Original: abrupt edges
The tone starts and stops at a nonzero sample level. Listen to its entrance and ending; no fade is applied.
Download original WAVFree tools for your next audio edit
Audio fade in and out. Give your edges room.
Bring a short recording in from silence or let its ending settle. Choose separate fade-in and fade-out lengths, choose WAV or MP3, compare the original with the exported file, and download it. Your audio stays in this browser.
No account neededNo audio uploadWorks in your browser
Troubleshoot clicks after cutting audio Fade in / fade out
Free · local0 disables an edge. Otherwise use 0.001–10 seconds. Default: 0.25 seconds at each edge. Both fades together must fit within the decoded recording.
Audio stays in this browser. The same linear amplitude envelope changes every channel and every sound at the selected edges. No trimming, crossfade, noise removal, or normalization is applied. Download the result before leaving and keep your original.
Hear the difference before choosing a length
Same tone. Different entrances and endings.
These original synthetic test tones demonstrate this tool's gain envelope. They are not recorded voices, music performances, or AI separation results. Keep the player volume consistent and start at a comfortable listening level.
Short edges: 5 ms each
A short fade changes only 240 frames at each edge. The middle keeps the original decoded level.
Download 5 ms fade WAVLonger: 0.5 s in, 1 s out
The tone has a slower entrance and ending. Much more of its duration receives a reduced gain, even though its highest sample peak is unchanged.
Download longer fade WAVEach file is a three-second, 330 Hz mono tone at 48 kHz, stored as 16-bit PCM WAV (288,044 bytes). All three retain 144,000 frames and a sample peak of 0.25, about −12.04 dBFS. The fades change the edges, not the peak in the unaffected middle. Equal peaks do not mean equal perceived loudness.
Download the original and choose it in the editor above to try 0.005 seconds at each edge, or 0.5 seconds in and 1 second out. Prepare from the original each time. Applying another fade to an already faded file compounds the gain change.
The samples are original Saturalabs learning assets, dedicated to the public domain under CC0 1.0. See the generation method and reuse notes. For level decisions, read the peak versus loudness guide.
One recording, two independent edges
How to add a fade in or fade out
Keep your original. Work on a copy that leaves enough room around the first and last wanted sounds.
- 01 / CHOOSE
Open your MP3 or WAV
Choose a mono or stereo recording up to 50 MiB and five minutes. Listen to the original player first so you know where a word, note, or effect starts and ends.
- 02 / SET
Choose each fade length
Use seconds, with a default of 0.25 at each edge. Set 0 to leave an edge unchanged, or 0.001–10 seconds for an enabled fade. Both fades together must fit inside the decoded file without overlap.
- 03 / COMPARE
Prepare, preview, and download
Apply the fades, then choose WAV or MP3, then alternate between the original and faded output. Check the first consonant, initial attack, and last tail before downloading. The preview plays the same file you download. Choose WAV when precise timeline placement matters.
Linear amplitude, linked channels
A fade changes the level at the edge.
A fade in begins at zero gain and rises to the decoded recording's original level. A fade out does the reverse at the end. The Audacity fades reference distinguishes these basic linear fades from shaped fades and crossfades.
Here, the multiplier changes linearly across the stored frames of each fade. Near its midpoint it is about 0.5, approximately −6.02 dB relative to the same decoded samples. A straight amplitude ramp is not a promise of a straight change in perceived loudness.
Every sound at that time receives the same gain. Fading in a noisy opening lowers the voice and the noise together. Both stereo channels receive that multiplier, preserving their relative balance rather than shifting the sound left or right.
The tool retains the full decoded frame count. It does not delete leading silence, shorten an ending, or add a pause. If the file needs a different time range, make that working copy with the free audio trimmer before choosing its fades.
Three-second clip at 48 kHz
What two quarter-second fades change
| Region | Decoded frame indices | Gain |
|---|---|---|
| Fade in | 0–11,999 | 0 at the first frame, rising to 1 at frame 11,999 |
| Middle | 12,000–131,999 | 1 throughout |
| Fade out | 132,000–143,999 | 1 at frame 132,000, falling to 0 at the last frame |
Choose the edit before the fade
Use the edge that suits the recording
Keep the beginning of a word
A long fade starting on speech can hide a soft opening consonant. Include a little room before the wanted word when possible, and compare the full phrase. A waveform boundary alone does not decide whether the phrase sounds complete.
Keep a deliberate musical attack
A fade can soften the first drum hit or picked note. Set the fade in to 0 when the original entrance should remain intact, then choose only the fade out. Listen in the arrangement rather than judging the edge in isolation.
Leave enough room for the tail
A fading end changes the decay you hear. Keep enough original duration for a sustained note, breath, or effect tail before choosing the fade. This tool cannot extend a tail that has already been cut off.
Fading, joining, and crossfading
Different transitions need different edits.
This tool changes the opening and ending of one file. The audio joiner puts several files one after another, with optional pauses. A crossfade overlaps clips while one becomes quieter and the other louder; use a multitrack editor for that transition.
If you fade a joined sequence here, only its outer opening and ending change. The internal clip boundaries do not receive fades. To prepare separate edges, work on each original clip first, download those copies, then arrange them in the joiner. Check whether the two faded edges create an unwanted dip or gap.
The trimmer's optional 5 ms edge fade is a short fixed treatment. This utility gives each outer edge its own length. Neither option guarantees click removal or fixes a discontinuity elsewhere in a recording. For a click inside a file, inspect that location in an editor rather than fading the entire file's beginning or end.
Know the exported file
Choose WAV or MP3, with the original kept.
The browser decodes a supported MP3 or WAV at 48 kHz and applies the fades before PCM16 preparation. Choose a 16-bit PCM WAV or a new MP3 at 128, 192, or 320 kbps. Mono stays mono; stereo stays stereo. Decoding can resample or include codec padding, so the reported duration follows the decoded frames. The WAV retains those frames. MP3 applies an additional lossy encode and can add delay or padding; its decoded sample values can differ throughout the file, not only at the edges.
Integer export rounds sample values and clips values outside full scale. Source tags, artwork, original bit depth, and other metadata are not copied. A WAV can be larger than an input MP3. Both input and output must fit within 50 MiB; use the size calculator to plan a shorter working copy.
For a WAV export, the WAV inspector can confirm the downloaded format and size. If your final project requires 44.1 kHz, the sample rate converter can make that delivery. Download each result before opening another tool and keep the originals.
Fades do not set a peak target or match loudness between clips. The peak and loudness normalization guide explains that separate decision. Source isolation is another workflow again: it estimates a described source inside the mix rather than lowering every sound at an edge.
MP3 encoder credits
Built with the LAME encoder via lamejs.
The optional MP3 download uses lamejs 1.2.1 by Alex Zhukov, a JavaScript port of the LAME MP3 encoder. Its unmodified script is served separately. See the attribution and source notes, upstream license notice, LGPL-3.0 and GPL-3.0 texts, and complete upstream source package.
A few useful answers
Free audio fade in and out questions
How do I fade audio in and out online?
Choose a supported mono or stereo MP3/WAV up to 50 MiB and five minutes. Set each fade length in seconds and prepare the WAV. Preview the faded result, compare it with the original, and download the same file. The default is 0.25 seconds at each edge.
Can I apply only a fade in or only a fade out?
Yes. Set an edge to 0 to disable it. An enabled fade can run from 0.001 to 10 seconds. The two fade lengths must fit within the decoded recording without overlapping. When both are 0, no edge gain is applied, but the tool still decodes and exports a new WAV.
Is it free, and is my audio uploaded?
Fading, previewing, and downloading WAV or MP3 are free, with no account or credits. The file is read in this browser, without an audio upload. Download the result before leaving or reloading. The separate Saturalabs AI isolation service has its own account and download-credit flow.
Does the tool fade each stereo channel separately?
The same linear amplitude multiplier is applied to both channels at every frame. This retains their relative balance during the fade. A mono input remains mono and a stereo input remains stereo; no panning or channel mixing is applied.
Can I crossfade two songs or remove background noise?
No. This tool changes the edges of one complete recording. It does not overlap files, mix music under speech, normalize loudness, or remove a competing source. A fade lowers every sound at that edge together. Use a multitrack editor for crossfades and layered audio.
Can I download the result as MP3?
Yes. Choose MP3 and 128, 192, or 320 kbps. The fades are applied before a new lossy encode, locally in a worker. WAV remains available for 48 kHz, 16-bit PCM and more predictable timeline placement. MP3 can add delay or padding. Neither output copies tags or artwork, and a higher bitrate does not recover lost detail. Both formats must fit within 50 MiB.
Does fading shorten the recording or guarantee click removal?
The fade envelope retains the complete decoded frame count. Only the selected edge regions receive a gain change before PCM16 preparation. WAV retains that frame count; MP3 adds a lossy encode and can include delay or padding. A fade can soften an abrupt edge, but cannot guarantee a suitable cut or fix a click inside the file. Listen for clipped consonants, missing attacks, and faded-out tails.
After preparing your audio