I wish to encourage lunacy that is Personal Animation Production.
How to make a render despite hardware(budget) limits common to personal productions?
Here are some techniques to help you get it done – these are possible in several softwares. This is my 2nd collection of Animation Production Tips.
Note that these are Tips and so very brief. Some could be expanded to full tutorials. I may do that for some, later. For now I just wish to get the ideas across.
Tips for animation production renders
- Render to image sequences(not video!) and combine to clips later. This is a no-brainer. There is so much you can do with the image sequences in post, and really to get the full benefit of post process you have to use image sequences. And don’t forget that if your video rendering crashes, you lose the video. But if your image sequence rendering crashes, you lose only one frame and can continue rendering from that point onward.
- Brake animation scenes to sub-scenes, render as separate passes and combine later. Main reason for this is to ensure you can render at all – a way to combat the oh-so common lack of resources. Other major benefit of having them all as separate passes is you can later adjust the look of each element and their timing. So, begin by breaking you scenes into character, background and effects scenes (and so on). Then render scenes separately. In each scene use full version of current main subject and very simplified versions of all the other stuff, if they are needed at all, and set them to render as masks/invisible/shadow casting depending on your needs. You might find different objects/effects work better with different lighting – here breaking to separate scenes helps too. In the end you bring all passes together in a video editor or in a post production tool.
- Break sub-scenes to sub-sub-scenes according to where the camera points at. This goes even futher in making sure you can make the render but isn’t needed if previous Tip is enough. If your camera shows only 180 degrees of your scene for a while, then that you can split that to a new scene and that’s where you can remove things. Always remove objects that won’t be seen, reflected, cast a shadow or otherwise have an effect to the render.
- Light scene without global illumination. Not using GI is one good way to save in rendering time. You may still want the GI-look. Check if there’s any plugins for faking it for your software. If there’s none, think how light would bounce around in your scene and place lights around to fake the effect. Like a red wall in a well lit room should have some red coloured lights pointing away from it(towards nearby objects, floor etc.). Granted this is more work for you, but it renders faster than GI and gives absolute lighting control.
- Create HDR-enviroment from you scene for rendering optimization. This approach only works in scenes where only your character moves and only in a limited area with no props(chairs, tables, anything) close by AND where your character is not casting a long shadow on the scene. If these limits suit your story, this tip is dynamite. In short you create a scene as normal but with one extra camera where the character is – lets call it character-camera. Hide your character and render one image of the scene with regular camera. Then use the character-camera to create a HDR-probe image(or similiar) of the scene, look up how to do this in your software, after which you no longer need the scene objects or lights – just the HDR and character. Create another scene with just the character and light the scene with your new HDR. Then render the character animation to image-series with transparent background. Combine this pass with your background image, the very first render you did, in post process.
- Move everything you can to post process. There is nothing unprofessional about ‘faking’ something in post process – only results matter. I would move all effects that are doable as 2D-versions to post production right away. 2D-effects are faster to set up, render very fast, and are non-destructive – you can change them without having to re-render your 3D-image. Lately post production softwares have begun to incorporate full or semi-3D tools tranferring even more power to post. You don’t necessarily need After Effects, Fusion or similiar specialized software, though – I have done lots of simple post processing in regular video editing software.
- Render different shading types as passes and combine in post process. There is no sense tweaking look of your render endlessly in 3D-software when you can do it faster and with more options in post process. So brake for example your character pass to image sequences for each shading type. They could be color, specular, reflection, ambient occlusion and depth. A good software should let you render these all at once. Fun part is that you can adjust intensity and pretty much everything else for each in post process and also how they interact as layers. One huge benefit is that you can use depth-pass to create depth of field effect very quickly, whereas it is very slow to render in 3D. In short you render once and get a million ways to change the result after. And no change you make this way is destructive.
Do you use these methods in your productions? Anything you would add or change? Is there a tip you’d like to see expanded to a tutorial?