im just getting into still image compositing and curious. when doing (just my current project for ease of use) a light saber and i blur the blade of a single still image, how does that translate to an animation sequence (say swinging the light saber to get that characteristic arc of light.) do i have to manualy composite each and every image in that image sequence?
Hello, I’m not sure what you asked, but I think you should animate the object, then assign keyframe and if you did a good compositing you’ll get the result! When you click the button “ANIMATION” compositing is applied to your image just ended each keyframe, and then processes the render will see the render without compositing, but when finished, will be applied and saved as an image
so each of the individual images it has in the animation render with have the compsiting attatched?
Yes. You set your movie node to Image Sequence and set the frame range. You want to render your initial animation as a series of PNG images (don’t use JPG because compression artifacts can make compositing difficult.)
ETA: Oops! You’ll find the Image Sequence in the Image node, not the Movie node.
In the Star Wars movies, the moving light sabers always had a lot of motion blur. The problem is that rendering motion blur can increase render times considerably.
… but there are other ways to do that, much less expensively, with compositing.
For instance, you can isolate “just the saber blade, ma’am,” so that you have one channel (so to speak …) of information in the eventual “mix-down” that consists only of the blade. You can now subject “just the saber blade, ma’am” to, say, a simple blur, or a vector blur. (This presupposes that you are using “MultiLayer OpenEXR” as your file-format, and that you are capturing the vector-pass along with everything else.)
You can also take this opportunity to change the color or tint of the blade, which initially might have been rendered in white.
It is ordinary practice to render image sequences as individual files, within a designated subdirectory, rather than to render to a “movie file” format. If the render crashes for any reason, you can simply restart it where it left off. If you have several computers working, each one can work on their own individual frames.
The individual files might contain much more information than a “movie file” needs or would contain. None of them are compressed. They’re just “files of floating-point numbers.” The “final cut” of the movie is also a whole lot of files in a subdirectory.
The “answer prints” that finally gets delivered or shown are generated from that original, in the very last step of the process. Many different versions might be called-for, with more or less compression and so-forth. (Notice that this final step is the only place where information is actually being “thrown away” to accommodate the requirements of a particular deliverable movie-file. And of course, nothing is actually being “thrown away” at all, because the original remains unchanged.)