How it works
WAV bytes = seconds × sample rate × (bit depth ÷ 8) × channels.
Compressed formats are estimated as a fraction of WAV — MP3 ≈ 8%, OGG ≈ 7%, Opus ≈ 5%.
Use the estimate to budget your final build size before encoding.
Free indie game dev audio size estimator. Enter minutes of voice / music, target format, sample rate, and bit depth; outputs total uncompressed and compressed MB / GB, with storage warnings for mobile deployments.
Built and maintained by Paul Clark, Redmoon Software
Use during pre-production to budget your final build size. Indie mobile games and itch.io uploads have hard size limits.
Audacity and other DAWs report file size after encoding. This tool estimates before.
WAV bytes = seconds × sample rate × (bit depth ÷ 8) × channels.
Compressed formats are estimated as a fraction of WAV — MP3 ≈ 8%, OGG ≈ 7%, Opus ≈ 5%.
Use the estimate to budget your final build size before encoding.
For looping music and ambience. OGG handles seamless loops better than MP3 at similar bitrates.
Acceptable for retro / SFX, but speech and music sound noticeably duller. 44.1 kHz is the safe default.
Multiply sample rate by bit depth by number of channels by seconds, then divide by 8 to get bytes. For example, stereo 44.1 kHz 16-bit audio is about 10.1 MB per minute.
Not usually. Music benefits from a full 44.1 kHz sample rate to preserve high frequencies, while voice often sounds fine at 22 kHz mono, which roughly halves the file size.
Input
30 minutes, OGG, 44.1 kHz, 16-bit, stereo.
Output
~22 MB total (~0.7 MB/minute).
WAV would be 30 × 60 × 44100 × 2 × 2 = ~317 MB. OGG ratio ~7% → ~22 MB.
File size for uncompressed PCM is sample rate times bit depth times channels times duration. At the CD-standard defaults — 44,100 Hz, 16-bit, stereo — that is about 10.6 MB per minute, so 30 minutes is well over 300 MB before any compression.
Every one of those four terms is a straight multiplier. Dropping to mono halves it; 22,050 Hz halves it again. That is why sound effects are frequently mono at reduced rates while music stays stereo.
Different assets want different handling. Short, frequently triggered sound effects are usually kept uncompressed or lightly compressed, because decoding cost is paid on every play and latency matters. Long music tracks are compressed and streamed, because size dominates and the decode is amortised.
Sample rate should follow content: a rate above 44.1 kHz is inaudible for most game audio and doubles storage for nothing, while dialogue and many effects are perfectly acceptable at lower rates. Choosing per asset rather than setting one project-wide standard is where most of the saving is.
Optimal grid + VRAM footprint for 2D animations.
Game dev & streamingGame Boy / NES / Genesis storage gauge.
Game dev & streamingWebcam + overlay sizing for 1080p / 1440p / 4K.