Free online tool

Slowed and Reverb Generator

Generate a slowed and reverb version of any song in your browser. Set speed, reverb decay, bass and tone, preview live, then export MP3 with nothing uploaded.

Open the free editor: upload MP3, WAV, AAC, OGG or FLAC, adjust speed, reverb, bass, tone, width and space, then export MP3 or a ZIP of many tracks. Everything runs in your browser.

A slowed and reverb generator takes a finished song and generates a new version of it: slower, lower in pitch, and wrapped in a reverb tail. The editor at the top of this page is that generator. It runs in your browser, never uploads your file, and hands you an MP3 at the end. The rest of this page explains what "generating" actually means under the hood, so you know what each slider does to the audio and why some settings sound right while others fall apart.

If you want the short version with fewer technical details, the slowed reverb maker page covers the same workflow step by step. If you have a whole album to process, start at bulk slowed reverb. If you are new to the style itself, what is slowed and reverb covers where it came from and why it sounds the way it does.

What generating a slowed version means

There is no neural network involved and no server rendering your song. A slowed version is generated by three signal operations applied in order: a playback rate change, a convolution reverb, and a pair of shelving filters plus a stereo width stage. Each step is deterministic. Feed the generator the same file with the same settings twice and you get the same output twice. That predictability is useful when you are building a batch of edits that need to match.

Resampling versus time-stretching

There are two ways to make audio play slower, and the Preserve pitch switch picks between them.

Resampling plays the samples back at a slower rate. The track lasts longer and every frequency drops by the same ratio. A turntable does this when you lower its speed, and a tape deck does it when the motor slows down. It is the classic slowed sound. With Preserve pitch off, Speed works this way. At 0.84x the song sits about 3 semitones lower, at 0.79x about 4, at 0.75x about 5. At the bottom of the range, 0.50x, it is a full octave down.

Time-stretching changes duration without changing pitch. The audio is cut into small overlapping grains that are spaced out and crossfaded back together. With Preserve pitch on, Speed uses the browser's built-in time-stretch. The song gets slower but stays in its original key. Preserve pitch changes the live preview only; exported files keep the speed-linked pitch. Time-stretching is harder than resampling, and at extreme settings you can hear a slight warble or a softening of drum hits. It is the right tool for practice tracks and dance edits, and the wrong tool for slowed + reverb, where the lower voice is the whole point.

Convolution reverb

Reverb and Space drive a convolution reverb. Convolution runs your audio through an impulse response, which is a recording of how a space answers a short burst of sound. The output is your song as if it were played in that space. Reverb is the wet mix, 0 to 100%. It sets how loud the reverberated signal is next to the dry one. Space is the decay time, 1.2 to 6.0 seconds. It sets how long the tail hangs after each sound.

The reverb is applied after the speed change, so the tail is not stretched with the music. A 2.8 second decay stays 2.8 seconds whether the track is at 1.00x or 0.75x. This matters in practice. Slowing a song opens up the gaps between phrases, which means you can use a longer Space on a heavily slowed track without the tail smearing into the next line. On a lightly slowed track, keep Space shorter.

Shelving filters and width

Bass is a low shelf from -6 to +12 dB. Tone is a high shelf from -10 to +10 dB, where negative values darken the sound. Width scales the stereo image from mono at 0 to wider than the source at 1.5. None of these are multiband EQ. They are broad, musical shelves meant for quick correction, not for carving out a single frequency. If you need surgical EQ, do it in a DAW before you bring the file here.

Step by step

1. Drop in a file. MP3, WAV, AAC, OGG or FLAC. It is decoded in your browser and never leaves your machine.
2. Choose a starting point. Dry is 1.00x with 0% reverb and flat tone, a blank slate. Haze Chamber and Wave Drift are ready-made slowed sounds you can adjust.
3. Set Speed. This decides how low the voice goes. 0.85x is a safe first value for most vocal tracks.
4. Set Reverb, then Space. Start around 40% and 2.8 s and move one control at a time so you can hear what each does.
5. Correct the tone. Slowing pushes energy downward and dulls the top. Try +4 to +6 dB Bass with -3 to -5 dB Tone and adjust from there.
6. Play and compare against Dry. The preview runs through the exact chain used for export.
7. Download. One MP3 at 96 kbps and 32 kHz. For several tracks, select them and use Download Selected to get a ZIP.

Settings by goal

Each row is a starting point, not a rule. Dense mixes want less Reverb than the table says. Sparse mixes can take more.

What the generated file is

The export is an MP3 encoded in a web worker at 96 kbps and 32 kHz. That bitrate is chosen so a full-length song encodes in seconds inside the browser, including on phones. It is fine for YouTube, TikTok and Shorts, because every one of those platforms re-encodes your upload anyway. It is not a distribution master. If you plan to release the result on streaming services, generate the edit from a WAV source and finish it in a DAW that can export WAV. The tool does not export WAV or FLAC, and it does not export video.

A frequent question is whether the preview is a lower-quality approximation of the export. It is not. The live preview and the rendered file go through the same nodes with the same settings. What you hear while the spectrum visualizer is moving is what lands in the MP3.

Generating many versions at once

Add several files, click Select All Tracks, and every change applies to all of them. Download All renders each track in turn and bundles the MP3s into a ZIP. If some songs need different settings, select only those, adjust, then select the others. Settings stick to the tracks they were applied to.

This is the main reason the tool exists as a generator rather than a one-shot converter. A channel posting slowed versions of a full album can set one speed, one reverb amount and one tone curve, then let the whole batch render while they prepare thumbnails.

Saving your own generator settings

When a combination works, click Create preset and give it a name. It appears beside the built-in presets and is stored in your browser's local storage on that device. There is no account, so presets do not sync between devices, and clearing site data removes them. Write down the values of presets you care about if you switch computers often.

Common mistakes with generated slowed versions

• Using a low-bitrate source. Resampling stretches every artifact in a 128 kbps file. Start from a lossless or 320 kbps file when you can.
• Reverb before speed. Decide on Speed first. The tempo change alters how much room there is for the tail.
• Treating Space as the volume of the reverb. Space is time, Reverb is level. If the mix sounds washed out, lower Reverb, not Space.
• Forgetting mono playback. Width above 1.3 can thin out on phone speakers. Check before you publish.
• Going below 0.70x with vocals. Resampling that far turns most voices into gravel. Save the deep end for instrumentals.

Copyright

Generating a slowed version of a song you do not own creates a derivative work. The original copyright stays with the rights holder, and platform matching systems recognise slowed versions. On YouTube that usually means a claim and the revenue goes to the rights holder, sometimes a block. Read the copyright and Content ID guide before building a channel around other people's music.