Audacity merge tracks: mixing down without losing quality

Select the tracks you want to combine by clicking in the first track's panel and shift-clicking the rest, then choose Tracks, Mix, Mix and Render. The selected tracks become a single track.

Mix and Render to New Track does the same thing but keeps the originals and adds the mix as an extra track. This is almost always the better choice, because it is reversible: if the mix is wrong you delete it rather than undoing your way backwards.

Rendering also applies every volume and pan setting and every effect permanently. That is the point of it, and it is also why you want a copy of the unrendered version somewhere.

When you mix tracks, their levels add together. Two tracks each peaking near the top produce a mix that goes past the top, and past the top is distortion.

Audacity does not warn you clearly about this. The render succeeds, the waveform looks flat-topped, and the result sounds harsh. Turn on View, Show Clipping before you render, and clipped regions appear in red.

The fix is to lower each track's gain slider before mixing. Bringing every track down by a few decibels leaves room for them to add together, and you set the final level afterwards with Loudness Normalization.

As a rough guide, mixing two tracks needs about 6 dB of headroom, four tracks about 12 dB. Better to land quiet and raise it than to land loud and have nowhere to go.

Everything you want to do differently to each track has to happen before the merge. Noise reduction on the noisy microphone, levelling on the quiet speaker, EQ on the boomy one. After rendering, they are one signal and any change applies to all of it.

Set the relative balance with the gain sliders and listen to the whole thing through before you commit. The relationship between two voices is far easier to judge in context than in isolation.

For music under a voice, the music generally wants to sit much lower than feels right while you are concentrating on it. Set it, then listen to a minute while doing something else, and it will usually still be too loud.

Exporting a multi-track Audacity project already mixes everything down into the exported file. If your only goal is one MP3 at the end, File and Export does the merge for you and leaves the project intact.

So merge only when you have a reason: applying one effect across a combined signal, reducing a heavy project to something that plays back without stuttering, or simplifying before a long edit.

Performance is a real reason on older machines. Twelve tracks with effects on each can stutter during playback, and rendering the settled ones frees up processing for the parts still being worked on.

Merging locks in whatever alignment the tracks currently have. If two recordings of the same conversation start at slightly different moments, the mix will have an echo, and it cannot be fixed afterwards.

Use the Time Shift tool to drag a track until a shared sound, a clap, a laugh, the first word, lines up on both. Zoom in far enough to see individual peaks, because close is not close enough at this scale.

Tracks, Align Tracks offers start-time alignment, which helps when recordings genuinely started together and were merely imported at different offsets.

For remote interviews, recording each person locally and aligning afterwards gives much better audio than recording the call itself. Hilite handles that alignment automatically and we make it, so read that accordingly.