To cut audio in a web browser, select a continuous time range between a non-negative start point and an end point up to the decoded file duration, then export the frame-aligned segment as a 16-bit PCM WAV file. Modern web browsers perform this operation locally by reading the sound file into system memory, decoding encoded streams into floating-point audio buffers, and extracting the exact sequence of audio frames requested. This workflow eliminates the need to transmit private media to remote servers while delivering frame-accurate cuts for speech, sound effects, or musical samples.
How Local Browser Audio Cutting Operates
Cutting audio entirely inside a web browser relies on client-side decoding frameworks rather than server-based rendering pipelines. When you select a file, the browser reads its raw byte stream locally and passes it to the Web Audio engine. The system relies on standard APIs like BaseAudioContext.decodeAudioData to parse compressed containers and decompress audio data into an uncompressed buffer stored in memory.
Once decompressed, the sound data exists as an AudioBuffer object containing discrete channels of 32-bit floating-point pulse-code modulation (PCM) samples. Choosing start and end timestamps converts those temporal values directly into integer frame indices based on the active sample rate. Because the processing occurs within the active browser session, audio files are not uploaded when you cut audio online using local web tools. The original file remains untouched on your disk, while temporary object URLs manage the source player and generated clip until you replace the file or close the tab.
Steps to Trim Audio and Export a WAV Segment
You can trim audio clips locally using the web tool Audio Cutter. The application processes your file inside your web browser and produces an uncompressed WAV download upon completion.
- Choose a browser-decodable audio file within the displayed file and decoded-audio limits. Select a local sound file from your device. The input file size must be 25 MB or smaller. The tool inspects the container format and decodes the stream into system memory, displaying the total duration, sample rate, channel count, and total frame count.
- Enter a non-negative start time and a later end time that does not exceed the decoded duration. Input your desired cut boundaries in seconds. The start boundary must be zero or greater, and the end boundary must be strictly greater than the start time without surpassing the exact decoded duration of the source file.
- Select Cut and export WAV, preview the PCM16 result, then download it. The browser extracts the designated frame range, converts floating-point samples into 16-bit signed PCM data, and generates a RIFF/WAVE download link. You can play the output directly in your browser to verify boundary accuracy before saving the file locally.
Audio Limits and Output File Size Calculations
Local browser editing requires explicit safety thresholds to prevent browser tabs from running out of heap memory during processing. Large input files or high-sample-rate multichannel recordings expand significantly once fully uncompressed into RAM. The trimming process enforces strict input and output boundaries: source files cannot exceed 25 MB, and decoded audio buffers must stay within 15 minutes of playback, 8 audio channels, 192 kHz sample rate, and a cumulative cap of 30 million samples across all combined channels.
Because the exported clip uses uncompressed 16-bit PCM encoding, output file sizes are strictly proportional to selection length, channel count, and sample rate rather than the compressed file size of the input. Each sample frame contains 2 bytes per channel. The exact file size in bytes is determined by the following formula:
Size (bytes) = (Selected Frames × Channel Count × 2) + 44
Consider a practical calculation for a 10-second selection taken from a stereo audio track sampled at 44,100 Hz:
- Selected Frames: 10 seconds × 44,100 Hz = 441,000 frames
- Substitution: (441,000 frames × 2 channels × 2 bytes/sample) + 44 header bytes
- Total Calculation: 441,000 × 2 × 2 = 1,764,000 bytes for audio data, plus 44 header bytes = 1,764,044 bytes (~1.76 MB)
The table below highlights how common input file types interact with browser decoding and local PCM WAV export capabilities:
| Container / Format | Decoder Support | Metadata Preserved | Export Format |
|---|---|---|---|
| MP3 (.mp3) | Universal across modern browsers | No (ID3 tags removed) | Uncompressed PCM16 WAV |
| AAC / M4A (.m4a, .aac) | Supported by most OS/browser pairs | No (iTunes/MP4 metadata removed) | |
| FLAC (.flac) | Supported in modern Web Audio engines | No (Vorbis comments removed) | Uncompressed PCM16 WAV |
| WAV (.wav) | Native browser support | No (Extra RIFF chunks discarded) | Uncompressed PCM16 WAV |
| Ogg Vorbis / Opus (.ogg) | Supported in Chrome, Firefox, Edge | No (Ogg comments removed) | Uncompressed PCM16 WAV |
Technical Differences Between Compressed Inputs and PCM WAV Exports
A primary consideration when trimming audio online is understanding why output files differ from input source files. Many popular consumer formats, such as MP3 or AAC, rely on lossy perceptual compression to reduce file sizes by stripping frequency detail imperceptible to human ears. Other formats, like FLAC, use lossless compression algorithms to reduce bandwidth while maintaining absolute sample fidelity. However, Web Audio decoders process all audio formats into raw, uncompressed floating-point buffers prior to manipulation.
When you export your cut audio clip, the editor builds a standard 44-byte RIFF WAVE header followed by interleaved, signed little-endian 16-bit integer PCM samples. A common question among audio editors is does cutting audio keep the original MP3 or AAC encoding? It does not. The export engine generates a clean PCM structure rather than re-encoding back into lossy formats like MP3 or AAC. Re-encoding lossy files in client-side applications introduces generation loss and variable codec delays, whereas standard PCM WAV guarantees immediate compatibility with editing tools, digital audio workstations, and playback systems.
During this conversion, existing container features—such as cover artwork, track titles, artist tags, loudness normalization flags, and embedded cue points—are stripped away. Floating-point audio values between -1.0 and +1.0 are scaled directly to signed 16-bit integer ranges between -32768 and +32767. If non-finite values occur within the decoded memory stream, they are converted to digital silence to prevent click artifacts. Consequently, trimming an MP3 file often results in a WAV output file larger than the original compressed input, but the resulting audio file is free from additional compression artifacts.
Sample-Frame Accuracy and Boundary Rounding
Precision timing in digital audio depends on sample rates rather than arbitrary clock seconds. In a file sampled at 48,000 Hz, each second consists of exactly 48,000 discrete frame positions. When you enter fractional seconds into the start and end input fields, the system rounds your input values to the nearest integer sample-frame boundary using the sample rate of the decoded AudioBuffer.
The slice boundary employs an inclusive start frame and an exclusive end frame. For example, selecting a start time of 0.0 seconds begins the slice at frame index 0. Selecting an end time of 2.0 seconds at a 44,100 Hz sample rate places the exclusive end boundary at frame index 88,200, resulting in a total selection of exactly 88,200 sample frames. If a user sets start and end values so close together that the calculated range rounds to less than one full sample frame, the system rejects the cut to avoid generating an empty audio file.
Because Web Audio decodes input streams into the hardware context's sample rate, subtle resampling can occur during the decode stage. The output sample rate displayed by the editor reflects the precise buffer rate used for extraction, guaranteeing that exported output durations match the exact sample count divided by the active sample rate.