I’ll chalk this up to my lack of post-processing knowledge in Blender.
I render out an OpenEXR Multilayer image sequence. I then want to import that sequence into the compositor and work some post- magic. I don’t see how to save out the resulting image sequence without having to re-render the entire animation again.
For still images, I’ve been going into the UV/Image Editor and “Save as Image,” but I’ve never done this with an image sequence, and I really don’t understand the workflow.
You already have a sequence of rendered images? No need to re-render.
The easiest way to let the compositor do it’s thing is to start a fresh new blender file. Then open the node editor, click the “compositing nodes” button (the one right by the “material” and “texture” buttons), check “use nodes,” and then add a movie clip input node. Press the button with a file folder icon to open a browser and select all the images in the pre-rendered sequence. Each frame that you now render will come from that image sequence.*
Alternatively you can use the video sequencer (without the compositor) by adding–this time–an image sequence strip.
When you are ready to process all the images, press the “Animation” button to render instead of the “Render” button.
*This works because at the bottom of the render panel (the tab in the properties pane with a camera icon on it), there is a “Post Processing” section and “Compositing” and “Sequencer” are checked by default.
Note: make sure the location you are saving the new images to is different from the location with the pre-rendered images. Be careful not to overwrite your original images. You might even want to uncheck “Overwrite” in the Output settings.
yes, of course, the only reason see360 suggested that is that a new blender file does not have any compositing nodes already in which might confuse the explaination he gave. But you can even hold two separate compositions in the node editor as only the one you plug into the composite output is the one used.
If you are using multilayer EXR, they even have all the different render passes (you actually rendered) in there, so that should give you a lot of control.
However, I would use the Image Input node instead. If you open a range of images, the node will give you a selection of Single Image which you can switch to Image Sequenze. Then you will get the number of frames to render (from the composit) a start frame and an offset and also cyclics (if you want to do multiple runs through an open end).
This way you have more control than in the movie node I expect. Look for matching the timeline with the input (as well as resolution, etc.). And as you can keyframe composite effects, you can change them over the duration of the animation you render out.
Yeah, my plan is to use OpenEXR Multilayer due to having the render layers available in the file so I can have the control of a fresh render without having to use many folders full of PNG render layer image sequences.
However, when I finish compositing, all I have is a preview image in the Viewer Node. If I hit “Render” or “Animation” it starts rendering the 3D scene again, THEN applying the compositing. I already have a rendered 3D scene–I just want the compositing effects to be rendered into a PNG sequence.
You’d need to still make a fresh scene. You can just add one to the same blend file if you want to. I like using a different blend though, just so I don’t have to load up the giant project file just to edit the composite though.
Also make sure in the Node Setup you have the output connected to a Composite Node in order to be able to render out the files with ‘Compositing’ checked in the Post Processing panel…
Maybe that’s trivial, but I didn’t see this mentioned above.
For animation & rendering, use 1 file. For post-pro compositing, use a new file. Separate areas, separate files. Files aren’t unnecessary big. And when something happens to 1 file, at least you still have the other.