Just Sharing My Progress

[smithb]

I grabbed these snapshots this morning to show how bad the IR reflects off the dome. When I first pulled the camera out of the box I thought that it wouldn't be a problem because there is a rubber hood that encircles the lens and seals against the dome to block out IR. In practice it doesn't quite do the trick.

These were taken about an hour before sunrise with my porch lights off. My neighbor to the right has some Christmas lights on but no porch lights or floods. It looks pretty darn bright, though, so he may have an IR flood that I don't know about. I'll have to ask. The light source to the left is the IR from my driveway cameras. I'll try to get some snapshots from all of my cams tonight without any ambient IR.

Across the street is one 60W decorative lamp post. No street lights. I really dig the STARVIS sensors in these cameras. I wish my night vision scope was half as clear and bright.

With SmartIR enabled:

With IR off:

 

[smithb]

One cool thing about having the camera installed to the side is that you can take full advantage of human ear recognition. I don't know much about it but I've read that the human ear is as individual as a fingerprint.

 

[bob2701]

Too bad you couldn't get to the other side of the door, much better face shot as they reach for the doorbell. Nice install.

 

[smithb]

Too bad you couldn't get to the other side of the door, much better face shot as they reach for the doorbell. Nice install.

I agree that the left side has the better vantage, but I tested it out quite a bit and found myself naturally drawn to look to the right of the door because the left is somewhat obstructed by the stone wall and the right has an open view of the driveway and neighborhood.

Plus, the wife did not approve the left location and that’s pretty much the end of my decision tree...

Combine all that with difficulty of installation and the decision was easy.

 

[looney2ns]

I grabbed these snapshots this morning to show how bad the IR reflects off the dome. When I first pulled the camera out of the box I thought that it wouldn't be a problem because there is a rubber hood that encircles the lens and seals against the dome to block out IR. In practice it doesn't quite do the trick.

These were taken about an hour before sunrise with my porch lights off. My neighbor to the right has some Christmas lights on but no porch lights or floods. It looks pretty darn bright, though, so he may have an IR flood that I don't know about. I'll have to ask. The light source to the left is the IR from my driveway cameras. I'll try to get some snapshots from all of my cams tonight without any ambient IR.

Across the street is one 60W decorative lamp post. No street lights. I really dig the STARVIS sensors in these cameras. I wish my night vision scope was half as clear and bright.

With SmartIR enabled:

With IR off:

That could be IR reflection from the post/ceiling. Not from inside the cam.

 

[smithb]

That could be IR reflection from the post/ceiling. Not from inside the cam.

Good point. I’ll take the dome off and see how it looks. I also have an IR flood that I can experiment with.

 

[nats]

One cool thing about having the camera installed to the side is that you can take full advantage of human ear recognition. I don't know much about it but I've read that the human ear is as individual as a fingerprint.

lol

Plus, the wife did not approve the left location and that’s pretty much the end of my decision tree...

Amen

I agree on the face shot comment, nevertheless you should get a reasonable shot for human based face rec, automated one will...well it works better when you have both eyes.

 

[FoxCR]

Like this? (does this count?)
https://photos.app.goo.gl/HqZBb5o5jHFcCfYN6

 

[nats]

That image should be good for facial. given you have something to compare against.

One note about facial and wide angle cameras. They have some relevant distortion towards the edges, if you must use one for facial only the center region will produce good images.

 

[FoxCR]

This is my package cam has been working well.

 

[FoxCR]

I would be afraid some would walk away with the rig and camera since we both work during the day. my HOA would probably also fine me - but I think what you are doing is great and the best way of doing it.
how did you deal with the ethernet cable going to the outside while testing?

 

[smithb]

I would be afraid some would walk away with the rig and camera since we both work during the day. my HOA would probably also fine me - but I think what you are doing is great and the best way of doing it.
how did you deal with the ethernet cable going to the outside while testing?

Dealing with HOAs can be difficult, but hopefully if you present your case up front they'll be amenable. Security cameras are becoming much more prevalent these days but most people still think of ominous bullets and box cams on prison walls. I would share pics of turrets with them to provide assurance that they're no more obtrusive than a light fixture.

For the test rig ethernet cable, I ran a cable a little longer than the full length of my first run and installed a temporary POE switch at the terminal end. From there, I ran patch cables to each camera location and secured them with cable ties. If you're in a high-risk area you can probably gather enough information to make an informed decision on placement so long as you have footage in both daylight and darkness. The risk with gathering less data is that you might need to relocate the camera or get creative with lighting if you miss something during testing.

 

[FoxCR]

yes HOA's need to get on with the changing times. I think Nest and Arlo have done a fairly good job educating folks.
so what I have been doing is trial an error, but I see the value of a test rig even to compare cameras. Lets say at 4 vs 6mm or 2.8 at a certain location.

 

[smithb]

After reviewing the footage of people coming to the door throughout the day I decided to rotate the lens and disable corridor mode. I think it's a better view.

 

[smithb]

Here's something that I need to start trying to figure out. The image appears to pixelate every few seconds. Maybe it's a noise filter setting or something?

 

[catcamstar]

Here's something that I need to start trying to figure out. The image appears to pixelate every few seconds. Maybe it's a noise filter setting or something?

Are these "snapshots" from the camera, or you having "paused" the video and grabbed a screenshot?

Can you report your camera settings, I'm especially interested in: fps, bandwidth, codec and more importantly: iFrame interval. I suspect the latter not being equally set as your fps. iFrame "buildups" can render these impurfections.

Hope this helps!
CC

 

[smithb]

I was having trouble taking snapshots in UI3 so I took screenshots and reduced the size a bit. I'll pull native images from my VMS and post them along with the settings later today, if that will help.

 

[smithb]

@catcamstar, I think you nailed it. I apparently didn't fix those when I installed it. The frame rate was 25 and iFrame was 30. I changed them both to 15.

 

[cuz]

Why is the correlation between frame rate and iframe?

 

[SouthernYankee]

Frame rate is the total number of frames per second.
Iframe is when a complete frame is sent.

So if frame rate is 10 and Iframe is 50 then a complete frame is sent every 5 seconds. All other frames are just the pixels that change. The problem becomes if an event occurs between IFRAMEs then it does not have all the frame information during play back. For example if an event occurs with the above setting, if an event occurs at 2 seconds, 20 frames. Then there is a 3 second period where the picture may be garbled.

The standard recommendation is thea the Frame rate and IFrame values be the same. So that there is a full frame every second.