Rendering video from exr multi layer

I have been using exr multi layer and it’s really great. What I have been doing is adding an input image node in the compositer, and importing all the exrs, duplicating the node and selecting the layers I need and then wiring them all up in place of the render layers and they all work. But If I then want to render out video I am rendering the scene at full size but 1 sample, which just can’t be the best way. Is there a way to just ignore rendering, and use the wired image sequences?

You don’t need to render your scene, just delete the render Layer node using just the image input node as the source for the compositor