Audio File Size Calculator: WAV, MP3 and Storage Time

How audio file size is calculated

Audio files come in two families, and each family has its own arithmetic. Uncompressed audio, which is what sits inside a WAV, an AIFF, or a Broadcast WAVE file, stores every sample at full size with no compression at all, so its size is exact and predictable. Compressed audio, which is what an MP3 or an AAC file is, discards information on purpose and stores the result at a fixed bitrate, so its size is predictable from that bitrate instead.

The Library of Congress format description for WAVE sets out the RIFF container and the LPCM encoding inside it, and notes that broadcast work usually uses the Broadcast WAVE variants that the EBU Tech 3285 specification defines. For lossy files, RFC 3003 registers the audio/mpeg media type and lists the MP1, MP2, and MP3 layers that produce the familiar .mp3 file, while ITU-R BS.1770 covers the loudness measurement that broadcast delivery depends on.

Two formulas cover everything on this page. For uncompressed PCM, multiply sample rate by bit depth by channel count by seconds, then divide by 8. For a compressed file at a known bitrate, multiply the bitrate in kbps by 1000 by the seconds, then divide by 8. There is no hidden factor, no variable quality setting, and no surprise.

Uncompressed PCM: sample rate, bit depth, and channels

Three settings decide how many bytes a second of recording takes, and they multiply together. The sample rate counts how many measurements of the sound are taken each second, so 48,000 samples per second is the standard for video work and 44,100 is the standard for CD music. The bit depth says how finely each measurement is stored, and 16 bits is the CD standard while 24 bits is the broadcast and studio standard. The channel count says how many of those measurements are made at once, one for mono, two for stereo, and six for a 5.1 surround file with five full channels plus a low frequency effects channel.

The AES standard for measurement of digital audio equipment, AES17, is the reference for how the full scale level of a digital file is defined, which is the part that matters when you compare a 16-bit recording to a 24-bit one. In plain terms, moving from 16-bit to 24-bit makes the file 1.5 times larger, and moving from stereo to 5.1 surround makes it 3 times larger. Nothing else changes it.

Compressed files: bitrate and duration do all the work

A compressed file has no sample rate to speak of, because the encoder threw most of the samples away. What it has is a bitrate, and that is enough. A 320 kbps MP3 stores 320,000 bits every second no matter what the music sounds like, so a five minute track is 320,000 times 300 seconds divided by 8, which is 12,000,000 bytes, and a one hour interview at the same bitrate is 144.00 MB.

That fixed cost is why compression earns its place in delivery and why it is a poor fit for a working master. The detail that lossless formats keep is discarded, and once it is gone no amount of bitrate brings it back. The RFC 3003 description of the MPEG audio layers makes the same point about the tradeoff, since the point of the layers is to remove data that the ear will not miss.

Worked example: one hour of CD quality stereo WAV

Take the format that fits on a CD: 44.1 kHz, 16-bit, stereo, and record for one hour. The calculator returns 635,040,000 bytes, which is 635.04 MB in decimal megabytes, 605.62 MiB in binary mebibytes, 0.6350 GB and 0.5914 GiB. The effective bitrate reads 1411.2 kbps, and the data rate is 10.584 MB per minute.

Run the arithmetic to check it. 44,100 samples per second times 16 bits times 2 channels is 1,411,200 bits per second, and dividing by 8 gives 176,400 bytes per second. Times 3,600 seconds for the hour is 635,040,000 bytes, which matches the tool exactly.

The storage line is the practical part. On a 64 GB card the tool reports 100 hours 46 minutes of room, and it also prints the compressed comparison for the same hour: 128 kbps = 57.60 MB, 192 kbps = 86.40 MB, 256 kbps = 115.20 MB, and 320 kbps = 144.00 MB. A 64 GB card holds one of those MP3 hours about eleven times over, while the WAV version fills the card in under seven hours.

Worked example: 5.1 surround at 48 kHz and 24 bit

Now the heaviest case a location recordist meets: 48 kHz, 24-bit, 5.1 surround, recorded for 2 hours 30 minutes. The calculator returns 7,776,000,000 bytes, which is 7776.00 MB, 7415.77 MiB, 7.7760 GB and 7.2420 GiB. The effective bitrate reads 6912.0 kbps and the data rate is 51.840 MB per minute.

Check it the same way. 48,000 samples per second times 24 bits times 6 channels gives 6,912,000 bits per second, and dividing by 8 gives 864,000 bytes per second. Times 9,000 seconds for 2 hours 30 minutes is 7,776,000,000 bytes, which is what the tool reports.

The storage line reads 82 hours 18 minutes on a 256 GB drive, and that figure is the one to carry on a shoot. A day of 5.1 location sound fills a 256 GB card in a few hours, and the same day recorded as stereo 48 kHz 24-bit uses 288,000 bytes a second, so it takes three times the card space. The EBU Tech 3285 Broadcast WAVE format exists for exactly this kind of professional work, and it carries the time reference and metadata alongside the audio.

Reading MB, MiB, GB, and GiB without getting confused

Two numbering systems run side by side, and the gap between them is what makes a card look smaller than its label. The decimal system counts in thousands, so a megabyte is 1,000,000 bytes and a gigabyte is 1,000,000,000 bytes. That is the system card makers, camera makers, and recorder menus use. The binary system counts in 1024, so a mebibyte is 1,048,576 bytes and a gibibyte is 1,073,741,824 bytes, and that is the system Windows, macOS, and most desktop file managers use.

The two drift apart as the numbers get larger, because each binary step is 2.4 percent bigger than the decimal step it replaces. A 64 GB card is 64,000,000,000 bytes, which is 59.60 GiB, so the system reports about 7 percent less than the box claims. That is not a fault and not a missing space. It is the same number written two ways.

How long you can record on an SD card or a drive

Recording time is the number people come for, and it falls out of the same division. Take the capacity in gigabytes, multiply it by a billion to get bytes, and divide by the byte rate for your format. For 48 kHz 24-bit stereo that rate is 288,000 bytes per second, so a 32 GB card holds 32,000,000,000 divided by 288,000, which is 111,111 seconds, or 30 hours 51 minutes. The same card holds 44.1 kHz 16-bit stereo for 50 hours 23 minutes, and a 320 kbps MP3 for 222 hours 13 minutes.

The pattern holds across capacities. Doubling the card doubles the hours, so a 256 GB drive at 48 kHz 24-bit stereo gives 246 hours 54 minutes, and a 1 TB drive, which is 1,000 GB, gives 964 hours 30 minutes. The line to watch is the 5.1 column in the table below, since surround is where a short card turns into an early stop.

Choosing settings for the job

Match the format to the delivery. For a podcast or a voice memo, 64 kbps to 128 kbps mono is enough and keeps the file small enough to email. For music, 192 kbps and above keeps the difference between compressed and original hard to hear, and 320 kbps is the setting most people settle on without being able to explain why. For video, 256 kbps AAC is the common pairing with a 48 kHz picture rate.

For anything you will edit later, record uncompressed. A 48 kHz 24-bit stereo master runs about 1036.80 MB an hour, and 5.1 surround runs 3110.40 MB an hour, so budget the card before the shoot rather than after. A 32-bit float file at 48 kHz stores the same 32 bits per sample as a 32-bit integer file, so the file size matches even though float keeps headroom that integer clips.

One last habit worth keeping. Check the byte rate your recorder menu shows against the figure here before you record anything you cannot redo. A recorder that says 1411 kbps for stereo 44.1 kHz 16-bit is working in uncompressed PCM, and one that says 320 kbps is writing a compressed file that no later step can widen back out.