Free Sample Rate Converter

Sample rate is how many audio snapshots are captured per second, measured in Hz. 44.1 kHz is the CD standard, 48 kHz is the video/broadcast standard, and 96/192 kHz are used for heavy studio processing or archival capture.

Answer a few questions about your project — use case, whether it syncs to video, heavy DSP or pitch-shifting, CPU/storage constraints — and the tool applies rule-based audio-engineering guidance to recommend a rate, with reasons and acceptable alternatives.

WAV, FLAC, OGG, and AIFF. Files are resampled using soxr, a high-quality bandlimited resampler, at a quality preset from QQ (fastest) to VHQ (highest quality, recommended for final masters).

48 kHz for anything that might touch video, broadcast, or a game engine — that whole world is built around it, and matching avoids resampling and sync headaches. 44.1 kHz is the CD and streaming delivery rate, so it is equally valid when you know the project is audio-only. For general music production with no constraints the tool recommends 48 kHz, because your release gets resampled to 44.1 kHz at export either way.

YouTube and effectively every video platform sit in the 48 kHz family. Deliver at 48 kHz and your audio passes through without an extra resampling step. If your session is at 44.1 kHz, convert before export rather than letting the upload pipeline do it for you.

Not for playback — human hearing tops out around 20 kHz, and 48 kHz already covers that with margin under the Nyquist theorem. 96 kHz helps during production: nonlinear plugins like saturation and distortion generate harmonics that can alias near Nyquist, and extreme pitch-shifting or time-stretching has more samples to work with. It also doubles your file sizes and CPU load, which is why the advisor only recommends it when you tell it those conditions apply.

A good resampler loses very little, and this one uses soxr's bandlimited algorithm — the same family SoX uses. Downsampling does discard content above the new Nyquist limit, which is unavoidable and usually inaudible. What you should avoid is resampling repeatedly back and forth; convert once, from your highest-quality source, at the VHQ preset for final masters.

Almost always a sample rate mismatch. If your audio is 44.1 kHz and the video timeline runs at 48 kHz, something has to resample, and sloppy handling shows up as gradual drift over a long clip. Convert the audio to 48 kHz first so it stays sample-locked to picture.

Half the sample rate — the highest frequency a given rate can represent. At 44.1 kHz that is 22.05 kHz, comfortably above the roughly 20 kHz ceiling of human hearing, which is exactly why 44.1 kHz was chosen for CD. Anything above the Nyquist limit has to be filtered out before sampling or it folds back as aliasing.

48 kHz. Most interfaces run natively at 48 kHz, and it matches video delivery if you ever cut clips for YouTube or social. Voice content gains nothing from higher rates. 44.1 kHz is fine if you are certain you only ever publish to audio-only podcast platforms.

48 kHz. Unity, Unreal, Wwise, and FMOD all standardize on it, so matching avoids per-asset resampling at import. Use 44.1 kHz only if your specific engine or pipeline is explicitly built around it.

96 kHz for one-shot preservation work you cannot redo. The headroom above 20 kHz protects against future needs — restoration, pitch correction, delivery formats that do not exist yet. 192 kHz only if the source or gear genuinely warrants it, such as ultrasonic research. If storage is a hard limit, 48 kHz is still safely above the audible range.

It is rule-based, not machine learning, because this is settled audio-engineering practice rather than something to learn from data. Video sync takes priority over everything else; then archival and heavy-DSP conditions push the rate up; then CPU or storage constraints pull it back down. Every recommendation lists its reasons and the alternatives that are also defensible.

Not directly — this tool reads WAV, FLAC, OGG, and AIFF. Resampling a lossy file is rarely what you want anyway, since it re-encodes already-discarded audio. Convert your MP3 to WAV first if you need to change its rate.

No. Bit depth (16-bit, 24-bit) is preserved from your source file — only the sample rate changes. Bit depth controls dynamic range and noise floor; sample rate controls frequency range. They are independent settings.

They are soxr's presets, from QQ (quick) through LQ, MQ, HQ, and VHQ. HQ is the default and is transparent for almost everything. Use VHQ for final masters and archival conversions, where the extra processing time does not matter.

Yes. It is completely free with no signup, no login, and no limits — get a recommendation and convert files instantly.