Pipeline

What VFX needs from your shoot — the one-page list

Five minutes on set saves five days in post. A practical list of what to capture, why each item matters, and what it costs us when you cannot get it.

  • 18 June 2026
  • 4 min read
  • Aniwala Studios

Most of the difficulty in visual effects is not simulation or compositing. It is reconstructing information that existed for free on the day and was not written down.

This is the list we send before a shoot. It costs almost nothing on the day. Print it, give it to whoever is closest to the camera, and post gets meaningfully cheaper.

The five-minute list

1. A clean plate. Roll the same shot with the actors and any rigs removed, camera untouched. Ten seconds is plenty.

Why: Anything we need to remove or replace has to be reconstructed from somewhere. With a clean plate, that is a copy-and-paste. Without one, it is hand painting, frame by frame, and the cost scales with the length of the shot rather than being flat.

2. A chrome ball and a grey ball. Held in the scene, in the same light, shot at the top or tail of the setup.

Why: The chrome ball tells us where every light in the room is. The grey ball tells us how bright and what colour they are. Together they let CG match the plate lighting on the first attempt instead of the fifth. If you can also shoot a bracketed HDRI panorama, better still — but the two balls get you most of the way for thirty seconds of work.

3. Lens data. Focal length, aperture, filters, and whether anything changed mid-shot.

Why: Lens distortion has to be undone before a camera solve and reapplied afterwards. Knowing the lens turns that into a lookup. Not knowing it turns it into a solve with an extra unknown, which is slower and less accurate.

4. Camera height and tilt, plus a tape measure to something in frame.

Why: A camera track with no scale reference produces a scene that is geometrically correct but sized arbitrarily. Every CG element then has to be scaled by eye, which mostly works until something is dropped, thrown, or has gravity applied to it — at which point it falls at visibly the wrong speed.

5. Tracking markers, and a photo of where they were.

Why: Markers on a featureless surface make a solve possible at all. The photo matters because we need to remove them afterwards, and knowing exactly where they were saves the painter finding each one by hand.

6. Reference stills of the set from four angles.

Why: Set extension, matte painting and any digital double of the environment all start here. Phone photos are fine.

7. A note on frame rate, shutter and any speed ramps.

Why: Motion blur has to match, and a ramp that nobody flagged will produce effects that blur wrongly in exactly the moments people are looking hardest.

What it costs when you cannot get it

Sometimes the shot is stolen, the location is a live street, or the schedule collapses. That is normal, and it is not a disaster. But the cost moves rather than disappearing:

Missing Roughly what it adds Clean plate Hand paint per frame — scales with shot length Chrome and grey ball An extra lighting iteration on every CG shot in the setup Lens data Slower, less stable camera solve Scale reference Guessed scale; visible in anything with physics Marker positions Manual marker removal

We would rather quote knowing you have none of it than quote assuming you have all of it and discover otherwise in week two. Tell us what you got and what you did not, and the number will be right the first time.

The one that matters most

If you can only do one thing on the list, shoot the clean plate. It is the cheapest item, takes the least time, requires no equipment, and it removes the single most expensive category of post work there is.

Nothing in visual effects is as costly as recreating something that was standing in front of the camera and nobody photographed.

Ask us before, not after

The other half of this is timing. A five-minute call during prep is worth more than a week of consultancy in post. We will tell you which shots are cheap, which are expensive, and — occasionally — which practical effect would be better and cheaper than anything we could do afterwards. That is a real answer we give, and we would rather give it before the money is spent.

Comments

Loading comments…

Leave a comment

Your email address will not be published. Required fields are marked *

Held for review before it appears.

Keep reading

More from the blog

All posts

Next step

Tell us the deadline first.

We'll tell you straight whether we can hit it, and what it would take if we can. No deck, no discovery phase. Just an answer within two working days.