What Is WAV? Everything You Need to Know

The Short Answer

WAV (Waveform Audio File Format) is the file format Microsoft and IBM jointly published in August 1991 as part of Windows 3.1's Multimedia Extensions. By default it stores uncompressed linear PCM audio — the raw, sample-by-sample digital representation of sound that comes off an analog-to-digital converter. The format is structurally trivial, decodes with zero CPU overhead, supports any bit depth and sample rate the hardware can handle, and is the lingua franca of professional audio: every DAW, every video editor, every sample library, and every broadcast workflow on earth understands WAV.

The trade-off is size. A three-minute stereo song at CD quality is roughly 30 MB. A one-hour 24-bit/96 kHz session capture is about 1.2 GB. WAV is what you record and edit in; you convert out of it before you distribute.

The RIFF Container

Underneath, a WAV file is a RIFF (Resource Interchange File Format) document. RIFF is Microsoft's tagged-chunk container, originally derived from Electronic Arts' IFF specification, and the WAV layout follows a strict pattern:

• RIFF header — 12 bytes at the very start: the four ASCII letters 'RIFF', a 32-bit little-endian length field, and the form type 'WAVE'.
• fmt chunk — describes the audio: format code (1 = PCM, 3 = IEEE float, 6 = A-law, 7 = µ-law, 65534 = WAVE_FORMAT_EXTENSIBLE), channel count, sample rate, byte rate, block alignment, and bit depth.
• data chunk — the audio samples themselves, stored interleaved (left, right, left, right…) for stereo files. The chunk header gives the byte length of the audio payload.

Optional chunks can appear before, between, or after the required ones: 'LIST' / 'INFO' for metadata, 'bext' for the Broadcast Wave timecode/originator info, 'cue ' for markers, 'smpl' for sampler loop points, 'fact' for non-PCM formats. A reader is supposed to walk the chunk list and skip anything it does not recognize, which is what makes RIFF mostly forward-compatible.

The container is technically able to hold compressed data — anything with a valid Microsoft format code, including ADPCM and even MP3 inside a 'WAVEFORMATEX' header — but in practice the moment you see a '.wav' file, you should expect uncompressed PCM. Compressed WAV exists mostly as a curiosity in legacy Windows software.

Why DAWs and Editors Love WAV

There are three real reasons WAV dominates production environments:

1. No decode overhead. PCM samples are just numbers in a file. A modern CPU can stream dozens of WAV tracks off an SSD without breaking a sweat. Compressed formats need decoding before each playback head can read them, which adds latency and complexity to a tightly timed real-time engine. 2. Sample-accurate seeking. You can jump to any sample in a WAV file by computing 'byte_offset = data_chunk_start + sample_index × bytes_per_sample × channels' and reading from there. Compressed formats have to decode from the previous keyframe — fine for playback, painful for waveform drawing and tight edit-point scrubbing. 3. Predictable file structure. Every DAW developer has written WAV reading code, and every WAV reader behaves the same way. There are no encoder-specific quirks, no LAME-vs-Fraunhofer differences, no padding ambiguities. What you wrote is what gets read back.

This is also why sample libraries, foley packages, and orchestral instrument collections ship as WAV. The streaming engines inside Kontakt, EastWest Play, Spitfire's BBC Symphony Orchestra, and others assume direct PCM.

File Size Math

WAV size is deterministic. The formula is:

bytes ≈ sample_rate × bit_depth/8 × channels × seconds + ~44 byte header

Plugging in standard values:

• CD quality (44.1 kHz, 16-bit, stereo): 44,100 × 2 × 2 = 176,400 bytes/second, or about 10.1 MB per minute.
• 48 kHz / 24-bit stereo (broadcast standard): 48,000 × 3 × 2 = 288,000 B/s ≈ 17.3 MB per minute.
• 96 kHz / 24-bit stereo (high-resolution): 576,000 B/s ≈ 34.6 MB per minute.
• 192 kHz / 32-bit float stereo (mastering): 1,536,000 B/s ≈ 92 MB per minute.

Because the data chunk uses a 32-bit unsigned length field, classic WAV maxes out at 4 GB. That is roughly 6.7 hours of CD-quality stereo or 1.7 hours at 96 kHz / 24-bit / 5.1. The RF64 extension (and its EBU-blessed cousin BW64) lifts this ceiling by allowing a 64-bit data size, and most modern DAWs auto-promote to RF64 when a recording crosses 4 GB.

Bit Depth and Sample Rate

Bit depth controls the dynamic range and noise floor; sample rate controls the audible frequency ceiling.

• 8-bit integer — 256 quantization levels, ~48 dB dynamic range. Historical only (early Windows wave sounds).
• 16-bit integer — 96 dB dynamic range. Audio CD, distribution masters, Bluetooth A2DP.
• 24-bit integer — 144 dB dynamic range. Tracking and mixing standard.
• 32-bit integer — 192 dB dynamic range. Mostly a curiosity.
• 32-bit float — IEEE 754 single-precision. The internal format of every modern DAW, also written to disk by 32-bit-float field recorders (Zoom F-series, Sound Devices MixPre II) so clipping in capture is recoverable in post.
• 64-bit float — Used internally by some mastering chains; rarely written to disk.

Sample rates run from 8 kHz (telephony) through 11.025 / 22.05 / 32 / 44.1 / 48 / 88.2 / 96 / 176.4 / 192 / 384 kHz. The Nyquist rule means a 44.1 kHz file can represent frequencies up to 22.05 kHz, comfortably above the human hearing limit. Higher rates exist mostly to give downstream filters and pitch-shifters more headroom, not because anyone can hear above 22 kHz.

Channels and Channel Order

WAV supports mono, stereo, and multichannel layouts. For surround content you can either store all channels in one WAV (ordered FL, FR, FC, LFE, BL, BR for 5.1, with the order specified in WAVE_FORMAT_EXTENSIBLE's channel mask) or split into individual mono "stems." DAW projects almost always use the mono-per-track approach because it makes routing and editing cleaner.

Broadcast Wave (BWF)

The European Broadcasting Union extended WAV in 1997 with a 'bext' chunk that adds timecode-aware metadata: an originator string, a creation date, an SMPTE time-of-day timestamp, and a free-text description. This BWF profile is mandatory for film and television deliverables — every Pro Tools recording on a film set is BWF, and broadcast servers like AVID iNEWS expect BWF on ingest. The audio inside is still ordinary PCM; BWF is purely additive metadata.

The Metadata Weakness

WAV's biggest practical flaw is metadata. The original spec had no clear story for tagging, and three competing approaches grew up around it:

• LIST/INFO chunks — Microsoft's original answer. Limited fields (IART, INAM, ICMT, IGNR, ICRD…), no album art, ASCII-only in older readers.
• ID3v2 tags — bolted on by encoders like LAME via an 'id3 ' chunk. Many players read these, but the placement is non-standard and some readers strip them.
• bext chunk — broadcast metadata only.

The result: metadata in WAV is fragile, and tag editors disagree about what to write where. This is one of the strongest reasons to use FLAC for archival instead — FLAC's Vorbis Comments are a clean Unicode key/value store with first-class artwork support.

When to Use WAV

WAV makes sense when:

• You are tracking, editing, or mixing in a DAW.
• You are sending masters to a mastering engineer or pressing plant.
• You are delivering broadcast content (use BWF specifically).
• You are building a sample library.
• You are interoperating with a tool that has poor compressed-format support.

WAV is the wrong choice when:

• You are distributing to listeners — the files are 5–10x larger than necessary. Use the WAV to MP3 converter for general distribution or WAV to FLAC for lossless distribution.
• You are uploading to a streaming service. Most ingestion pipelines re-encode, but you will burn far more upload bandwidth and storage than needed — compress a WAV file first if you must keep PCM and just need to trim its footprint.
• You are putting audio in a mobile app. Bundle compressed audio and decode at runtime.
• You are archiving a music collection long-term — see WAV vs FLAC for archiving and What is FLAC for why FLAC wins on storage cost without losing anything.

For one-off conversion jobs the WAV converter handles WAV-to-anything in your browser, and MP3 to WAV and FLAC to WAV round-trip the other way for editing.

Real-World Use

The places WAV still dominates in 2026:

• Recording studios — every Pro Tools, Logic Pro, Cubase, and Reaper session is WAV-backed.
• Film and TV — production sound is BWF, post-production deliveries are BWF, archive masters are BWF.
• Sample libraries — Native Instruments, Spitfire, Output, Splice all ship WAV.
• Broadcast servers — radio playout systems like RCS Zetta and ENCO DAD ingest WAV/BWF.
• Mastering houses — final masters delivered to streaming services, vinyl cutters, and DDP authors are WAV at the project's working sample rate.

Outside those contexts WAV is overkill. Inside them it has not been seriously challenged in 35 years and is unlikely to be — the format is simple enough that the only thing that could replace it is a similarly simple format with better metadata, and FLAC already plays that role for everyone who needs compression.

For the broader story on uncompressed digital audio see the PCM audio explainer.