I’ve noticed more than a few articles on software-based camera mapping in the Blender literature, and after giving it some thought, I beleive I could produce a kit that would attach to most tripods to make tilt and pan angles available to your computer as joystick inputs.
I’m wondering if there’s enough interest in something like this to make it worth my time and money (I don’t really have much of either). I’m thinking right now that I’d be able to sell this for something between $100 and $150.
Any comments? Anyone out there think they’d be willing to pay for something like this?
He means he’s gonna use 2 potentiometers (a few cents each), which you can attach to a tripod and then record the “joystick” movements.
That data then could be used to control the camera movement in blender.
And all that for a lousy $100.
I doubt there’s real use for that because the really interesting part would be to MOVE the camera around. But that’s not possible with his idea.
It’d be far more useful (and fun) to attach an acellerometer to the camera, and feed blenders camera the data, that way you could motion capture the camera as if you where holding it, making camera moves far more realistic. (see…I’m giving ideas away for free…)
To get realistic and nice camera moves you need a heavy camera.
So how about tieing an iPhone or iPod Touch to a large brick and pretending it’s a camera, then write a small app for the iPhone/Touch to record the accelerometer data and a python script to get it into blender.
Something more complicated but with interesting potential would be to get a two way communication going between virtual camera and Blender. you move your cam that has a lcd screen on that you can view a 3d scene through and it sends the movement back to Blender -ie a cheap version of what James Cameron used for Avatar.
@spacetug I think yellow meant using the ipod touch’s accelerometer (no need for an actual camera) the brick was to give you the weight of a real camera which would affect how you moved it.
I think he’s talking about getting camera motion without having to do tracking, but it’s pretty useless for a 3d track unless he can find a way to get the camera’s position, not just the rotation.
It seems like a good idea. May seed onto other greater things. All the best for your project. I’ve have to research it before deciding if I can spare $1.50 from me wallet. Oh wait, it’s $150.00.
Sorry I didn’t get back to this thread sooner. My concept is for a tripod mount, a bit more complicated than potentiometers (most computers these days don’t have game ports, you have to and provide USB inputs. Most of the cost was going to be in custom mounting hardware (created in Blender and fabricated at Shapeways), plus the cost of the interface board.
This would only work for a stationary camera, and possibly provide zoom information (via contacts mounted on the zoom “paddle” on the camera). If you want to be able to support a steady-cam, that would be more than a matter of an accelerometer, you’d need a gyro to get the pan angle (tilt you can get from an accelerometer), and you really need to use a full-bore inertial navigation system (several hundred dollars), in order to integrate roll, yaw, and tilt rates with x, y, and z accelerations to get attitude and position.
I mostly do archvis stuff and I would find this incredibly useful. Camera tracking software seems to be a little limited for simple pan/tilt tripod shots so if I could film a scene from my tripod and just import the data to Blender it would save me a lot of work and (hopefully) look even better than I can do it at the moment.
I’m interested to see if you get anywhere with this - good luck!
