LEARN · Mixing vs. Mastering

What’s the Difference Between Mixing and Mastering?

Mixing balances the individual tracks of a song — vocals, drums, bass, instruments — in level, tone and space, and combines them into a stereo mix. Mastering then processes that finished mix as a whole: tonal balance, loudness and peak level are set so the song sounds good on every system and is ready for the streaming platforms. Mixing shapes the song; mastering makes it release-ready.

What happens during mixing?

The mix decides how a song sounds: which part sits up front, how vocals and instruments share the frequency range, where each element sits in the stereo image and how much reverb and effects it carries. Typical tools are level, panning, EQ, compression, reverb, delay and automation. The mix works on many individual tracks at once — which is why it can solve problems that mastering can no longer reach, such as a vocal that’s too quiet or a kick drum clashing with the bass.

What happens during mastering?

Mastering only works on the finished stereo mix. It fine-tunes the tonal balance, sets loudness and dynamics, and limits the peak level so no distortion occurs when the file is converted to streaming formats. For EPs and albums, it also makes sure all tracks fit together in loudness and tone. The end result is the file you upload to your distributor — the master from which every other format is derived.

Loudness and true peak: the numbers that matter for streaming

Streaming services normalize the loudness of songs, measured in LUFS. An extremely loud master gets turned down and loses dynamics without sounding any louder. So rather than chasing the highest possible number, aim for a master that sounds good after normalization. For peak level, Spotify recommends keeping true peak below −1 dBTP so no clipping occurs during encoding; very loud masters benefit from even more headroom.

Can mastering save a bad mix?

Only to a limited extent. Mastering can improve the overall sound, but it can no longer target individual tracks. If the vocal is too quiet in the mix, mastering only makes it louder along with everything else. So the rule is: fix problems in the mix before the song goes to mastering. Also leave your mix some headroom — a mix that’s already slammed against the limiter leaves mastering hardly any room to work.

AI mastering: when it fits

AI mastering analyzes your mix and matches it to a reference, such as a genre profile or a track whose sound you like. It’s fast and affordable, and for many releases it delivers a solid, streaming-ready result. For highly idiosyncratic productions or albums with wide dynamics, an experienced mastering engineer still has the edge. AI mastering at OGRECORDS™ lets you compare the original and the master in an A/B test before anything is applied.

Frequently asked questions

Which comes first: mixing or mastering?

Always mixing first. Mastering builds on the finished stereo mix and is the last sonic step before release.

Can I mix and master my own music?

Yes, many independent artists do both themselves. Mixing and mastering with some distance and on different playback systems — headphones, monitors, phone — makes a big difference to the result.

How loud should my master be for Spotify?

There’s no required value. By default, Spotify normalizes to around −14 LUFS; a master in that range will barely be changed. Many genres are mastered louder — what matters is that the track sounds good after normalization.

What are stems?

Stems are submixes of a mix, such as all drums, all vocals or all synths as separate files. They’re used for stem mastering, remixes, Atmos mixes and sync licenses.

What file format does my master need for distribution?

For streaming releases, an uncompressed WAV (or FLAC) at a minimum of 16-bit and 44.1 kHz is the standard; many distributors also accept higher resolutions. Never upload an MP3 as your master: any further encoding by the stores would degrade the quality even more.

Make your mix streaming-ready.

Upload your track, pick a reference and compare the original and the master in an A/B test before you apply anything.

Go to AI mastering →Release now →