AI Video QC Scorer · Trust, but verify

CineQC

The broadcast eye for AI video.

Automated artifact detection across seven metrics, then a broadcast-trained scoring rubric and a defensible QC report.

Built by Marc Warfield — two-time Sports Emmy-winning editor, Emmy judge since 2017. Read the methodology →

100% client-side — your video never leaves this device.

Drag & drop a video — or several to batch-QC

MP4 / WebM / MOV · analyzed locally in your browser

Analysis runs best on desktop Chrome or Edge. On mobile you can review and read reports, but decoding a clip needs a desktop browser.

~3× faster — plays the clip at higher speed and samples fewer frames, so the metrics are coarser.
Steps through the clip seek-by-seek instead of playing it: every source frame is captured even if your machine is busy, and the same clip always gets the same scores. Slower than playback — the most defensible pass. Overrides Fast mode.
Mean Luma / per frame
Temporal Flicker / Δ luma
Color Drift / histogram distance
Sharpness / detail loss
Motion Coherence / block flow

Waiting for a clip…

Optional local AI quality gate

AV Sync · SyncGuard

Local engine · media stays here

Measure global AV offset, progressive drift, and speech-to-mouth alignment. A conclusive result affects the final CineQC verdict; unavailable or inconclusive analysis never penalizes the clip.

Not analyzed.

Module B — human rubric

Broadcast Scoring Pass

Automation can't judge anatomy, physics, text, continuity, or composition — so this is the pass your eye does. Click 1–5 on each category as you scrub the clip; the anchored definition explains each score, and the verdict below updates live. Temporal Stability is pre-filled from the automated flicker and motion scores — adjust it if the numbers miss what you see. Categories you leave unscored simply don't count toward the verdict.

—/100
Grade —
—
Sign-off