To add reverb and echo to audio online, you can apply one of six bounded effects — echo, lightweight reverb, reverse, fade in, fade out, or normalization — locally in your browser with Audio Effects Online, choose a whole-number amount from 0 to 100 percent, and download a fresh PCM16 WAV without uploading anything to a server. The tool is designed for quick sound-design drafts rather than full studio production, so each run applies exactly one transformation to every decoded sample frame. Input caps include a 50 MiB file size, a 30 million channel-sample budget for the decoded data, and the same ceiling for the complete post-effect output including any echo or reverb tail. Because every step happens in the current browser tab through the W3C Web Audio decode contract, you keep your master file untouched, can preview the result on the page, and decide whether to re-run the same WAV through a different effect to build a multi-stage chain.
Six Audio Effects in One Local Tool
Audio Effects Online applies one transformation per run to a single browser-decodable audio file and exports the result as a PCM16 WAV. The list of choices is wider than a typical reverb-only utility: echo, lightweight reverb, reverse, fade in, fade out, and normalization. Each option is bounded, reproducible, and disclosed on the page, which makes the tool practical for sound-design drafts, podcast polishing, and quick comparison renders without installing software or signing in to a cloud service.
Normalization here is peak normalization, not loudness matching: the largest absolute decoded sample is scaled toward a 0.95 peak, so the perceived loudness between two different recordings is not balanced automatically and silent input stays silent. Reverse reads every decoded frame in the opposite order and blends with the original according to the amount, so at 0% the audio plays unchanged and at 100% it plays fully reversed. Fade in scales from a selected dry proportion at the start to full level at the end, while fade out performs the inverse across the full duration.
Echo and reverb are the two spatial effects in the set. The echo is a single 200 millisecond delay with a half-strength copy added back into the dry signal at 100% amount; lower amounts scale only the delayed copy, so the original sample is always retained verbatim. Reverb is a deliberate three-tap early-reflection design with disclosed delays and weights, not a hall simulation. The table below summarizes what each of the six options actually does, which is the simplest way to decide which one fits a given source.
| Effect | What it does at 100% amount | Best used for |
|---|---|---|
| Echo | Adds a half-strength copy 200 ms after each input sample; full tail allocated before processing | Slap-back vocals, dub-style repeats, post-production accents |
| Lightweight reverb | Adds three early reflections at 29.7, 37.1, and 41.1 ms with weights 0.50, 0.35, and 0.25 | Sketching room sense, thickening dry voice or guitar takes |
| Reverse | Reads every decoded frame in opposite order and replaces the dry signal at full strength | Hidden-message intros, transition FX, ear-training material |
| Fade in | Ramps from selected dry level at the start to full level at the end | Smooth openings, podcast cold opens, cinematic risers |
| Fade out | Ramps from full level at the start to selected dry level at the end | Natural endings, podcast sign-offs, looping cue beds |
| Normalization | Scales the largest absolute sample toward a 0.95 peak | Recovering headroom after a quieter-than-intended render |
Reverb vs Echo: Which Effect Fits Your Goal
People who search to add reverb and echo online usually want either a sense of space around a recording or a clear, repeated delay line. The two effects in this tool sound very different even though both add something after the original signal. Echo creates a single delayed copy you can count out by ear, while reverb adds short, partly overlapping reflections that smear into a wash. Choosing the right one up front saves you from running the same file twice and stitching exports by hand.
Pick echo when you want a single, recognizable repeat at a fixed distance — a slap-back vocal, an old-style phone effect, or a rhythm-tightened snare. The 200 ms delay is short enough to feel like a tight double and long enough to read as a discrete repeat. Pick lightweight reverb when you want the source to sit slightly forward or back in a synthetic acoustic without committing to a real room; the three fixed taps give shape but never model diffusion, damping, or a decay tail. If a true hall, plate, spring, or convolution reverb is the goal, this tool will not substitute for a dedicated room or impulse-response processor.
For combined treatments the workflow is sequential: run echo first, download the resulting WAV, and run reverb on that file as the next input. Each run applies only one effect, so chains are built by exporting the intermediate and re-using the local WAV. Changing the input file, effect, or amount at any point invalidates prior work, so keep the original master untouched and process copies only.
How to Add Reverb and Echo to Audio Online
The page is built around three explicit actions: pick a file, pick an effect and amount, then apply and download. The same path works for echo, reverb, reverse, fade in, fade out, and normalization — only the chosen effect and its integer percentage change between runs.
- Open Audio Effects Online in your current browser tab and confirm the file you intend to process is browser-decodable and no larger than 50 MiB. MP3, WAV, M4A, AAC, Ogg, WebM, and FLAC filenames are accepted, but actual decodability depends on your browser and operating system.
- Choose one audio file using the file input. The browser decodes it locally through the Web Audio implementation; no upload is sent to a remote server.
- Select exactly one effect — echo, lightweight reverb, reverse, fade in, fade out, or normalization — and set a whole-number amount from 0 through 100. Fractions are not accepted.
- Apply the effect locally. The page reports channel count, sample rate, decoded and output frame counts, output peak, and the exact WAV byte size after processing.
- Inspect the reported values, then download the PCM16 WAV. The source file stays untouched in your downloads folder; only the new WAV is created.
- To chain a second effect, use the freshly downloaded WAV as the input for another run with a different effect and amount.
If decoding fails or the post-effect output would exceed the 30 million channel-sample ceiling, the tool creates no file at all and surfaces the relevant frame or peak numbers on the page. That worst case shows up most often with very long inputs where the reverb or echo tail pushes the result past the budget; trimming the source first is the standard workaround.
Choosing an Amount from 0 to 100 Percent
Every effect on this tool shares the same amount control, but each effect responds to that control in its own way. For echo, the dry signal is always preserved and the delayed copy is scaled linearly with the amount, so 50% produces a half-strength 200 ms repeat and 25% a quarter-strength one. For reverb, the three tap weights — 0.50, 0.35, and 0.25 — are each multiplied by the selected amount, so 50% reduces the whole wet mix by half and the early reflections keep their relative balance against each other.
Reverse, fade in, and fade out all use the amount to define either the blend ratio with the original or the starting level of the ramp. At 0% reverse returns the file unchanged; at 100% you get a fully reversed render. Fade in at 0% plays the file at full level from the start, while fade in at 100% ramps from silence to full level across the whole duration. Normalization targets a 0.95 peak, and silent input stays silent.
A practical habit is to start at a low amount, between 10 and 25%, for echo and reverb, listen to whether the effect is audibly present, then increase in 10-point steps until you hit either the tone you want or audible clipping in the reported peak. Strong settings on echo and reverb can drive overlapping samples past the legal -1 to +1 range, and the encoder clips them to that ceiling before quantization rather than scaling the whole file down.
File Limits, Output Specs, and Local Processing
The platform-level guardrails exist for a reason, and they interact. The first ceiling is the 50 MiB file size on the encoded input. The second is the 30 million channel-sample budget that applies independently to the decoded input and to the post-effect output. Echo and reverb can pass the input budget while still failing the output budget, because their tails add frames beyond the original duration; in that case nothing is truncated and no file is produced.
The download is a brand-new interleaved little-endian PCM16 RIFF/WAVE. Microsoft documents the WAV block alignment as channels multiplied by bits per sample divided by eight, which at sixteen bits means two bytes per mono frame and four bytes per stereo frame. Floating samples are clipped to the legal -1 through +1 range before they are quantized to signed sixteen-bit PCM. The W3C Web Audio specification defines how an AudioBuffer length is measured in frames, and that same contract is what the page uses to size the WAV header.
| Limit | Declared value | Behavior when exceeded |
|---|---|---|
| Encoded input file size | 50 MiB | File rejected before decoding |
| Decoded channel samples | 30,000,000 | Decode refused, no output produced |
| Post-effect channel samples | 30,000,000 | No truncation applied; no file produced |
| Channels | 1 to 8, equal length | Decode refused |
| Sample rate | 8,000 to 192,000 Hz | Decode refused |
| Effect amount | Integer 0 to 100 | Field rejected until corrected |
Export drops the source codec, bitrate, compression quality, tags, artwork, chapters, cue points, loop markers, and every other piece of container metadata, so the resulting WAV is not a re-mux of the original. All file reading, decoding, sample processing, WAV encoding, and download creation happen in the browser tab, audio never reaches a server, and there is no account or login. Decode contexts are closed between jobs and temporary object URLs are revoked on replacement so stale results cannot replace newer ones. Repeated lossy decode and PCM processing may reveal source artifacts, and repeated clipping or normalization can change dynamics, so keeping an untouched master copy is the safest habit before running any chain.