Swap Left and Right Audio Channels

How the Audio Channel Swapper works

Stereo files hold two streams of samples, one for each speaker, and a surprising number of problems come down to those two streams being in the wrong place or one of them being empty. This page moves them around without touching the audio inside them: no resampling, no gain change, no filtering, just samples moved from one lane to the other.

Why you might swap left and right audio

Stereo only works because everyone agrees which side is which. The convention goes back to the Blumlein patents of 1931 and the stereo LP of 1958: the first channel is left, the second is right. WAV, AIFF, FLAC and MP3 all store samples interleaved in that order, and the AES3 digital interface used in studios and broadcast carries channel 1 as left by the same rule. When a single link in the chain gets it backwards, every later stage faithfully preserves the mistake. A drummer who was panned stage right ends up on the left, a presenter looks left on camera while their voice comes from the right, and a film mix that was checked in a room sounds wrong on headphones.

Silent channels are the other common fault. A lapel or shotgun microphone has one output, but most camera and field recorder inputs are stereo pairs. Plug the microphone into input 1 and the recorder writes the voice to the left channel and a flat line to the right. On speakers this is easy to miss; on headphones, which is how most podcasts and online videos are heard, it is a voice in one ear and silence in the other.

Where each option is used

Swap left and right. Cables get plugged in the wrong way round, camera microphones label their channels backwards, and sometimes an editor simply wants the guitar on the other side. Swapping exchanges the channels sample for sample with no other change. It is also the fix for XLR-to-RCA looms wired to the wrong jacks, and for footage from cameras that were mounted upside down and flipped in the edit, where the picture is corrected but the sound field is still mirrored.

One channel on both sides. Choosing the channel that has the audio and copying it to both sides turns a one-sided file into a proper centred recording. This is different from the Stereo to Mono tool, which averages the two channels — averaging a live channel with a silent one halves the level, a drop of about 6 dB, whereas copying keeps it exactly as it was. The same option rescues a two-microphone interview where one microphone failed, or a dual-mono recording where the second input was set as a safety track 10 or 20 dB lower and you want the louder take everywhere.

Invert right channel polarity. Flipping the sign of every sample in one channel makes the two channels push and pull against each other. On a mono-compatible file the result mostly cancels when summed, which is why engineers use polarity flips to check for phase problems and why a flipped file sounds oddly hollow and hard to locate on headphones. It is also the trick behind "vocal remover" demos: inverting one side and summing cancels whatever was in the centre. This page does the flip; the Stereo to Mono tool can do the sum, and the Vocal Remover does both in one go. It also repairs a real fault: a balanced cable with pins 2 and 3 crossed flips one channel, and the Phase Checker will show a strongly negative correlation when that has happened.

Tips and pitfalls

• Check the peaks before choosing. The stats panel shows the peak level of each channel. A dead channel reads around −60 dBFS or lower, or no signal at all; a live voice usually peaks between −20 and −3 dBFS. Pick the channel with the real signal.
• Near-silent is not silent. A channel carrying only hiss or a faint bleed of the other microphone will still show a low peak. Copying the good channel over it is usually the right call, because summing in a little noise and a delayed copy of the voice makes the result worse, not better.
• Swap is not a polarity fix. If a file sounds thin and phasey, swapping the sides changes nothing. Inverting one side is the tool for that.
• Don't invert a file for release. A polarity-flipped stereo file cancels badly on mono speakers, phones and smart speakers. Use inversion to diagnose or to feed another process, then undo it.
• Keep WAV for further editing. WAV export is bit-exact for the swap and copy operations, so nothing is lost if the file is going back into an editor. Compressed formats add a generation of loss each time they are re-encoded.

The blog post on fixing a stereo recording walks through the whole diagnosis, from spotting a one-sided file to deciding between copy, swap and mono.

Mono files

A mono file has one channel, so the tool copies it to both sides and the export is a stereo file with the same audio centred. The operation setting is ignored for mono input.

How to read the result

The player plays the processed file so you can compare it by ear with the original. On headphones a swapped file should put every sound on the opposite side, a copied file should sit dead centre, and an inverted file should feel diffuse and hard to place. The stats panel reports the peak of each output channel: after a copy the two numbers should match exactly, after a swap they should have traded places, and after an inversion they are unchanged because flipping the sign does not change the level. Sample rate and length are unchanged, so the output lines up sample for sample with the original if you drop it back onto a video timeline.

Related tools