BIG Owl on my Trail

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
And sometimes the animals are vewwy, vewwy quiet
:lol:
, so the auto-align doesn't work, and I have to do the alignment graphically, by burned-in timestamp (BlueIris and/or my cameras don't include a timecode, AFAICT)*.

The results are usually good, and this would be very, very difficult to do in ffmpeg without the graphical interface. I also need to nudge the angles on the cameras and IR illuminators a bit:


On that note, is there a way for Amcrest cameras and/or BI to add timecode to the video streams without re-encoding? I'm using direct-to-disk in BI.
 
Joined
Dec 28, 2019
Messages
8,069
Reaction score
17,558
Location
New Jersey
If you're using bvr then BI is recording everything the camera sends. So if you turn on the time display in the camera, position it strategically, and sync it to a local NTP server, like NetTime running on your PC, everything will be pretty much synced down to at least a few tenths of a second on the display. That's my strategy for time sync and I can see them all pretty much rollover on the minute. (hard to watch all 21 at once)

Incidentally, I don't know what it is about your videos, but I always watch them, multiple times. Actually the same is true of lulucamz, especially his wildlife stuff. He just posted one of a coyote having Colonel Sanders on the hoof.
 

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
If you're using bvr then BI is recording everything the camera sends. So if you turn on the time display in the camera, position it strategically, and sync it to a local NTP server, like NetTime running on your PC, everything will be pretty much synced down to at least a few tenths of a second on the display. That's my strategy for time sync and I can see them all pretty much rollover on the minute. (hard to watch all 21 at once)
Ah, I don't mean the burned-in timestamp that gets displayed; I mean the binary timecode that tells the computer exactly when each frame occurred. This would allow applications like Resolve to sync multiple clips even in the absence of audio.

But now I'm seeing that the MP4 specification doesn't actually have a provision for binary timecode, and workarounds for embedding it can cause problems.
Incidentally, I don't know what it is about your videos, but I always watch them, multiple times. Actually the same is true of lulucamz, especially his wildlife stuff. He just posted one of a coyote having Colonel Sanders on the hoof.
Thanks! I only post the interesting ones ;)

LOL! I just watched lulu's rooster video :oops:

Here are my coyotes again. The lead coyote has a bit of a limp:

 

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
The framing matches a lot better as well! Great job!
I nudged the cameras and the illuminators a bit, and I think this is the best that I can do. The cameras' FOV is wider at the bottom than in the middle, so I either double-up my legs a bit, or else I miss my head a bit:


I wish Davinci Resolve supported 20 fps output, because then I could get the frames in better sync, but you really only see the artifacts when you step through the video frame-by-frame. At full speed it's not very noticeable.

Now I'll wait for tonight's coyotes . . . .
 

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
Wow! I didn't get much video action this morning; in fact BI didn't even trigger, but the sound woke me up. There's not much to see, but they enter from my neighbor's paddock, top-center, and then proceed to scream their bloody heads off:


I thought it was a pack of 10!! Amazing how much sound 2 medium coyotes can make. My neighbor came hiking up the trail 60 seconds afterward.:oops:
 
Joined
Dec 28, 2019
Messages
8,069
Reaction score
17,558
Location
New Jersey
I'm guessing here, but those cameras look like either different models or different focal lengths. If that's the case getting them stitched together will be impossible. If it isn't the case it's still pretty much impossible to stitch them together without some serious post-processing. That said, you've done a great job to say the least!
 
Last edited:

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
I'm guessing here, but those cameras look like either different models or different focal lengths. If that's the case getting them stitched together will be impossible. If it isn't the case it's still pretty much impossible to stitch them together without some serious post-processing. That said, you've done a great job to say the least!
They're the exact same model, Amcrest IP5M-T1179EW-28MM, but they're mounted like this:



That's the view from where the coyotes are. Heres the closeup:



With that vertical gap between them, aligned to match up on a object 6-8' away, the parallax at any other distance is going to be significant, which is why the top of the combined image is so far off.

But this has me woolgathering a bit. The top-down view shows that the yaw angle between them is very roughly 80 degrees:





and there's a bit of a pitch angle as well.

But what if I mounted them opposing each other, e.g. straight up and straight down on that fence post, and then put a double-sided mirror, or just two thin mirrors, in the middle? Could I get optical centers of both cameras closer together that way?



I don't think well enough in 3D to be able to visualize how much freedom I'd have to adjust the angles and get the images aligned any better.
 

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
Here's some more epic howling. The coyotes walk by the cameras about a minute after they stopped howling (edited out), and they're not carrying any prey, so it was just howling for the sake of it:


and then here's my first coyote footage of 2022. Happy New Year, everyone!!

 
Joined
Dec 28, 2019
Messages
8,069
Reaction score
17,558
Location
New Jersey
I think you'd lose way too much light fooling with a two sided mirror like that Then there's the problem of dust/dirt/moisture messing up the mirror and at the short distance between the lenses and mirror it would be very noticeable.

Another trick might be a "boobie" camera.

Boobie camera
 

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
I did think about the Boobie camera. It's a bit expensive.

For $2.89 each I might buy a couple of these mirrors and play around with the idea. I have a bunch of Amcrest bullets lying around that I won't need until swarm season starts up in ~2 months.
 

TheWaterbug

Getting comfortable
Joined
Oct 20, 2017
Messages
281
Reaction score
383
Daytime coyotes! I've seen them at 8:00 AM and 5:00 PM before, but this is the closest to mid-day that I've seen them yet. Interesting that they're different colors! One is more reddish, and the bigger one is grayer:

 

NightLife

Getting comfortable
Joined
Sep 10, 2021
Messages
354
Reaction score
728
Location
Canada
Howling for the sake of pretending to be something they're not .. a pack. Wolves try and fake out others like that as well, howling so it sounds like they're larger in numbers, when in reality there's like 2 or 3 bro's. :rofl:



Thanks for the cool videos!
 
Joined
Dec 28, 2019
Messages
8,069
Reaction score
17,558
Location
New Jersey
From my experience all of them have hot spots, some worse than others. This is my favorite from CMVision. I have six of them now with four of them in service for three plus years with no problems. I haven't bothered to measure current on them but they do a pretty good job. There is a definite hot spot on the center but it isn't too pronounced in the FOVs I have, say 20 feet or a little more. I prefer the higher LED count units on the theory that if a few fail the unit will still provide enough IR. If that happens with a four LED unit, it's pretty much useless.

 

Terri Underhill

Getting the hang of it
Joined
Mar 23, 2019
Messages
127
Reaction score
76
Location
oklahoma
Top