Audio & Video

Converts between MP3, WAV, FLAC, AAC, OGG and Opus with bitrate and sample-rate settings.

Cuts a section out of an audio file on the waveform, with fade in and fade out.

Renders an SRT, VTT, ASS or SMI file permanently into the video frames with size, position, colour and outline control, drawn with the browser’s own fonts so Korean needs nothing extra.

Pulls the audio track out of a video as MP3, WAV, M4A or FLAC, or copies it without re-encoding.

Re-encodes to a target file size or a quality preset, stepping the resolution down when the bitrate would starve it, and shows the projected size before it starts and the actual size after.

Reads MP4, MOV, WebM and MKV and writes MP4 (H.264 + AAC) or WebM (VP9 + Opus) with quality and resolution settings. When only the wrapper changes — MOV to MP4, MKV to WebM — the streams are copied untouched.

Crops the frame with a draggable window locked to 9:16, 1:1, 16:9, 4:5 or free, or pads it with bars to a target ratio, then re-encodes on your device.

Assembles clips into a YouTube or Shorts upload in the browser: trim, order, transitions, a music bed and subtitles. Same-batch AI clips with hard cuts export losslessly without re-encoding.

Turns a clip into an animated GIF or WebP with frame rate, size and palette controls.

Cuts a clip between two timestamps, copying streams without re-encoding when the cut is on a keyframe.

Shifts pitch and formants independently, so a recording reads as a different person rather than as a sped-up tape. Presets for male↔female, child, giant, helium, robot (ring modulation), telephone (300–3400 Hz band) and whisper (noise excitation). Records from the microphone or takes a file; writes WAV.

A chainable rack in one page: reverb by convolution with generated impulse responses, delay, chorus, flanger, EQ, compressor, fade and peak normalisation. The order of the chain is editable, because it changes the result.

Joins audio files end to end or mixes them onto one timeline with per-track gain.

Subtracts what is identical in both channels, band-limited so the bass and the top survive. This is not stem separation: it removes the centred kick, bass and snare along with the vocal, leaves any stereo reverb behind, and does nothing at all to a mono file — which it detects and refuses up front.

Exports frames at a timestamp, at a fixed interval, or every frame in a range as images.

Converts an animated GIF to MP4 or WebM, which is usually a fraction of the size.

Measures integrated loudness and true peak, then normalises to a target LUFS such as -14 or -16.

Reads and edits container metadata — title, artist, album art, codecs, duration and streams.

Removes hiss, hum and room noise with spectral gating: a noise profile from a marked region or the quietest tenth of the recording, subtracted per frequency band with a stated floor. Reads audio files and the audio track of MP4/WebM; writes WAV, M4A or Opus.

Detects a monophonic vocal’s pitch with YIN, snaps it to a chosen key and scale, and resynthesises with TD-PSOLA so the formants stay put. The retune-speed control runs from natural correction to the hard-tuned staircase; the detected pitch track is drawn.

Strips the audio track from a video without touching the video stream.

Finds quiet stretches in a recording and removes them, with a threshold and padding setting.

Speeds up or slows down a video from 0.25× to 4×. The picture is retimed frame-exact; the audio keeps its pitch through WSOLA time-stretching, or can follow the speed tape-style, or be muted.

Converts between SRT, VTT, ASS, SMI and plain text, fixing the encoding along the way.

Shifts or stretches subtitle timings by a fixed offset or by matching two reference points.

Changes playback speed without shifting pitch, or transposes pitch without changing length.

Concatenates clips in order, normalising resolution and frame rate when they differ.

Rotates a sideways phone recording by 90° steps, or rewrites just the rotation metadata.

Dudley’s channel vocoder: splits a voice into N bands, follows each band’s envelope and paints it onto music or a built-in synth drone. Controls for band count, attack, release, formant tilt, sibilance pass-through and dry/wet.

Renders an audio track as a waveform, spectrogram or frequency-bar image (PNG), for artwork and thumbnails.

Estimates a track’s tempo in BPM by onset autocorrelation, with a confidence score and alternative candidates.

Lists the soft subtitle tracks inside an MKV or MP4 and exports the chosen one as SRT or VTT.

Plays a clip backwards, decoding one keyframe interval at a time so memory stays bounded, with the audio reversed sample by sample or muted. Refuses, with the numbers, when a keyframe interval would not fit the memory cap.

Analyses camera motion across frames and re-crops each one to smooth out handheld shake.

Builds a grid of evenly spaced frames with timestamps, the classic preview contact sheet.

Overlays a logo image or a line of text at one of nine anchors with margin, size, opacity and an optional time window.

Draws the waveform of an audio file as an image or SVG with colour and size options.