Capture Resolution and Memory Usage

[spammy]

I have a 4GB machine with 7 cameras of various resolutions. I have set up each with the source settings in ZM (capture width and height) set to what the camera outputs. Using the formula at https://wiki.zoneminder.com/Math_for_Me ... o_optimize, my total memory usage comes to 3.2GB.

I have some questions:

1) I have moved from Milestone Xprotect and while I immediately acknowledge the value ZM has brought us since the migration, memory wasn't a limiting factor in that set up (although I do now recall the disk being hammered). Is there a reason why it is here? Please note that this isn't a criticism, just a curiosity.
2) If I just adjust the Capture Width and Height, will this allow ZM to require less memory? Or does this have to be changed in the cameras and ZM always has to match that resolution?
3) If the latter, why do we need to enter it at all? Isn't resolution of the stream something that can be autodetected?

[spammy]

To answer my own question, I have put in a lower resolution (while keeping the aspect ratio) in the ZM capture settings and memory usage has reduced.

I'm still interested in some of the answers above though

EDIT: and what about disk (space and IO) and load? Would understating the resolution lower those too?