Visual Services
We clean, restore, and prepare your footage frame by frame — removing wires, rigs, blemishes, and unwanted objects, then prepping mattes and plates for flawless downstream compositing.
What we do
Paint & Prep is the invisible art of making VFX work look real. We remove everything that shouldn't be in the shot, repair damaged frames, and deliver pristine plates ready for compositing, colour grading, or final delivery.
Every frame is handled with hand-crafted precision — our artists use a blend of digital paint, clone tools, frequency separation, and AI-assisted infill to achieve seamless, undetectable results.
Stunt harnesses, support wires, tracking markers, and lighting rigs — painted out frame by frame so nothing breaks the illusion.
Blemish removal, tattoo hiding, continuity fixes, and object replacement. If it's in the frame, we can remove or change it.
Grain reduction, scratch removal, flicker correction, and damaged-frame repair to restore archive footage or fix on-set mishaps.
Pre-viz mattes, background extensions, and pre-keyed plate preparation delivered in the exact format your compositing pipeline requires.
Our Process
We review every frame, identify problem areas, and build a shot breakdown with estimated frame counts and complexity ratings.
Clean plates, location stills, and on-set photography are gathered to rebuild obscured backgrounds accurately.
Artists paint over unwanted elements using frame-accurate clone stamps, frequency separation, and AI-assisted infill.
Outputs are played at full speed and checked for temporal consistency, edge stepping, and motion smear before delivery.
Clean EXR, DPX, ProRes, or DNxHD sequences — delivered to your pipeline spec with naming conventions and colour metadata intact.
Methods
Traditional clone stamp, patch, and heal tools applied by hand on a drawing tablet — offering the greatest control for complex, moving subjects.
A clean section of background is tracked and warped to cover the problem area, following the camera motion automatically across the whole shot.
Generative models reconstruct complex textures behind removed elements, then refined by hand to remove any artefacts.
High-frequency detail and low-frequency tone are corrected independently — preserving texture while fixing colour or light inconsistencies.
Motion vectors from adjacent frames are used to warp background pixels into removed areas, creating seamless temporal consistency.
Temporal smoothing, grain matching, and de-flicker algorithms stabilise exposure and grain variance across long shot sequences.
FAQ
Ready to clean up?
Share a frame grab or shot description and we'll reply within 4 hours with a quote, timeline, and a test frame so you can see the quality before committing.