A volume changer is a linear amplitude multiplier, and avoiding mistakes with one starts with treating its percentage slider as a multiplier rather than a perceived loudness control. Setting the slider to 100% leaves every decoded sample at its original amplitude, 50% halves each finite sample, 200% doubles it, and 0% writes complete silence while preserving the source duration, sample rate, and channel layout. Doubling a sample is a roughly 6.02 dB amplitude change, not a doubling of perceived loudness, so a slider value of 400% does not push the file anywhere near twice as loud in human terms. Once a multiplied sample crosses plus or minus full scale, the browser clamps it to the supported range, the operation counts that as a clipped sample, and the resulting waveform contains distortion that no later slider move can repair. Treating the slider as a precision tool means picking a supported browser-decodable file, setting a sensible starting percentage, and reading the clipping report before downloading the WAV. The strongest safeguard is the preview: when the full file sounds clean through suitable speakers or headphones and the clipped-sample count sits at or near zero, the export is ready to use.
What Linear Gain Actually Does to Audio
The slider controls a straight multiplier applied to every finite decoded sample, so the relationship between percentage and amplitude is fixed and predictable. At 100% each sample keeps its decoded value. At 50% every sample is multiplied by 0.5, which halves the amplitude. At 200% the multiplier becomes 2, doubling each sample, and a value of 0 produces silence while preserving the source duration, sample rate, and channel layout.
The mathematics for translating a linear multiplier into decibels follows a standard formula: amplitude change in dB equals 20 × log10(gain). Working out the most common case gives 20 × log10(2) = 20 × 0.30103, which simplifies to roughly 6.02 dB. Pushing the slider to 400% corresponds to a multiplier of 4, which by the same conversion works out to about 12.04 dB before any clipping is applied.
The Volume Changer uses this rule because the underlying stage treats the percentage as a direct linear amplitude multiplier, not a loudness compensation curve or a broadcast target. Doubling the slider value does not double perceived loudness, and moving from 100% to 400% does not push the output four times louder in human terms. Treating the slider as a precision knob rather than a "make it louder" dial is the largest single safeguard against distortion, because the moment a multiplied sample crosses plus or minus full scale the browser clamps it back into range and records a clipped sample. A deeper treatment of this relationship lives in the Volume Changer Accuracy guide, which walks through the same derivation with extra worked examples.
Three Steps to Adjust Volume Safely
These three steps cover the full safe workflow for adjusting audio without introducing distortion, decode errors, or unwanted format surprises.
- Choose a supported browser-decodable file no larger than 50 MiB and wait for the filename to appear in the upload area. MP3, WAV, M4A, AAC, Ogg, WebM, and FLAC are the formats the contract names, but a filename ending in one of those extensions does not prove the bitstream can actually be decoded. If decoding fails, the tool stops and shows an error rather than producing an empty or partial WAV.
- Set the linear output volume from 0% through 400% and select Change volume. Starting near 100% confirms the export matches the source, then raise or lower the percentage in small steps while listening and watching the clipped-sample count grow as you cross full scale. Each new percentage invalidates the previous result and revokes its temporary object URL, so an older asynchronous decode cannot overwrite a newer setting.
- Preview the WAV, inspect the peak and clipped-sample count, and download only after listening to the complete result. A nonzero clipped-sample count is evidence that the gain may have introduced audible distortion, even when the waveform still looks normal.
Each step belongs to a different phase: file selection before the gain is applied, gain selection before the multiplier runs, and verification before the export is committed. Running all three turns a single slider move into a documented edit that can be replayed later with predictable results.
File Choice Errors That Trigger Early Stops
Many volume changer mistakes start before the slider is ever touched. The 50 MiB input limit, the roughly five-minute decoded duration limit, the eight-channel cap, and the 30-million-channel-sample cap all run before any gain is applied, so picking the wrong file produces a stop signal rather than a partial output. A 30 MB compressed MP3 can easily decode into more than five minutes of stereo audio, and a long multitrack session can quietly exceed 30 million samples even when the file itself sits well below 50 MiB. The tool checks the decoded duration and channel count, not the on-disk size, so confirming both before uploading prevents the most common early stop.
Format support is handled by the browser's Web Audio decoder, which the W3C Web Audio API specification describes in the decodeAudioData stage. A filename extension is not proof that the media can be decoded, so the practical rule is to wait until the filename appears in the upload area after a successful decode. Treat a decode failure as an early stop, not a malformed file you can retry into working order, and confirm the audio really is what you think before you click Change volume.
Reading the Peak and Clipped-Sample Report
The clipping report is the most reliable single indicator of safe-versus-distorted output. Each finite decoded sample is multiplied by the selected gain, so any value above +1 or below -1 is clipped to the supported full-scale range and counted. The peak after clipping is also reported, because two clips at the boundary can show the same peak yet still cover a meaningful share of the file. A peak that sits very close to plus or minus full scale paired with a nonzero clipped-sample count is strong evidence that the gain should be lowered, while a peak well below full scale with a zero clip count usually means the gain sits in a safe zone.
A zero clip count does not by itself guarantee ideal perceived loudness, because perceived loudness depends on frequency balance, dynamics, speakers, and listening conditions, none of which a linear multiplier changes. The tool does not calculate LUFS, true peak, replay gain, or platform-specific targets. Use the clipping counter as a distortion detector rather than a loudness meter, and treat a count of zero as permission to download rather than proof that the output is mastered.
| Clipped-sample count | Peak after clipping | What it usually means |
|---|---|---|
| 0 | well below plus or minus 1 | Headroom remains; gain is unlikely to introduce distortion |
| 0 | at or near plus or minus 1 | Existing peaks already touch full scale; loudness may still be limited |
| small but nonzero | at plus or minus 1 | A few samples crossed full scale; small risk of audible distortion |
| large and nonzero | at plus or minus 1 | Substantial clipping; lower the percentage and re-render |
What the Exported WAV File Includes
Every export is a 16-bit PCM WAV written to a RIFF container with little-endian integer fields and format code 1 for uncompressed PCM. The downloaded filename ends in -volume-adjusted.wav. Negative full scale maps to -32768 and positive full scale maps to 32767, matching the asymmetric range of signed 16-bit audio. Channel samples are interleaved frame by frame, the sample rate matches the decoded source, and the channel count matches the decoded source. The container fields follow the structure documented in the WAVEFORMATEX reference, which is what most players still read.
The tool does not preserve source tags, album art, chapters, loop points, cue markers, encoder metadata, or the original compression settings, so any metadata belongs in the WAV's downstream workflow rather than in the gain operation. WAV is broadly inspectable and avoids adding a new codec dependency, but it is usually much larger than MP3, AAC, or Ogg. Plan storage accordingly, especially when the source was already a compressed format, and remember that the downloaded file is temporary and disappears when the tab closes or when a new operation replaces it.
| Field | Value after export |
|---|---|
| Container | RIFF / WAVE |
| Format code | 1 (uncompressed PCM) |
| Bits per sample | 16 |
| Sample rate | matches decoded source |
| Channel count | matches decoded source |
| Sample layout | interleaved, frame by frame |
| Filename suffix | -volume-adjusted.wav |
Habits That Keep Your Output Safe
A few consistent habits close most of the remaining gaps. Keep the original recording as the master and re-export from it rather than re-running the gain stage on a previously adjusted WAV, because repeated PCM conversions add quantization noise that no slider change can hide. Verify the downloaded duration and channels before replacing any production audio, because the operation cannot restore information already clipped in the source and the file's intended use often depends on those exact values.
For a safe edit, begin near 100%, preview the full output through suitable headphones or speakers, then check the clipped-sample report. Reduce the percentage if any distortion appears, and remember that the downloaded WAV is temporary: it disappears when the tab closes or when a new operation replaces it, so save a working copy somewhere durable as soon as the preview confirms the result. Linear gain is not loudness normalization, mastering, limiting, compression, or broadcast compliance, so for any task that depends on perceived loudness, use a measured audio workflow and monitor the exported file on the same speakers your audience will use.