Independent, objective analysis, comparison and validation of digital video.
- Does your organisation use tools to analyse or convert digital video?
- How reliable and effective is that analysis or conversion?
- What changes occur silently, without your knowledge?
- Are your tools doing what they say they are?
FVT provides independent, objective analysis of digital video and a reliable way to compare originals against converted files. It then goes one step further, with automated single or multiple tool and method validation, against acceptance criteria that you define.
CCTV and surveillance video is everywhere, and comes in many different formats.
Often, to review, analyse and present video, it needs to be converted. Any conversion can silently alter the integrity and authenticity of the video.
Are you obligated to validate the effectiveness of a video conversion tool or method? SOPs, ISO 17025, FSR Codes of Practice?
The tool
One tool, three modes, single and batch.
FVT removes manual steps and mitigates against human error. It speeds up the process and marks the homework of your tools. It does not replace them. It increases assurance and compliance where required.
A detailed forensic report of a file's technical properties: container and stream details, frame-level analysis and GOP structure, with file hashes and per-frame hashes for integrity. Review the report and export the raw data as CSV.
A side-by-side comparison of two files, typically an original against its converted output. Built on the analysis foundation, it adds objective, perceptual quality metrics to enable faithful measurement of the conversion.
Compare many converted files against their originals automatically. You define the tools, the methods, and your acceptance criteria; FVT runs the lot and returns an overarching pass/fail report plus an individual report for every comparison. Days or weeks of manual work delivered in minutes, measured against your own ground-truth reference clips.
See it in action
The tool, and the report it produces.
Configure an analysis on your own machine. Every run writes a self-certifying, court-ready report.
Why FVT
Built to be trusted in court.
Five principles behind every measurement, the reasons a report holds up under cross-examination.
Objective
Reports the data as it finds it. No interpretation, no computed opinion, just the measurements.
Independent
Not associated with any tool manufacturer. FVT measures the output of any tool that converts video.
Local & deterministic
Runs entirely on your machine, no cloud, no AI, no telemetry. Same inputs, same results, every time.
Self-certifying
Every report embeds the FFmpeg binary SHA-256, a unique Report UUID, and a UTC timestamp.
Customisable
Tailor the report to your workflow, your branding and disclaimer, and which analyses and fields appear.
Who it's for
Wherever converted video has to be trusted.
Built for police forensic video labs, and just as applicable anywhere video evidence is converted and relied on.
Why now
It stopped being guidance and became the law.
Work in law enforcement, based in England or Wales? If you provide video evidence into the criminal justice system, the following apply.
Statutory since 2 October 2023 under the Forensic Science Regulator Act 2021; Version 2 in force since 2 October 2025. Validation of the software you use is required, not optional.
Commercial off-the-shelf software must be validated before use, including updates. FVT's Validate mode is how you evidence that, using your own data sets, against acceptance criteria that you define.
When digital evidence is altered, you must be able to explain your actions and their implications. An FVT comparison report is that explanation, original versus converted, in data form.
FVT wraps FFmpeg and libvmaf, open-source, industry-standard primitives that an independent expert can reproduce.
Technical
Runs on the hardware you already have.
No specialist infrastructure, no GPU clusters, no cloud dependency.
| Formats | H.264, H.265, AVI, MKV, MP4, MOV and more, anything FFmpeg can read, including awkward CCTV exports. No fixed format allowlist. |
| Operating system | Windows & macOS. Lightweight local install. |
| Hardware | Standard workstation. No GPU required (GPU acceleration supported). |
| Network | None required. Fully offline operation supported. |
| Engine | FFmpeg + libvmaf, open-source, auditable, industry-standard. |
| Outputs | Self-contained HTML and PDF reports, plus CSV exports, bundled as a ZIP. |
FAQ
Questions we're asked first.
No. FVT is deterministic: the same inputs produce the same numbers every time. Every report records the exact FFmpeg binary used (by SHA-256), so an independent expert can reproduce the result exactly.
FVT wraps open-source, auditable primitives, FFmpeg, libvmaf and standard cryptographic hashes. The numbers are the same a hand-written script would produce; FVT just produces them repeatably, with audit shape, and in a report your SOP can act on.
Minimal. FVT is a local-only tool: no external API calls, no telemetry, and no data leaves the machine. That materially reduces the IA surface compared with cloud-connected or AI-dependent alternatives.
No. FVT measures using open-source FFmpeg, so your historical evidence stays analysable with any FFmpeg-compatible tool. You set the standards and own the criteria, FVT does the measurement.
See it on your own footage.
A short demonstration for your team, run on your machine. No commitment.
Request a Demo