Is there a way to set Blender up before a render so that when I press F12 it will render a camera view, save that to an image file then switch to a 2nd assigned camera and render that+save it, and so on with more cameras? I have long renders that I need alternate camera angles to that I want to run over night.
Send copies of a scene to VSE, each with a different active camera. No need to interleave frames is there? Just render back to back? Would take same amount of time.
Or the compositor with multiple scenes rendered to File Output nodes
You can also animate the camera, see the video of Viralata āT&T Blender for Architecture nĀ°5 : the cameraā
see http://vimeo.com/23575591
Hereās a quick tut.
- Set up a few cameras as you like.
- Select first camera and make it active (ctrl+0)
- Place the timeline on the frame where you want the camera to become active and place a Marker (M)
- With marker selected, bind camera to marker (Ctrl+B in Timeline)
- Select next camera and make it active (ctrl+0)
- Repeat from step 3 and so onā¦
Hereās a simple 3 camera scene
In camera view play/scrub the timeline to see the camera switches
3-cameras.blend (496 KB)
Simple animation using three cameras in one continuous render
So the first 4 frames can each have different active cameras and the output can be a PNG frame sequence?
Yesā¦ if you just click Animation instead of Render on my test file it will render the animation which only has three frames, each with different cameras.
Youāll get three pngās (camera0001.png, camera0002.png and camera0003.png)
You can also specify these things as command-line parameters to the Blender program.
Believe it or not, and being the nerd that I am, I use Unix makefiles to generate portions of a film, and I have written some Python scripts to auto-generate portions of those files. So, I can literally start a job that consists of: āmake movie.ā Change stuff around, and issue āmake movieā again, and āmakeā will determine what needs to be re-done and what doesnāt, and in what sequence.
āLinked Scenesā can be useful here, with a base Scene in which you set everything up, and shot-specific Scenes that are linked to it, e.g. for specific camera-setups.
Instead of trying to render one shot that contains camera switching, all of the shots are handled separately, and editing is used to put-together the final piece. Furthermore, this editing takes place before final-rendering of the shots that will go into the edit. OpenGL quick-render the sequences, note the shots & cameras and frame-number ranges from the final cut edit, then detail-render only what you have determined that you need.
Well the animated link camera to marker trick works perfect for what I want to do. Iām not doing much animation yet just different schematic views of models. But it will work fine for animations too when I get more in to that. Thanks guys!
If youāre shooting ājust one frameā or somesuch, a little script or what-have-you can be useful. But when your projects get a bit more complicated (and they will ā¦) the customerās/bossās going to want to ātweak just one little thing ā¦ā and plan your workflow in such a way that you can accommodate that.