Audio File Size Calculator
Budget an audio file from timecode. Compare uncompressed PCM sample data with an entered compressed average bitrate, add an explicit container and metadata allowance, and see decimal and binary storage without treating codec quality as a file extension.
Set the session and encoding lanes
One hour of 48,000 Hz, 24-bit, two-channel PCM contains 1.037 GB of raw samples; a 1.50% allowance raises the estimate to 1.052 GB. The 192 kbps compressed lane estimates 87.696 MB with the same allowance.
PCM and compressed size equations
Pulse-code modulation stores a numeric sample for every channel at every sample instant. Ideal PCM payload therefore grows linearly with duration, sample rate, bit depth, and channel count. Compressed size is instead estimated from an average encoded bit rate. The sample rate and bit depth used during production do not directly multiply the final lossy file size once average codec bitrate is known.
compressed bits = seconds × average kilobits/second × 1,000
estimated bytes = payload bits ÷ 8 × (1 + overhead percent ÷ 100)
decimal MB = bytes ÷ 1,000,000; binary MiB = bytes ÷ 1,048,576
The overhead input represents container structures, headers, tags, artwork, padding, indexes, error data, and other non-audio bytes as one planning percentage. Real overhead is not necessarily proportional to duration, especially for short files with large artwork.
Sample rate, bit depth, and channels
Sample rate states how many samples are represented each second for each channel. Bit depth states the nominal number of stored bits in each PCM sample. Channels multiply the sample stream: mono is one, stereo is two, and multichannel projects can be much larger.
A 48 kHz, 24-bit stereo stream carries 48,000 × 24 × 2 = 2,304,000 raw bits each second, or 2.304 Mbps. Changing to 96 kHz doubles payload; changing from stereo to six channels triples it; changing 24-bit to 16-bit multiplies it by two-thirds.
Some containers pad samples to larger word boundaries or use floating-point data. A nominal 24-bit signal may occupy 24 or 32 stored bits per sample depending on format and software. Enter stored bit depth or use an explicit allowance.
Average bitrate and variable bitrate
For constant-bitrate audio, entered kbps is a direct payload rate. Variable-bitrate codecs allocate more bits to complex passages and fewer to simpler material, so only the file’s average bitrate predicts size. A target or nominal setting may not equal the finished average.
IETF Opus specifications permit a broad bitrate range and dynamic changes. RFC guidance discusses different useful ranges for speech and music, but bitrate alone does not define audible quality. Encoder version, mode, bandwidth, channels, complexity, source, and listening conditions matter.
Measure a representative encoded file when storage must be guaranteed. Divide its payload or total bits by duration to obtain an observed average and keep overhead separate if comparing codecs.
Worked one-hour studio example
One hour contains 3,600 seconds. At 48,000 samples per second, 24 stored bits per sample, and two channels, PCM payload is 8,294,400,000 bits or 1,036,800,000 bytes. Adding 1.5% produces approximately 1.052 decimal GB, equal to about 0.980 GiB.
At a compressed average of 192 kbps, one hour contains 691,200,000 encoded bits or 86.400 MB before allowance. With 1.5% overhead, estimate 87.696 MB or 83.633 MiB. Raw PCM payload is 12 times the compressed payload in this example.
Ten selected PCM files require about 10.524 GB. A 64 GB decimal storage budget holds 60 whole one-hour PCM files under the same assumptions and represents approximately 60.82 hours before the next full file would exceed the budget. File-system allocation, reserved capacity, redundancy, and backups reduce practical availability.
What can change the finished file
| Factor | Size effect | Planning action |
|---|---|---|
| Container header | Usually a fixed plus per-chunk or per-packet amount. | Measure representative short and long files instead of assuming one percentage. |
| Tags and artwork | A cover image can outweigh audio in a short clip. | Set artwork dimensions and metadata policy. |
| Variable bitrate | Average changes with content and encoder choices. | Encode a diverse test set and retain a high-percentile size. |
| Lossless compression | Ratio depends on waveform predictability. | Do not enter a lossy bitrate as though it guarantees FLAC or another lossless size. |
| Multichannel mapping | Codec coupling and stream mapping may not scale exactly by channel count. | Use codec/container documentation and actual output. |
| File-system allocation | Small files consume whole allocation units and storage may reserve space. | Test the destination volume and include operational headroom. |
File size is not audio quality
A larger PCM file can preserve a higher sample format, but it does not repair clipping, noise, poor microphone placement, bad room acoustics, resampling artifacts, or an unsuitable master. A higher compressed bitrate can reduce coding artifacts, yet codec design and source material matter. Do not rank two formats by megabytes alone.
For preservation, distribution, streaming, telephony, podcasts, music, machine learning, and legal evidence, the appropriate format and metadata differ. Preserve an authoritative master, document processing, verify checksums, and create delivery derivatives under a controlled workflow.
Storage, transfer, and backup planning
Nominal drive capacity is usually decimal, while operating systems may display binary units with ambiguous labels. This calculator reports both. A 64 GB device does not expose all 64 billion bytes to one project after formatting, file-system metadata, wear provisioning, recovery partitions, snapshots, or organizational reserves.
Batch size multiplies one selected estimate by whole-file count. Real projects also contain session files, proxy audio, peaks, render caches, alternate mixes, stems, video, notes, and duplicates. Build an inventory from the production workflow rather than budgeting only final masters.
Transfer time depends on usable throughput, not storage size alone. Maintain at least the backup copies and geographic separation required by the project’s risk. Verify every copy with checksums and restoration tests; a file that occupies the expected bytes can still be corrupted or undecodable.
How to verify an encoded result
After export, inspect duration, channel layout, sample format, codec, average bitrate, container, and exact byte size with a trusted media tool. Play the beginning and end, seek through the file, and check for truncation. For variable bitrate, divide total encoded bits by duration rather than reading only a configured target.
Separate audio payload from container overhead only when the tool can report both. Otherwise use total finished bytes for operational storage. Compare several files, including silence, dense music, speech, and noise, because encoders respond differently.
Document software version, codec settings, sample conversion, loudness processing, metadata, and artwork. A reproducible export preset is more useful than a size estimate with no production context.
Archive manifest and retention record
For every delivery, preserve a manifest containing filename, exact bytes, duration, format, codec, channel layout, sample rate, stored sample format or average bitrate, checksum, creation tool, and creation date. Define which file is the preservation master and which files are replaceable derivatives. A retention policy should also state how long each copy remains, where independent copies live, who can delete them, and how migrations will be validated when storage media or codecs change.
Recalculate capacity from actual manifests after the first production cycle. Forecasts improve when average size, high-percentile size, daily file count, revision count, and annual growth are measured separately. Keep free-space alarms and a migration reserve; filling a working volume to its nominal last byte can harm performance and leave no room for indexes, temporary renders, or recovery.
Audio file size FAQs
Why multiply PCM by channel count?
Each channel has its own sample value at every sample instant. Stereo therefore stores twice the ideal sample data of mono at the same sample rate and bit depth.
Is kbps the same as kB/s?
No. kbps is kilobits per second. Divide by eight to obtain decimal kilobytes per second before adding container overhead.
Why are GB and GiB different?
GB is one billion bytes. GiB is 1,073,741,824 bytes. Reporting both prevents a unit-label difference from looking like missing data.
Can sample rate predict MP3 or Opus size?
Not by itself. For compressed output, average encoded bitrate is the direct size input. Sample rate, channels, encoder mode, and content influence which bitrate and quality are appropriate.
How accurate is the overhead percentage?
It is a user-entered planning allowance. Headers, tags, artwork, packetization, indexes, padding, and file duration make real overhead format-specific and sometimes partly fixed.
Does lossless audio have a fixed bitrate?
No. Lossless compression ratio depends on the signal. Use PCM as an upper planning reference or measure representative lossless encodes with the intended codec and settings.
References
These standards and U.S. metrology sources support audio bitrate, Opus container behavior, and digital unit definitions.