Stereo interview input
Speaker A is on the left. Speaker B is on the right. The voices overlap near 7.5 seconds.
Download Stereo interview inputFind out whether a two-person recording has one speaker on each channel. Split the 12-second interview and compare your exports with the two reference tracks. Then listen to the mixed version to hear what changes when both people share a single channel.
Speaker A is on the left. Speaker B is on the right. The voices overlap near 7.5 seconds.
Download Stereo interview inputMono WAV containing only Speaker A. The silences keep the original timing.
Download Expected left outputMono WAV containing only Speaker B at the original times.
Download Expected right outputBoth speakers share one channel. Open this file to hear how their overlap sounds in a single mix.
Download Mixed mono comparisonSeparate original tracks give you the most control. If one speaker appears on each stereo channel, split the channels and check for bleed. If one person is simply louder on one side, channel splitting preserves that difference along with every other sound on that channel.
When a mix places both speakers on one channel, keep the best original. You can edit passages where people take turns. Overlapping speech needs a different method and should be checked word by word.
The exercise voices are original scripted lines generated with eSpeak NG formant synthesis. The CSV lists all four lines and their source times.