Mixing game footage and model viewer.

[Waverian]

This is more a question of curiousity than anything else..

But for some of the more experienced machinima artists around, when you're using modelviewer to create a character with the intention of keying it over game footage, do you use any particular method to give it a more realistic appearance- proper scale, standing straight, right angle, etc.

Is it trial and error of filming multiple angles and sizes in viewer and trying them individually in the editor of your choice or just having a good eye and sense of how the character would actually appear?

[Thiras]

Originally Posted by Waverian

Is it trial and error of filming multiple angles and sizes in viewer and trying them individually in the editor of your choice or just having a good eye and sense of how the character would actually appear?

Mostly the latter for me. It just looks "right". Of course it gets wrong sometimes and then the only choice is to redo it.

[Ceejay9]

The angle is the toughest part, I've found. I use vegas 7, and I've found that to resize a character if it's too big/small, all I have to do is zoom in or out in the crop pane.

But the angle is the toughest part. And matching surrounding lighting and shadows.

[soosisti]

I take screenshots from my Vegas timeline (best settings) and then load the screen as a background to WoW model viewer. Then I align the character to correct position, hide the background and do my recording. This way I have gotten a lot better results than just by estimating the angle etc.