Rendering multi-camera sequences from one scene

[skovron]

What is the best method for rendering many sequences of animation (through different cameras) from one scene? Do I have to setup many passes, one for every camera? Or maybe should I prepare many scenes an use batch rendering?

I'd like to model one scene, setup some cameras with animation and then just render all sequences. In 3dsmax for example I used to create one scene for every scene in a movie and then just rendered many sequences using video post or batch rendering.
Is it hard to achive this way in XSI? How do You do this?

Thanks ( and forgive me my bad english :( )

[cryptosporidium]

I think its pretty simple to do that using passes. you have to setup one pass and make copy of it for each camera you have in scene. Then just hit Render all passes and thats it.

[skovron]

Thanks for your answer. Is it common method for rendering sequences? I mean do the proffesionals use passes or rather many scenes?
For me the "problem" with passes is this: "setup one pass and make copy of it". When I use for example 6 different passes and have got 8 cameras/sequences in a scene then if I change some pass for particular camera, I need to make same changes to other 8 passes. Well it gets harder to manage with many more cameras, and this is only for one scene.

[mantom]

Thanks for your answer. Is it common method for rendering sequences? I mean do the proffesionals use passes or rather many scenes?
For me the "problem" with passes is this: "setup one pass and make copy of it". When I use for example 6 different passes and have got 8 cameras/sequences in a scene then if I change some pass for particular camera, I need to make same changes to other 8 passes. Well it gets harder to manage with many more cameras, and this is only for one scene.

I would suggest creating a null for each camera you intend to animate, then constrain the camera to the null. Animate the null until you get desired results, then store the constraint in the mixer as an action source. Rinse and repeat for each camera null. After you've set up all your nulls (cameras), you'll have several action sources. Load each action source on it's own track as an action clip in the animation mixer and mute the track it resides on. When you need to render a camera out, simply unmute the track containing the camera constraint you'd like to work with.

Matt

[cryptosporidium]

skovron: im not sure what problem you got with passes. its the same like you would do separate scene for every camera (3ds max workflow), but you basically have them stored in single scene. You will finnish pass setup for one camera, and then you just copy it, no matter whether you got 1 pass or 10 passes per camera. Only thing you have to specify for each pass is a camera you want to use. Thats the advantage of using passes. You can basically keep more scenes in single one. And of course its more comfy to have all in one than having separate scenes for everything. Especially when your setup is done and you need to change something. And of course professionals use passes this way. Its one of big advantages over 3ds max... at least for me. Working with more scenes at the same time was pain for me.

[skovron]

Don't get me wrong. It's not that I've got some really big problems with using passes this way. I know It's huge win to be able to render many sequences from single scene. Actually I do it this way right now. I asked you about this because I felt I was doing this wrong. You know, I was wondering if I'm using right technique with right place and if this is some kind of hack or trick.
Ok so now I know it's way to go.

@ mantom:
I need to try te technique You have described.

Thanks.

[kim aldis]

Use the mixer to switch path constraints on a single camera. The advantage over using passes is that you can see and judge the cuts in the wireframe. See this article for details:-

http://kim-aldis.co.uk/drupal-6.1/node/7

[franky]

i always use one shot (one camera) in one scene to keep track of a more complex setup.
so i have Shot01_V001.scn, Shot02_V003.scn and so on.....
each shot is one camera move/setup.

the scenes usually use reference models so you still can make changes to the set/characters/whatever once and get it distributed through all shots.

[cryptosporidium]

skovron: well,as you can see here, there is no absolutely right way of doing something, neither there is absolutely wrong way. I prefer using passes, others prefer using mixer sequences, others rather create scene per camera. Its only what you personally find most efficient and comfortable and dont worry that most people might do it another way. Whats fine for others doesnt neccessarily have to suit you.

[franky]

amen!

the point is, in xsi you have the choice of YOUR way :)

[skovron]

Ok then I think I just need to explore all the possibilities and find out what the best for me :)