A few technical thoughts:

A few technical thoughts:

Automatic feeder:

To monitor the amount of corn in the current 225–250 kg feed tank and to control the feeder units mounted on the tank, the Moultrie Feed Hub caught my attention.
A promising device for a task that has been sorely lacking a good solution.

My opinion of the Moultrie feeder heads, in terms of material quality and construction, is rather disappointing (the bayonet-lock mounting solution snapped off after just a few uses, and the distance between the spinner plate and the outlet can only be adjusted within a very narrow range), but despite this I will still be applying this sensor and controller to my current dual-feeder large tank.
With this device I’ll be able to continuously monitor how full the tank is, fine-tune the feeding cycles to match the game’s movement patterns, and trigger a feeding at any time I choose.

Cameras:

I found a very significant difference between the Zeiss Secacam 7 and the Spypoint trail cameras.
While the Spypoint cameras upload whatever photos/videos were captured in a given period (however many there may be) to the cloud app at intervals matching the configured "Sync Frequency", the Zeiss follows a completely different logic.
With the Zeiss cameras, the "Communication interval" only serves to check status and apply any setting changes.
It plays no role in sending photos.
It’s actually the other way around: whenever a photo is sent, that’s when the status check and any setting changes are applied.

What’s very strange, though, is that the Zeiss camera sends the photo immediately whenever there’s an event.
You can set pauses between individual motion detections, and you can also limit the number of photos sent within an hour, or pause uploading entirely — but for example, if a sounder shows up at the feeder and I set a one-minute pause between two detections, roughly 60 photos would be taken per hour, and it would send each photo individually, once per minute, which means a huge amount of energy consumption.

If I limit the number of photos sent within an hour to, say, 5, then since I’ll only get the photos from the first 5 minutes, I won’t have any idea how long the sounder stayed at the feeder.
The 55 photos that weren’t sent will still be sitting on the camera, but the system won’t upload them to the cloud server. 🙁

Perhaps the solution is to set a 5-minute pause between two detections, which, if there’s continuous activity at the feeder, means 12 photos transmitted per hour.
That way you can determine, to within 5 minutes, when the sounder left the feeder and how long it stayed.

Sending 12 photos an hour isn’t all that many, but it’s exactly 12 times more than if it only sent the total volume of photos taken over the full hour just once.

The solution is probably to run the camera off a system that provides renewable power. . .