Video Blackout

Video Blackout — Free Download. Video redaction

Video Blackout is desktop video redaction software for Windows 10 and Windows 11. It detects faces and license plates on your computer, lets you draw manual boxes for badges, screens, or other identifiers, and mark audio ranges to mute or beep. Export writes a new H.264 MP4 with mosaic, blur, or solid fill burned into the frames. The original file is never modified, and footage is not uploaded to the cloud. After install and license activation, you can work offline. Free mode includes a watermark and length limits. Paid plans remove the watermark and unlock longer videos, batch processing, and extra devices. Built for bodycam review, CCTV, FOIA releases, GDPR and DSAR responses, and any workflow where sensitive video must stay on the machine.

★★★★★
5.0(1 ratings)
Download Video Blackout (Official links)
File size: 130 MB
The latest version of Video Blackout is: 1.0.0.0
Operating system: Windows
Languages: English
Developer: FlyAllRisk
Price: $0.00 USD ($29 or $69 per month or free version)
  • Irreversible pixel rewriting. Every export rewrites the actual pixel values in the output H.264 MP4 file. Masked regions are not stored as a separate layer, metadata flag, or removable overlay. A video editor, hex editor, or frame extraction tool cannot restore the hidden content because the original pixels are replaced during encoding. This differs from timeline-based masks in non-linear editors that can be toggled off by anyone who opens the project file.
  • AI face detection. The application includes on-device neural network models that scan video frames and propose bounding boxes around human faces. Detection runs locally without sending frames to a remote server. Operators can accept or reject each proposed box. Faces that should remain visible can be unchecked, so only selected individuals are masked in the exported result.
  • AI license plate detection. Separate detection models identify vehicle license plates in footage from CCTV cameras, body-worn cameras, dashcams, and handheld recordings. Proposed plate boxes can be reviewed before export. This function supports law enforcement release workflows, insurance claim processing, and public records requests where vehicle identifiers must be removed before publication.
  • Manual box drawing. When automatic detection misses a face, plate, badge, computer screen, document, tattoo, or any other identifier, the operator draws a rectangle directly on the frame. Manual boxes are treated identically to detected boxes during export. This function covers edge cases such as reflections, unusual angles, partial occlusion, and objects that detection models were not trained to recognize.
  • CSRT object tracking. After a box is placed on a frame, the CSRT (Channel and Spatial Reliability Tracking) algorithm follows that region through subsequent frames. Tracking reduces the amount of manual repositioning required when a subject moves through the scene. The operator can correct drift at any point by adjusting the box, and the tracker continues from the corrected position.
  • Static rectangle lock. A rectangle can be locked to a fixed screen position for a defined time range. This function is used for stationary objects such as wall-mounted screens, license plates on parked vehicles, signs, whiteboards, or computer monitors that remain in one place while the camera stays still. The locked box is applied to every frame inside the chosen range.
  • Manual audio range marking. The timeline allows the operator to define spans of audio that should be muted or replaced with a beep tone. The application does not perform speech recognition or automatic transcription. All audio redaction decisions are made by the person reviewing the footage, which keeps the process predictable and auditable.
  • Mosaic masking style. Selected regions can be exported with a mosaic effect that replaces the original pixels with a coarse block pattern. Mosaic is a common requirement in broadcast, law enforcement, and privacy compliance workflows. The mosaic is rendered into the exported frames and cannot be reversed by image processing tools.
  • Blur masking style. A Gaussian-style blur can be applied to selected regions. The blur radius and intensity are controlled by the application, and the result is written into the output file. Because the pixels are permanently altered, the blur cannot be removed by sharpening filters or AI upscaling tools after export.
  • Solid fill masking style. Regions can be covered with a solid color fill, typically black or white. Solid fill provides the strongest visual guarantee that no information from the original pixels remains visible. This style is used when policy requires complete occlusion rather than partial distortion.
  • Batch processing. A folder of video files can be added to a processing queue. The application applies the same detection and masking configuration to each file in the list. Batch mode reduces repetitive work for organizations that handle large volumes of bodycam footage, CCTV archives, or disclosure packages.
  • Offline and on-premise operation. After installation and license activation, the application functions without an internet connection. No video frames, audio samples, or metadata are transmitted to external servers. This function supports air-gapped workstations, secure facilities, and jurisdictions with strict data residency requirements.
  • Original file preservation. The source video is opened in read-only mode. Export always produces a new file, so the original recording remains unchanged on disk. This function supports chain-of-custody requirements, forensic review standards, and any workflow where the unaltered original must be retained as evidence.
  • Keep selected people clear. In group scenes, the operator clicks the faces that should remain visible. All other detected faces are masked during export. This function is used when a subject or interviewee must be shown while bystanders, minors, or protected individuals are obscured.
  • Selective detection modes. Detection can be limited to faces only, plates only, or both. Operators can disable detection categories that are not relevant to a specific file, which reduces review time and prevents unnecessary boxes from appearing in the results list.

Video Blackout is developed by FlyAllRisk. The project was created to address a specific gap in video privacy tooling: most redaction workflows either depended on cloud services that require uploading sensitive footage, or on professional editing suites whose mask layers can be removed by anyone with access to the project file. The first public release appeared in 2024, targeting Windows 10 and Windows 11 desktop systems. The application is written primarily in C++ with supporting components in Python for the detection pipeline. The neural network models run through an embedded inference runtime, and the export stage uses FFmpeg libraries for H.264 encoding.

Latest News

Latest Programs