Cut, join, level and extract audio without uploading a single file
Everything here runs in your browser. Files are never uploaded.Most audio jobs are small ones. A recording that needs its first thirty seconds gone. Two files that should be one. A voice note nobody can hear. A video whose sound is the only part worth keeping. Each of those has its own tool here, and all of them decode and re-encode on your own device, so an interview, a lecture or a client call is never uploaded to anyone.
Audio Cutter takes one or more pieces out of a file and saves them separately or joined back together, which is how you drop the silence at the start and the goodbye at the end in a single pass. Merge Audio puts several files end to end in the order you set, for stitching a recording that stopped and restarted.
Volume Changer raises or lowers a file by an exact number of decibels and counts how many samples would clip before it does anything, so you find out that a recording cannot be made louder without distortion instead of discovering it afterwards.
Video to MP3 keeps the audio track exactly as it was recorded wherever the format allows, rather than decoding and re-encoding it. That matters: a re-encode of an already compressed track loses quality for nothing.
Tone Generator plays a sine, square, triangle or sawtooth at any frequency, or sweeps between two. It is the signal source for the test pages, and a way to find a hum in a room.
MP3 and AAC are lossy: every time a file is decoded and written again, a little detail goes, and it never comes back. Do the work in one pass rather than four, and if you have the original as WAV or FLAC, edit that and export once. Cutting at an exact timestamp always means a re-encode, because a compressed file can only be split cleanly on a frame boundary.
No. The file is decoded by your own browser and written back there. Nothing is sent to a server, so an interview or a recorded call stays on your device.
MP3, WAV, M4A, AAC, OGG, FLAC and the audio inside common video files. Your browser does the decoding, so the exact list depends a little on the browser you use.
A little, yes. A compressed file can only be split on a frame boundary, so cutting at an exact second means decoding and re-encoding that section. One pass is barely audible; four passes are not.
Until it clips. Volume Changer counts the samples that would exceed full scale before applying the change, so you see the damage in advance rather than hearing it afterwards.
The limit is your device memory, because the whole file is held while it is worked on. An hour of speech is comfortable on a phone; a three-hour recording wants a computer.
Not in these tools. The Speaker and Mic Test and the Sound Level Meter use your microphone for measurement, but nothing on the site keeps a recording.