Découpeur vidéo

Trimming with this tool is not lossless in the container-edit sense used by tools like FFmpeg's stream copy mode. The tool works by seeking a hidden video element to your chosen start point, playing it forward to your chosen end point, and recording the live stream through the browser's MediaRecorder API. That process re-encodes the video using the best codec MediaRecorder can access on your browser and platform — typically VP9 with Opus in WebM, or H.264 with AAC in MP4. Re-encoding is a generational process: even at the highest quality the MediaRecorder bitrate allows, compressed artifacts from the source file become inputs to the new encoder. For most practical purposes — social media clips, presentation excerpts, tutorial segments — the quality difference is imperceptible. For video production workflows where the trimmed segment will be re-encoded again downstream, the accumulated loss can become visible, particularly in high-motion scenes with dark gradients. True lossless trim requires a container-aware tool that can cut on keyframe boundaries without touching the encoded bitstream. FFmpeg with the -c copy flag achieves this for MP4 and WebM sources. The advantage of this browser tool is that it requires no installation, runs entirely in your browser with no file upload, and handles any video format your browser can decode. Practical tip: if your workflow involves multiple editing passes, do all your cuts and trims in this tool first in a single session, then do one final export — minimizing re-encode cycles reduces the cumulative quality impact.

Speed distortion in the trimmed output is caused by a mismatch between the MediaRecorder's capture rate and the source video's native frame rate. The pipeline plays the source through a video element and records its captureStream(). The captured stream's timestamp track is written by MediaRecorder based on real elapsed time during recording. If the browser throttles the tab — which happens in background tabs in Chrome, Firefox, and Safari — the video element plays slower than real time. MediaRecorder continues timestamping at wall-clock rate, so frames appear denser in time than they should. The result plays faster than intended when the tab was throttled. Conversely, if the source video has variable frame rate content — common in screen recordings and some smartphone-recorded MOV files — the decoder may deliver frames at inconsistent intervals, and MediaRecorder samples them unevenly. The most common cause is users switching to another tab or app during the trim operation. The fix is straightforward: keep the WikiPlus tab active and visible in the foreground throughout the entire trim process. Do not minimize the window or switch to another application until the progress indicator shows 100 percent and the download prompt appears. On macOS, disabling App Nap for the browser in Activity Monitor can help on older systems. Practical tip: start a trim on a test 30-second clip first to verify your setup produces the expected duration before processing a longer recording.

The tool currently supports a single continuous trim selection per session. You set one start point and one end point, export the clip, then reload a new file or set new in and out points for a second clip. Joining multiple trimmed segments into a single file is not handled within the trimmer itself. For non-destructive multi-segment editing entirely in the browser, the practical approach is to export each segment separately as individual files, then use the Video Compressor tool on each segment if size reduction is needed. Combining those segments into one final file currently requires a local tool. FFmpeg handles this cleanly: create a text file listing each segment path in order, run ffmpeg -f concat -safe 0 -i segments.txt -c copy output.mp4, and all segments are joined without re-encoding in seconds. Alternatively, the browser-based video editor CapCut Web or ClipChamp in Windows handle multi-segment timelines entirely in the browser. A multi-segment mode for the WikiPlus trimmer is a frequently requested feature and is on the development roadmap. All trimming and any joining runs locally — no file is uploaded at any point. Practical tip: name your exported segments with a numbered prefix like 01-intro.webm, 02-main.webm, and 03-outro.webm before running the FFmpeg concat command — this makes the ordering unambiguous and avoids accidental sequence errors when passing them to the concat list.

The tool does not work with DRM-protected content, and this is both a technical constraint and a legal boundary. DRM-protected video — content from Netflix, Disney Plus, Amazon Prime, HBO Max, Apple TV Plus, and most streaming services — is delivered through encrypted media extensions, a browser API that decrypts the video stream only for the platform's own protected renderer. The decrypted frames are never exposed to JavaScript APIs like canvas.captureStream() or MediaRecorder. Attempting to load a DRM-protected URL in the tool will either produce a blank video element or fail to load entirely. Files obtained by circumventing DRM protection are covered by anti-circumvention provisions in copyright law in most jurisdictions, including the US Digital Millennium Copyright Act and the EU Copyright Directive. Processing those files using any tool, including this one, may constitute infringement. For legally obtained personal recordings — screen recordings of your own content, recordings of events you filmed, video files you purchased in unprotected formats, Creative Commons licensed material, or public domain content — the tool works without restriction. Files stay entirely in your browser, so no content is transmitted anywhere. Practical tip: for videos you own or have licensed, ensure you have the video file in an unprotected format such as MP4, MOV, MKV, or WebM before loading it into the tool — the trimmer works with any format the browser can decode natively.