Exposure & Color
DYNAMIC RANGE · WHITE BALANCE
A
ΔE 3.4 delta-E
PRE-PUBLISH VIDEO QC
01 / READ THE CRAFT
Know exactly what to fix before you publish.
Give CutScore a file or a link. It reads the craft across 13 families, auto-detects your video type, and returns a 0–100 score, the evidence, and the concrete fixes — as a coaching PDF. For every kind of video, not just film.
13craft families
0–100objective score
A–Mevery axis, measured
8genres, auto-detected
Vlog
AUTO-DETECTED
3 fixes before you publish.
FIXES ADVISED · WEIGHTED FOR VLOG
Captions too small to read — hold them ≥ 2.0s02:14
Window blown out — pull exposure −1 EV02:32
Loudness is platform-perfect — −14 LUFS✓
TIMELINE · WHERE WE FOUND THEM
SCRUBBING
01 / SEE THE FIX
From flagged to fixed. Drag to see it.
The same vlog frame, before and after CutScore's top three fixes — recovered highlights, readable captions, and platform-safe loudness.
02 / BUILT FOR EVERY KIND OF VIDEO
One tool. Every genre — judged by the right bar.
CutScore detects your video type and changes its weighting and targets to match. A vlog and a brand film are not graded the same way — automatically. Tap a genre to see how the weighting shifts.
Vlog
Talking-head
Tutorial
Product demo
Reaction
Podcast
Short
Brand / Cinematic
03 / THE GAP
Why your videos look amateur even when the content is good.
The story can be great and still land flat. What separates it is craft — and craft is mostly measurable. These are the four most common reasons a strong video reads as unpolished.
Exposure drifts
Blown windows, crushed shadows, a white balance that swings shot to shot. The eye forgives a lot — the meter doesn't.
HIGHLIGHTS+2.1 EV clipped
Loudness is off
Too quiet next to everything else in the feed, or peaking into distortion. Platforms normalize to a target you can't hear by ear.
INTEGRATED−9.4 LUFS hot
Rhythm sags
Shots that hold too long, a hook that lands late, cuts that fight the energy instead of riding it. Retention leaks here.
HOOKlands at 0:11
Text is unreadable
Lower-thirds that flash by, captions under the contrast floor, type too small for a phone. Half your audience never reads it.
CONTRAST2.9:1 below AA
04 / HOW IT WORKS
From a file to fixes in three steps.
No timeline to learn, no preset to tune. Hand it the video; read the number; ship the fixes.
STEP 01
Give it a video
Drop a file or paste any link — YouTube, Vimeo, a raw export. Up to 4K. CutScore detects the genre on its own.
STEP 02
It reads the craft
Everything measurable is computed deterministically — the same numbers every time. AI handles only the subjective calls: aesthetics, text design, synthesis.
A B C D E F G H I J K L M
STEP 03
You get the fixes
A 0–100 score, per-family breakdown, prioritized fixes with timestamps, and the evidence behind every call — as a coaching PDF.
82 / 100
05 / WHAT WE MEASURE
Thirteen families of craft, A through M.
Each family is a cluster of objective measurements with a target. Together they describe the whole signal of your video — and every one of them applies to an ordinary upload, not just a film.
Sharpness & Technical
ACUTANCE · NOISE
B
0.78 acutance
Stability & Motion
SHAKE · MOVEMENT
C
0.11 shake
Composition & Framing
THIRDS · HEADROOM
D
0.92 ratio
Audio Loudness
LUFS · TRUE PEAK · R128
E
−14.0 LUFS
Audio Quality & Balance
SNR · SIBILANCE · STEREO
F
42 dB SNR
Voice & Speech
PACE · FILLERS · SILENCE
G
162 wpm
Editing & Rhythm
SHOT LENGTH · HOOK · CUTS
H
ASL 2.4 s
On-screen Text
READABILITY · CONTRAST
I
4.8:1 contrast
Narrative Structure
HOOK → BODY → CTA
J
0:08 hook
Platform & Metadata
CODEC · RES · ASPECT
K
16:9 · H.264
Accessibility & Safety
CAPTIONS · PHOTOSENS.
L
98 % captions
Overall Aesthetics
AI · SUBJECTIVE
M
AI judgment
See all 13 scored in a report
Each family links to a glossary entry in the full build.
06 / A DIFFERENT CATEGORY
Not a growth tool. A craft coach.
vidIQ and TubeBuddy help people find your video. CutScore makes the video worth finding. Different jobs — run both.
CutScore Pre-publish quality
- Measures the craft — exposure, loudness, rhythm, text, accessibility, across 13 families.
- Objective numbers from ffmpeg, OpenCV, PySceneDetect and EBU R128 — not vibes.
- Genre-aware weighting — a vlog and a brand film are judged by different bars.
- Concrete, timestamped fixes you apply before you hit publish.
Growth tools Discovery
- Keywords, tags and thumbnails to help people find the video.
- Trend data, competitor tracking, SEO for the platform's algorithm.
- Audience and channel analytics — what happens after publish.
- Silent on whether the video is actually well made.
08 / METHOD
Objective by design.
Anything that can be measured is measured deterministically — the same way every time. AI is reserved for the judgment calls a number can't make, and your footage is never handed to a third-party model to produce a score.
Measured, not guessed
ffmpeg, OpenCV, PySceneDetect and EBU R128 do the quantitative work. The numbers are reproducible — re-score and you get the same read.
AI only where it belongs
Aesthetics, text design and synthesis are subjective — that's where a model helps. Everything objective stays deterministic and reproducible.
Held to real standards
We grade against published references, not house opinion — so the targets mean something outside CutScore.
EBU R128 loudness
WCAG ≥ 4.5:1 text
−1 dBTP true peak
09 / EARLY ACCESS
Join the waitlist.
The analysis engine produces real coaching reports today. The hosted dashboard and API are in private development. Add your email for early access — no spam, just the launch.
You're on the list.
We've noted your email. Check your inbox for a welcome note — then we'll only email again when early access opens.
10 / QUESTIONS
Frequently asked.
No. CutScore judges the craft of your video — image, sound, editing rhythm, on-screen text and platform compliance. It doesn't do tags, keywords, or thumbnails for discovery. It's pre-publish quality control, complementary to growth tools like vidIQ or TubeBuddy.
Yes — that's the point. CutScore auto-detects the genre (vlog, talking-head, tutorial, product demo, reaction, podcast, short-form or brand film) and adapts its weighting so every video is judged by the right bar.
A weighted read across 13 families (A–M): exposure and color, sharpness, stability, framing, loudness (EBU R128), audio quality, voice and pace, editing rhythm, on-screen text, narrative structure, platform metadata, accessibility, and overall aesthetics.
Most issues are measurable. CutScore runs ffmpeg, OpenCV, PySceneDetect and EBU R128 deterministically, so the numbers are objective and repeatable. AI is used only for subjective judgment like aesthetics and synthesis.
The analysis engine exists and produces real coaching reports today. The hosted dashboard and API are in private development — join the waitlist for early access.