Sample video file for testing

A downloaded clip tells you little when something goes wrong. A test video shows its own timecode and frame number on every frame, so a player that drops frames, seeks to the wrong position, or starts a few frames late is visible at a glance, and a transcoded copy can be compared with the original frame by frame.

Each pattern is built to make one kind of fault obvious, and the sync beep and flash make audio drift something you can measure rather than guess at.

Every option maps to a real test. Choose the combination your upload pipeline, player or transcoder has to handle.

Each pattern stresses a different part of the video pipeline.

With Sync beep on, a white flash fills the frame for a tenth of a second at every whole second, and a 1 kHz beep plays at the same instant. In a correctly synced file, each beep starts on its flash. Step through frame by frame in a player, or measure it with a tool such as ffmpeg, to see how far a player, a transcode or a streaming pipeline moves the sound.

Browsers delay recorded audio by a small, steady amount, which differs by codec. In Chrome and Edge the recorder corrects for it: in testing, every beep landed within a frame of its flash, in MP4 and WebM alike. Other browsers record without the correction, so for precise sync work in them, use the ready-made downloads, which are made with ffmpeg and are exact. Pairs start at one second, because a beep at zero would have to sound before recording begins.

The left and right test plays the beep in one channel at a time, alternating each second, with LEFT or RIGHT on screen. It catches swapped channels, and pipelines that mix stereo down to mono.

Upload and processing: size and duration limits, accepted formats, frame rates and orientations, how long transcoding takes, and whether the processed video keeps its length, resolution and frame rate. Playback: start time, seeking to a timestamp, pausing and resuming, and switching between quality levels. Thumbnails: whether the generated thumbnail comes from the frame the requirement says.

Use each size in turn. Problems often appear only at 4K, where file sizes and decoding load are highest, only in vertical video, where players letterbox or crop it wrongly, or only at 360p, where scaling up exposes blurring and layout issues.

MP4 with H.264 video and AAC audio plays almost everywhere. WebM with VP9 or AV1 video and Opus audio plays in every current browser and on Android, but some older systems and editing tools reject it. The ready-made downloads come in both; the recorder offers every format your browser can produce, and says which you got.

Browsers write WebM files without a duration in the header, so some players show the length as unknown until the file has loaded. That is worth knowing, and worth a test of its own if your system accepts uploads from browsers.

Above 1440p, recording runs at the limit of most computers, because the browser encodes every frame in real time. If a 4K recording stutters or comes out short, lower the frame rate or the size.