Mono compatibility checker

Does your mix survive a phone speaker?

Phone speakers, smart speakers and many laptops play in mono. A mix with phase problems loses level, or loses a voice or an instrument outright, when its two channels are added together, and on headphones you’ll never hear it. Drop in a file to fold it to mono, see what it loses and where, and check whether a “stereo” file is really stereo. It runs in your browser; nothing is uploaded.

Drop an audio or video file here

WAV, AIFF, MP3, M4A, MP4 and MOV at any length. Other formats (FLAC, Ogg…) up to 64 MB. Analysed in your browser; nothing is uploaded.

What it checks

The loudness change in mono. The file is measured to ITU-R BS.1770 as stereo, then folded to mono as (L+R)/√2 and measured again, over the whole file and second by second. That fold keeps audio that’s the same in both channels at the same loudness, so the change is purely what phase costs: nothing for a centred voice, 3 dB for two unrelated channels (wide reverb, two different instruments), and everything for a channel that’s the other upside down. Devices fold at either −3 dB or −6 dB; a −6 dB fold makes everything 3 dB quieter again, for every file alike.

Phase correlation. How alike the two channels are, from +1 (the same signal) through 0 (unrelated) to −1 (one inverted), the reading a correlation meter shows.

Dual mono, one-sided and inverted channels. A stereo file with the same audio in both channels carries it twice, and a mono file would do. One with a silent channel plays in one ear on headphones. One with a channel polarity-inverted cancels in mono.

Where the thresholds come from

No standard publishes a pass mark for phase correlation or for loss in mono; correlation meters are read by eye. So the lines this tool draws are its own, each set with room to spare from a case that’s fine:

Cancellation: losing more than 4 dB in mono. Unrelated channels lose 3 dB by arithmetic, so it takes a full decibel past that before something is cancelling.
Inverted channel: correlation below −0.5. Flipping one channel of a mix turns its correlation negative; real stereo sits well above zero.
Dual mono: correlation of 0.98 or more, which means what differs between the channels is 20 dB under what they share. MP3 and Opus keep dual mono identical, but some AAC encoders leave it at about 0.99. A centred voice with stereo ambience 12 dB down reads about 0.94, and stays stereo.
One-sided: one channel 40 dB or more below the other.

How it’s checked

The loudness meter is the one behind the loudness checkers here, tested against the EBU’s conformance signals and against ffmpeg. The fold-down and correlation are tested against ffmpeg’s own measurements on generated files (true stereo, dual mono, an inverted channel, a silent channel, partial cancellation), where each problem file must raise its finding and the true stereo file must raise none. Only stereo is checked: a surround mix folds to mono through a downmix that differs from player to player.