WAV

WAV, short for Waveform Audio File Format, is a file format that stores digital audio inside a RIFF container, almost always as uncompressed linear PCM samples. In audio production, broadcast and post-production, WAV is the most widely used interchange format for full-quality audio, because practically every editor, recorder and operating system reads it without conversion.

WAV is a container, not a codec. Its header declares how the audio data is encoded, and while that encoding is nearly always plain PCM, the format can also carry IEEE floating-point samples or, more rarely, compressed data. Its main costs are large file sizes, limited metadata and a 4 GB size limit in the classic version.

How a WAV file is structured

A WAV file is built from chunks, each consisting of a four-character identifier, a 32-bit size field and the chunk data. This structure comes from the Resource Interchange File Format (RIFF) that Microsoft and IBM introduced in the early 1990s. All numeric values are stored in little-endian byte order.

The file begins with a RIFF header that names the form type WAVE. Two chunks are mandatory: the fmt chunk, which describes the encoding, and the data chunk, which holds the samples. Software that does not recognize an optional chunk can skip it using its size field, which keeps the format extensible.

The table below lists the chunks most often found in practice.

In the data chunk, samples of all channels are interleaved frame by frame: left, right, left, right for stereo. A canonical PCM file has a 44-byte header, although real files often contain additional chunks before the audio data.

Encodings, bit depths and file sizes

The format tag in the fmt chunk tells a reader how to interpret the samples. Tag 1 means integer PCM, tag 3 means 32-bit or 64-bit floating point, and the tag WAVE_FORMAT_EXTENSIBLE adds a channel mask for multichannel layouts and an exact definition of valid bits per sample.

Because PCM WAV is uncompressed, its size follows a simple formula: sample rate × bit depth × channels × duration. Stereo audio at 44.1 kHz and 16 bits runs at 1,411.2 kbit/s, which equals about 10.6 MB per minute. The following table shows how quickly size grows with resolution and channel count.

WAV compared with other audio formats

WAV differs from its neighbors in the formats and codecs family mainly in whether the audio is compressed and how well the format carries metadata. The comparison below places it among the formats it is most often weighed against.

FLAC and WAV contain bit-identical audio after decoding, so the choice between them is about workflow rather than sound. WAV is preferred where editing speed and tool support matter; FLAC is preferred for archiving and distribution where storage and tags matter.

Broadcast Wave and large-file extensions

Professional workflows rely on extended forms of WAV that add metadata and remove the size limit. The classic RIFF header uses 32-bit size fields, so a standard WAV file cannot exceed 4 GB, which is reached after roughly 4.1 hours of 48 kHz, 24-bit stereo audio, or much sooner in multichannel recordings.

• Broadcast Wave Format (BWF), defined by the EBU, adds the bext chunk with a description, originator and a sample-accurate time reference used to align audio with picture.
• RF64 replaces the 32-bit size fields with 64-bit values, allowing files far larger than 4 GB while remaining compatible with BWF metadata.
• BW64, standardized by the ITU, builds on RF64 and supports object-based and immersive audio metadata.
• Wave64 (W64) is an older 64-bit variant from audio software vendors and is less widely supported today.

Field recorders used in film production typically write BWF files with `iXML` metadata, so the editing system can sort takes by scene, track name and timecode automatically.

Limitations and common pitfalls

The main weaknesses of WAV are storage cost and inconsistent metadata support. Most practical problems arise when files move between systems that interpret optional chunks or extended headers differently.

• Weak music tagging: INFO tags are inconsistently supported, so cover art and detailed tags often disappear in music libraries.
• Silent truncation at 4 GB: long multichannel recordings in classic WAV can lose audio beyond the limit unless RF64 is enabled.
• Assuming WAV means lossless: a WAV file made from an MP3 still carries the lossy artifacts of its source.
• Float files in older software: some players misread 32-bit float WAV and play noise or silence.
• Mismatched sample rate on import: a 44.1 kHz WAV placed in a 48 kHz session plays at the wrong speed and pitch without conversion.

In plain words

A WAV file is like a full, unedited copy of a recording. It stores the sound exactly as the recorder captured it, without squeezing it down to save space. That makes it large, but nothing is thrown away.

Because almost every computer, phone and audio program can open it, WAV is the safe choice when people need to pass recordings to each other without losing quality. When the music finally reaches listeners, it is usually converted into a smaller format for streaming.

A real-world example

A documentary sound recordist captures an interview on a portable recorder set to 48 kHz, 24-bit Broadcast Wave, with eight tracks for two lavaliers, a boom microphone and backup channels. Each file carries iXML metadata with the scene, take and track names, plus a timecode stamp in the bext chunk.

The recorder is configured to write RF64 because a single long interview on eight tracks exceeds 4 GB in just over one hour. Back in the edit suite, the picture editor imports the files, and the software lines each take up with the camera footage automatically by timecode.

The mixer later exports the final stereo program as a 48 kHz, 24-bit WAV for the broadcaster, while the streaming version is encoded to AAC from that master. The WAV remains the archival reference for every future deliverable.