Stereo to Mono Converter & Channel Router

Most channel converters change the channel count and tell you nothing. This one measures your file first, because whether a mix survives mono is a question about the file rather than about the setting you picked.

• Phase correlation. A number from +1 to −1 comparing the two channels. +1 means they are the same signal, so folding to mono loses nothing. Around 0 means a wide image. Below zero the channels are cancelling each other, and mono will be noticeably quieter and thinner.
• The curve, not just the number. One figure for a whole track hides a chorus that collapses inside verses that are fine. The curve shows correlation over the length of the file, so a problem section is visible as a dip.
• Cancellation by band. The same fold-down measured across bass, low mids, high mids and treble. This is the reading that tells you whether the loss matters: thirty decibels missing above 5 kHz is a stereo widener on a hi-hat, and six decibels missing in the bass is a part that vanishes on a phone speaker.

All three are measured on the file you dropped in, using the same fold-down the tool would actually apply — not an idealised one that flatters the result.

• Mono sum Both channels at half each. This is the correct downmix for speech, podcasts, phone playback and anywhere a single channel is required.
• Dual mono The same sum written to both channels. Use it when a platform insists on two channels but the content should be centred.
• Left only / Right only / Extract One channel on its own. This is the fix for an interview recorded with one microphone into one side, and the only way to hear a surround centre by itself.
• Swap left and right For a recording wired backwards, and for checking whether a stereo image is the way round you think it is.
• Balance Turns one side down. It never turns the other side up, so it cannot clip a file that was already fine.
• Stereo width Narrows or widens the image through the mid/side pair. 0 % is mono, 100 % is untouched, 200 % is double width — and wide settings are exactly what the band table is there to warn you about.
• Mid/side encode and decode Converts left/right into mid and side, and back again. The pair is exactly invertible here, so an encode followed by a decode returns the file you started with.
• Surround to stereo ITU-R BS.775 Lo/Ro for ordinary stereo playback, or Dolby Lt/Rt for a surround decoder. The LFE is excluded by default, which is the standard.
• Custom matrix Type how much of each input channel goes into each output channel, as a percentage. Negative values invert polarity. This is the control a desktop channel mixer gives you.

Summing two channels can produce a signal louder than either of them. The usual answer is a fixed safety trim — turn everything down by a decibel and hope. That is wrong in both directions: too little for material that sums hard, and needless attenuation for material that does not.

This tool measures the routed signal's actual peak, then applies the exact gain needed to land on the ceiling you chose. A file already below the ceiling is left alone, because the guard is a ceiling and not a normaliser — raising a quiet file is the volume booster's job.

Then it reads the finished file back. The peak, the loudness and the bit depth on screen come out of the bytes you are about to download, not from the settings that produced them.

A clipped file does not necessarily have a high peak. Once samples have been flattened against full scale, any trim applied afterwards scales those flat tops down with everything else — so the file reads comfortably below zero and is still distorted.

That is not hypothetical. The page this tool replaces asked for the peak, saw it sitting below its threshold, and printed PASS for a file measured carrying 9.66 % total harmonic distortion.

The reading that separates them is flat factor, which counts how much of the waveform sits flat against its own maximum. A loud-but-clean file and a clipped one are indistinguishable by peak and thirty units apart by flat factor. The number is shown beside the verdict so the verdict can be checked.

Input: MP3, WAV, FLAC, M4A, AAC and OGG. Output: MP3, M4A, AAC, OGG, FLAC and WAV. The sample rate stays at the source's own unless you change it, so nothing is resampled behind your back.

MP3 carries at most two channels, so a six-channel matrix into MP3 is capped at two and the tool says so at the control rather than silently downmixing.

Every routing, the free-form matrix, per-channel gain and polarity, the surround downmix, the clip guard and the whole mono check are free on both plans. Pro adds batches of up to 25 files as one ZIP, 48 kHz, 320 kbps and 24-bit output.

Powered by optimised processing algorithms.

Every measurement and every conversion happens on your own machine. Your audio is never uploaded, never stored and never sent anywhere — which is why the tool works with the network off once the page has loaded.

• LUFS meter
• Audio normalizer
• Resample & bit depth
• Audio merger