The latest articles on DEV Community by 雷笛歌 (@_e7d82c08f61159ad3b2ea). https://dev.to/_e7d82c08f61159ad3b2ea https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4089563%2F9b66a8c1-6e2e-4036-8b87-8c4139ca658d.jpg https://dev.to/_e7d82c08f61159ad3b2ea en 雷笛歌 Sun, 23 Aug 2026 03:06:30 +0000 https://dev.to/_e7d82c08f61159ad3b2ea/keep-backing-vocals-with-an-ai-vocal-separator-1momirr-1dop https://dev.to/_e7d82c08f61159ad3b2ea/keep-backing-vocals-with-an-ai-vocal-separator-1momirr-1dop <p>Most <strong>AI vocal separator</strong> / <strong>online vocal separator</strong> tools optimize for karaoke: remove the lead. Producers often need the harder job: <strong>keep backing vocals</strong> and clean <strong>backing vocal separation</strong> without wiping the harmony stack.</p> <p>This post walks the practical pipeline in the <strong>MO Mirror audio engine</strong> on <a href="https://1momirr.com" rel="noopener noreferrer">1momirr.com</a>: separate → acoustic scan → smart mastering → listen.</p> <h2> What usually breaks </h2> <ul> <li>Doubles get glued to the lead and disappear together</li> <li>Residuals smear into the instrumental</li> <li>You only hear it after bouncing stems in a DAW</li> </ul> <p>Always judge on the densest chorus, not the intro.</p> <h2> Step 1 — Vocal separation </h2> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F1drfmik0uilh4db5mhbm.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F1drfmik0uilh4db5mhbm.png" alt=" " width="800" height="450"></a></p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F6q8ku9kx683d68edc422.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F6q8ku9kx683d68edc422.png" alt=" " width="800" height="450"></a></p> <p>Treat the first pass as a stem split, not the final bounce.</p> <h2> Step 2 — Acoustic scan </h2> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fy6xughoxstbp4c47f735.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fy6xughoxstbp4c47f735.png" alt=" " width="800" height="450"></a></p> <p>Scan after the split. Muddy leftovers and harsh edges show up here before you commit.</p> <h2> Step 3 — Smart mastering </h2> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fbkckvm6dn5uzbqztd7wj.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fbkckvm6dn5uzbqztd7wj.png" alt=" " width="800" height="450"></a></p> <p>Stem splits often leave uneven loudness. Keep mastering light and musical.</p> <h2> Step 4 — Listen / export </h2> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fb2h3cndcddawg3tgsyex.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fb2h3cndcddawg3tgsyex.png" alt=" " width="800" height="450"></a></p> <p>If backing vocals vanish on the chorus, you over-separated.</p> <h2> Web vs desktop (do not mix these up) </h2> <ul> <li> <strong>Web (1momirr):</strong> quick online try of the pipeline above</li> <li> <strong>Desktop:</strong> seamless <strong>ABCD</strong> A/B listening with no scrub gap</li> <li>The <strong>website does not</strong> provide AB compare audition — that UX is desktop-only</li> </ul> <p>Dragging files between apps hides differences because of timing gaps. Seamless same-timeline switching does not.</p> <h2> Try it </h2> <p><a href="https://1momirr.com" rel="noopener noreferrer">https://1momirr.com</a></p> <p>Primary terms: <strong>1momirr</strong>, <strong>MO Mirror audio engine</strong>, <strong>keep backing vocals</strong>, <strong>backing vocal separation</strong>, <strong>AI vocal separator</strong>.</p> ai audio music productivity 雷笛歌 Sun, 23 Aug 2026 02:06:47 +0000 https://dev.to/_e7d82c08f61159ad3b2ea/ai-audio-music-productivity-1dam https://dev.to/_e7d82c08f61159ad3b2ea/ai-audio-music-productivity-1dam