Blender Camera flaw?

I am currently building a world with mountains. I have a simple box (temperal character) with movement controls and a camera that follows. As the box moves up any gradual incline the camera wants to see from below (only). It will ignore the height setting and watch the box go up the incline from below- it seems the camera does not have proper Z-axis controls.
I experimented with the grid and made it higher than everything else. The camera then will view the box from a birds eye view and follow it along the X, Y axis. This showed me that the camera is in relation to the grid and not any object you attach it to.
Is this a unworkable flaw, or is there something that could be done to fix this? (Also I have tried parenting the camera to the box which causes them to fall through the ground even when trying numerous force settings.)

the default is it being set to the grid…i once found a tutorial on how to make a third person camera…sorry i cant find it…

try looking for it…if you dont then…look at first person camera and see what they do…because they also cant just run into the landscape…

worst comes worse you make a script…or use a diff game engine

I am not 100% for sure what your trying to do.

I know for alot of the things i have done I will make a empty.
Camera to empty. empty to object. Then you can have any camera options or movements done with the empty, I belive the camera can still track this way too.

If any of that helps

try searching for “phil demo” it has a pretty good 3rd person camera

There are two ways of having the camera track an object, to simulate a third person view.

  1. Using ‘slow parent’.

Parent camera to the object which is your main actor. Select the camera. Go to ‘Object’ buttons (F7), In ‘Anim settings’ press the ‘SlowPar’ toggle button. Now deselect ‘Offs Par’ (Offset parent - You don’t want the time offset work on your actor, only the camera). Finally, add a number in the TimeOffset box. For example 20.

Change the numbers and see which works best for you. Press ‘p’ to see the effect.

  1. Using the ‘Camera’ actuator.

Add logic bricks to your camera: ‘sensor’ (set to ‘Always’), connect to ‘controller’, connect ‘controller’ to ‘actuator’ (set to ‘Camera’).

You can fill in the height, minimum and maximum distance to tracked object, and also fill in the object to be tracked.

There is a limit to the height and distances you can set. If your tracked object is too big in camera-view, scale this object down.

victor

Things always work best if you use an empty to control objects not the object it self

Dr S