Checking the result after using a volume changer comes down to three concrete checks: listen to the complete preview from start to finish, read the peak value and clipped-sample count the tool reports, and confirm the downloaded WAV opens at the original duration and channel count. Volume Changer performs decoding, sample multiplication, peak detection, clipping counting, preview playback, and WAV encoding entirely in the current browser tab, so the numbers you see on screen describe the exact bytes that will be written to the downloaded file. The percentage you selected is a direct linear amplitude multiplier, not a decibel control, which means doubling the slider from 100% to 200% raises each finite sample by roughly 6.02 dB before the encoder clamps anything that crosses digital full scale. Treat the on-screen peak and clipped-sample count as the inspection product of the run, not a guess, and use the steps below to verify what came out of Volume Changer before you replace any production file.
What the Result Page Actually Shows
When Volume Changer finishes processing, it stops at a result state rather than handing you a file blindly. Two figures are the core of that state, and they describe the multiplied-and-clipped audio buffer, not the original source.
- Peak after clipping — the largest absolute sample value in the result, expressed in the same floating-point range the browser decoder produced. A peak of 1.0 means the loudest sample sits exactly at digital full scale. A peak under 1.0 means the loudest moment is quieter than maximum. A peak reported above 1.0 is impossible, because any sample above the limit was already clamped to ±1.0 during the run.
- Clipped-sample count — the number of finite decoded samples that crossed digital full scale and had to be clamped to -1.0 or +1.0. Each finite sample is multiplied by the selected gain, and any value above +1 or below -1 is clipped to the supported full-scale range before PCM encoding. Zero clipped samples means the gain stayed inside the legal output range. Any nonzero count is evidence the gain may have introduced audible distortion, although zero clipping does not guarantee ideal perceived loudness.
These two numbers come from the same buffer that the WAV encoder writes out, so they are the most authoritative check you have without leaving the page. Treat them as a snapshot of the exact file you are about to download.
How to Check Your Volume Changer Result Step by Step
Use the three steps below in order. The order matters: the file name has to land first, the percentage has to be applied, and only then does the preview, peak, and clipped-sample report describe something real.
- Confirm the source file is decoded. Choose a browser-decodable MP3, WAV, M4A, AAC, Ogg, WebM, or FLAC file no larger than 50 MiB, then wait for the file name to appear in the tool. A filename extension is not proof the media can decode, so the appearance of the file name is the signal that decodeAudioData succeeded. Decoded duration is also limited to five minutes, channel count to eight, and total decoded channel samples to 30 million; if decoding fails, the page shows an error rather than producing an empty or partial output.
- Apply a percentage and run Change volume. Set the linear output volume between 0% and 400%, then select Change volume. At 100% samples keep their decoded amplitude; at 50% every sample is multiplied by 0.5; at 200% every sample is multiplied by 2. Zero percent writes silence while preserving the decoded duration, sample rate, and channel count. Changing the file or percentage after a run invalidates the previous result and revokes its temporary object URL.
- Preview, read the report, then download. Play the preview from start to finish and listen for distortion, sibilance, or pumping. Read the peak and clipped-sample count. Download only after listening to the complete result, because the temporary object URL disappears when the tab closes or when a new operation replaces it.
Reading the Peak and Clipped-Sample Report
Read the two numbers together rather than separately. A run that ends at a peak of 0.94 with zero clipped samples is healthy and indicates the gain stayed under digital full scale with headroom to spare. A run that ends at a peak of 1.0 with a clipped-sample count of 12,480 is loud but distorted, because those 12,480 samples were flattened against the ceiling. A run that ends at a peak of 1.0 with zero clipped samples is the loudest clean result possible inside the supported range.
| Selected percentage | Multiplier on each sample | What the report typically shows on a typical source |
|---|---|---|
| 0% | 0.0 (silence, original duration preserved) | Peak near 0.0, zero clipped samples |
| 50% | 0.5 | Peak roughly half the source peak, zero clipped samples unless the source already exceeds ±1 |
| 100% | 1.0 (unchanged) | Peak matches the source, zero clipped samples unless the source itself was clipped |
| 200% | 2.0 | Peak usually reaches 1.0 with clipped samples unless the source was unusually quiet |
| 400% | 4.0 | Peak at 1.0 with a large clipped-sample count on most sources |
The exact peak and count depend on the actual source audio, so read the values on your screen rather than guessing from the table. If the count is nonzero but the preview sounds acceptable, the distortion is below your perceptual threshold for that material. If the count is zero but the preview sounds harsh, perceived loudness depends on frequency balance, duration, dynamics, speakers, and listening conditions, and this tool does not calculate LUFS, true peak, replay gain, or platform-specific targets.
Confirming the Downloaded WAV Is Correct
Once you click download, the file ends in -volume-adjusted.wav and contains the same interleaved little-endian PCM16 buffer the preview played. Confirm four properties before you treat the download as the new master:
- Duration matches the original source. Volume Changer preserves the decoded duration because the operation is a per-sample multiplier, not a time stretch.
- Channel count matches the original source, capped at eight channels at decode time.
- Sample rate matches the original source, because the tool writes the WAV header with the decoded sample rate rather than resampling.
- Bit depth is always 16-bit signed PCM in a RIFF container with format code 1 for uncompressed PCM, where negative full scale maps to -32768 and positive full scale maps to 32767, matching the asymmetric range of signed 16-bit audio. The output is always WAV even when the source is compressed, which is broadly inspectable but usually much larger than MP3, AAC, or Ogg.
The tool does not preserve source tags, album art, chapters, loop points, cue markers, encoder metadata, or the original compression settings, so do not rely on the WAV to carry ID3 or chapter information. For a deeper look at how linear gain shapes that buffer, the linear gain explanation walks through the math behind the multiplier and the asymmetric signed range of PCM16.
When to Repeat the Volume Change
Repeat the run when any of the following is true. First, when the clipped-sample count is nonzero and the preview shows audible distortion, drop the percentage until the count returns to zero, the peak stays inside ±1.0, and the preview sounds clean. Second, when the peak is well under 1.0 and the source is too quiet for its intended use, raise the percentage in small steps, checking the report each time, until the peak sits near 1.0 with zero clipping. Third, when you change the file or percentage, the previous result is invalidated and its temporary object URL is revoked, so you have to run Change volume again before a new download is valid. Fourth, when the source itself was already clipped, repeated PCM conversions cannot restore the lost information, so keep the original file and re-export from it whenever possible. Linear gain is not loudness normalization, mastering, limiting, compression, noise reduction, or broadcast compliance, so for professional delivery use a measured audio workflow and monitor the exported file through suitable headphones or speakers before you replace any production audio.
What the WAV Format Looks Like in the Download
The exported WAVE file uses a RIFF container, little-endian integer fields, format code 1 for uncompressed PCM, the original decoded sample rate, the original number of channels, and 16 bits per sample. Channel samples are interleaved frame by frame, so a stereo file writes left, right, left, right through the entire buffer. This deterministic layout is why the same source and same percentage always produce the same byte count on disk, and why the peak and clipped-sample count from the report can be trusted as a description of the file you save. The 50 MiB input cap, the five-minute decoded duration cap, the eight-channel cap, and the 30 million channel-sample cap exist to bound memory allocation and main-thread work on ordinary laptops and mobile devices, so a compressed file can still exceed limits after decoding even when the selected file is small.
For a deeper look, see Plan the Steps to Use an Audio Pitch Changer.