Every control, with its real range.
No “magic” slider that hides what it does. Below is the complete list from the pipeline, including the step size the app randomises on, and the order the filters run in.
colour → vignette → noise → perspective → curves → zoom → rotation → flip → chroma shift → scale → speed
Audio is processed on its own graph, then the container flags are written last. The order is fixed so a preset produces the same structure everywhere; only the random draws differ per copy.
Frame controls
Each control can be switched on or off per preset. When it is on, every copy draws its own value from the range you set.
| Effect | Range | Step | What it changes |
|---|---|---|---|
| Brightness | −0.20 … +0.20 | 0.05 | Luma offset across the frame. |
| Contrast | 0.80 … 1.20 | 0.05 | Distance between shadows and highlights. |
| Saturation | 0.70 … 1.30 | 0.05 | Colour intensity, independent of brightness. |
| Gamma | 0.80 … 1.20 | 0.05 | Mid-tone curve without clipping the ends. |
| Vignette | 0.20 … 0.80 | 0.10 | Radial falloff toward the corners. |
| Film noise | 4 … 12 | 1 | Temporal grain, uniform or patterned, seeded per copy. |
| Perspective | 0 / 1 | — | Corner shear of the frame geometry. |
| Luma curves | 0 / 1 | — | S-curve remap of luminance. |
| Zoom | 1.02 … 1.12× | 0.02 | Crop then rescale, centred or anchored to a preset corner. |
| Mirror (hflip) | 0 / 1 | — | Horizontal flip of the whole frame. |
| Pixel shift | 0 / 1 | — | Offsets the colour channels by one pixel. |
| Random size | 0 / 1 | — | Width and height jitter within the preset bounds. |
| Random 9:16 size | 0 / 1 | — | Vertical framing jitter for short-form outputs. |
| Blurred border | 8 … 24 px | 4 | Blurred fill around a slightly scaled frame. |
| Speed | 0.95 … 1.05× | 0.01 | Timing change with matching audio tempo. |
Audio controls
Applied only when the source has an audio stream. A silent video skips this graph entirely rather than failing.
| Effect | Range | Step | What it changes |
|---|---|---|---|
| Volume | 0.90 … 1.10 | 0.02 | Overall gain, kept subtle. |
| Waveform shift | 1 … 20 ms | 1 | Offsets the waveform against the video clock. |
| Pitch | −1 … +1 st | 0.25 | Semitone shift without a tempo change. |
| Spectral tilt | 0 / 1 | — | Broad high/low emphasis. |
| Reverb | 40 … 120 ms | 10 | Short room tail. |
| Equalizer | −6 … +6 dB | 1 | Multi-band tone shaping. |
| Chorus | 20 … 60 ms | 5 | Modulated short delay. |
| Flanger | 1 … 25 ms | 1 | Swept comb filter. |
| Phaser | 1 … 10 | 1 | Moving notches across the spectrum. |
| Tremolo | 2 … 8 Hz | 1 | Amplitude modulation. |
| Vibrato | 2 … 8 Hz | 1 | Pitch modulation. |
| Compression | −30 … −10 dB | 2 | Dynamic range control. |
| De-esser | −4 … −2 | 0.5 | Tames harsh sibilance introduced by other stages. |
| Room echo | 20 … 50 ms | 5 | Early reflections without a long tail. |
| Soft limiter | 0.85 … 0.98 | 0.01 | Prevents clipping after the chain. |
Encoder and container
The layer that most “uniqueizers” leave untouched. RepurposeThis rewrites it for every copy, and can rebuild the metadata as US-region device output.
| Control | Range | Step | What it changes |
|---|---|---|---|
| Frame rate | 24 … 30 fps | 1 | Frame timing and container header. |
| Video bitrate | 1500 … 6000 kbps | 500 | Encoder bitrate, and therefore file size. |
| Audio bitrate | 96 … 256 kbps | 32 | Audio track encoding. |
| Trim start | 0 … 5 % | 1 | Trims the head of the clip per copy. |
| Trim end | 0 … 5 % | 1 | Trims the tail of the clip per copy. |
| US metadata | on / off | — | Strips source metadata and writes fresh US-region device tags. |
Image transforms
Applied to each file in the batch, with the source metadata removed before the new camera EXIF is written.
| Transform | Range | What it changes |
|---|---|---|
| Saturation | 0 … 3 | Colour intensity. |
| Contrast | 0.5 … 3 | Linear contrast around mid-grey. |
| Brightness | −0.5 … +0.5 | Exposure offset. |
| Gamma | 1 … 3 | Mid-tone response. |
| Grain | 0 … 20 | Per-pixel noise, seeded per image. |
| Vignette | 0 … 1 | Corner darkening. |
| Zoom | 1 … 2× | Centre crop and rescale. |
| Rotation | −10° … +10° | Straightens at a new angle, then re-crops. |
| Pixel shift | 0 … 20 px | Sub-pixel channel offset. |
| Crop | 0 … 20 % | Removes a percentage from every edge. |
| Blurred border | 0 … 100 px | Adds a blurred frame around the image. |
| Flip / flop | h / v / both | Mirrors the image. |
| Pattern overlay | grid / stripe | Low-opacity texture across the frame. |
| EXIF rewrite | on / off | New make, model, software and capture time. |
| Format | JPG / PNG / WebP / TIFF | Output encoding, with optional size reduction. |
Geometry and grain, not colour alone.
On most footage, colour tweaks barely move a perceptual hash. Small geometric changes and per-copy grain do. That is not folklore — it is what the Fingerprint Lab reports when you measure it.
- High impact: pixel shift, zoom, perspective, film noise, crop.
- Medium: luma curves, vignette, blurred border, rotation.
- Low alone: brightness, contrast, saturation, gamma.
These are tendencies, not guarantees. The lab ranks effects on your source because a still portrait and a fast-moving clip do not respond the same way.
More is not automatically better
A preset that pushes every slider to the limit produces visibly degraded copies, and platform moderators notice artefacts faster than hashes notice colour. The useful strategy is a few strong geometric effects, a light grain pass, a real audio change and a rebuilt container.
Start from a built-in preset, run 10 copies, look at the contact sheet, then widen only the effects that measured low in the lab.
Bring your own source.
Ten copies on your footage will tell you more than any table. The app is the same $25 either way.