, while convenient, shows the same set of problems. Back before cable wasn't around there was subscription TV, pay UHF channels. They wanted a signal level of 0dbm to be able to provide service and they were providing antennas with 10 to 15db gain not the minus db gain of stub or no real antenna ( a quarter wave at 2.4ghz is really short but has 0db gain or even a loss. Even then, there were drop outs. I've always wondered why dish antennas, which have some real gain, aren't used for point to point WiFi just for applications like this. It would probably eliminate a lot of the problems. I guess they're not aesthetic enough.

Incidentally, my first camera was a WiFi in that same tree. I used antennas that were advertised as 6dbm gain, allegedly multiplying the signal level by four times. It would hold for a few hours then drop and need a reset. Same thing you're experiencing.
 
Joined
Feb 2, 2018
Messages
16
Reaction score
1
So how do I prove that I'm having an RF environment problem? The symptoms seems so much more like some kind of "bandwidth trap", as I keep calling it. Like if the camera fell behind for just a tick and then got stuck there.
 

awsum140

Known around here
Joined
Nov 14, 2017
Messages
1,254
Reaction score
1,128
Location
Southern NJ
Run a cable is the simplest way. I dug a trench, put in conduit and sealtite and pulled cable. No problems with dropouts anymore. String one overhead or whatever, but I think you'll see the same result...solid signal and no dropouts.
 
Top