Get in touch
Method
In practice

One frame, start to finish A Level 3 post-viz, and the last stretch done by hand.

A 1970s street, a hero car, a Rolls-Royce behind it. The look was agreed on AI stills; the last stretch was done by hand. Step through what that means on one frame.

Where it sits in the method: post-viz, atLevel 3 on the scale.

Level 3: IMG 2D refine & layering

Press play, or step through it.The steps, in order.

A 1970s street of luxury shops, palm trees and a silver sports car in low sun, with a black Rolls-Royce behind it

The frame

The generation we chose: the right light, the right street, the right era. This is a Level 3 post-viz still, made to agree the look before anything moves.

Where it’s wrong

Look at the Rolls-Royce behind the hero car. The details a car person would spot at once are the ones AI gets wrong most often.

  1. 1 Headlights: the two pairs run into each other
  2. 2 Grille and mascot off the centre line
  3. 3 Headlights: the wrong shape

Look closer

At full size the mistakes are plain. A new prompt would give a new car, and a new set of mistakes.

The fix

Not a new prompt. The chosen frame is kept and the mistakes are corrected by hand: painted and layered on the still, in 2D. Drag the line to compare.

The frame, fixed

The same generation, with the details right. That is what Level 3 buys: the look from AI, the control from the artist.

Level 3 on the scale: IMG 2D refine & layering

Your frame next?

Send us a still. We'll tell you what AI gets right, and what needs a hand.

Share a frame, a board or the script on WhatsApp. You'll get a level for the shot, what the AI route can carry and where the artists come in, before anyone spends on the expensive part.

Let's talkConnect on LinkedIn