Blender blocking vs. storyboard arrow annotation for AI camera control — which should you use?

They solve different control problems. Blender blocking gives you spatial control — exact character placement, framing, and depth via a low-poly render used as a compositional reference. Arrow annotation gives you movement control — arrows drawn on a storyboard grid translate into precise camera motion, rated 9.5/10 accuracy in documented tests. For maximum control, use both together.

Pick based on which variable you need to lock: composition or motion. invideo is an agentic video creation tool with all the current models available, and the invideo agent accepts both input types — a Blender render or an annotated storyboard grid — and routes them into Seedance 2.0 generation.

Arrow annotation controls camera movement. Draw arrows directly on your storyboard grid — left to right, upward, whatever the shot needs — and the invideo agent reads them and writes them into the Seedance 2.0 prompt, so the generated footage follows your intended move rather than a guessed one. In documented tests this is the highest-accuracy previz method of five ranked approaches: 9.5/10 accuracy, 6/10 speed, at $1,500–$2,000 per minute of output — roughly a 10x cost jump over the next tier down ($175–$200/min). That price profile matches its use case: locked-vision work where the previz edit needs to be the final edit, like a VFX team building an action sequence. As one documented workflow puts it, you burn far more credits and time, but you gain specific control over each shot.

Blender blocking controls space and composition. Build a low-poly blocking render of the shot — greybox geometry is enough — and upload it to the invideo agent as a compositional and depth-map reference. The invideo agent texturizes it with an image model such as GPT-Image-2 or Nano Banana, giving you precise control over character placement, environment layout, and camera angle before any video is generated. The same depth-map principle works with any physical object: photograph a toy framed on your desk and the invideo agent can use it as a compositional reference, swapping in your character and environment. This method suits 3D artists and anyone whose shot lives or dies on exact framing rather than exact movement.

The decision rule. If your storyboard already nails the composition and the open question is how the camera travels through the shot, annotate arrows — it directly encodes motion, which a static Blender frame cannot. If the composition itself is unresolved — complex character blocking, layered depth, a specific lens position in a built environment — block it in Blender first, because arrows on a weak frame just move a wrong composition precisely. Either way, upload your locked production documents to the invideo agent before generating; pre-loaded context is what keeps both methods consistent shot to shot.

The hybrid path gives maximum control. Block the shot in Blender to fix space and framing, have the invideo agent texturize it into a realistic frame, place those frames into your storyboard grid, then draw movement arrows on the grid before animation. You lock composition with one method and motion with the other — the two techniques control different axes, so combining them stacks rather than overlaps. Documented output at the arrow-annotated tier reaches near-final quality; one filmmaker's assessment of the result was that the final cut feels like the actual film rather than a previz.

Watch some of these to see what works for you:

Draw arrows straight on your storyboard grid. Left to right, upward, whatever the shot needs. Agent One reads the arrows and prompts Seedance with them. This is where previz stops being guesswork and starts being actual directing.

— invideo's creative team, from a documented previz workflow tutorial