{"rewrite":{"id":"r_0da319854182870dc6536470","clusterId":"c_ffd838ee9a595f87f7c2309b","slug":"writer-builds-personal-diffsinger-vocal-synth-from-his-own-voice","model":"deepseek-v4-flash:free","headline":"Writer Builds Personal DiffSinger Vocal Synth From His Own Voice","summary":"A writer for CloseBox and TechnoEdge documented building a personal DiffSinger vocal synthesis model, recording about an hour of his own singing over two days, then training it on an NVIDIA DGX Spark. The goal is to convert his recordings with an RVC voice changer to recreate his late wife's voice for new songs.","whyItMatters":"The piece shows how open-source AI vocal synthesis now lets individuals create custom singing voices from scratch, bypassing the cost and data demands of commercial vocaloids.","webCardHtml":"\u003cp\u003eThe writer\u0026#39;s previous experiment used SoulX-Singer, a zero-shot synthesizer, to make his wife\u0026#39;s voice sing in Japanese. That required only 10 to 20 seconds of data but remained unstable for full songs. DiffSinger, by contrast, demands about an hour of singing with each phoneme recorded at least 20 times, so he recorded his own voice over two days, then planned to convert it with an existing RVC model of his wife\u0026#39;s voice.\u003c/p\u003e\u003cp\u003eHe trained the model on a DGX Spark, reaching a preliminary checkpoint after several hours. The approach sidesteps the need for new recordings of his wife, whose singing exists only in a few tracks and an RVC model. The project builds on his 2013 UTAU-Synth voicebank, which used just three songs and took seven years to produce over 100 tracks.\u003c/p\u003e","blueskyPost":"A writer built a personal DiffSinger vocal synth from his own voice, aiming to convert it to his late wife's voice via RVC. Open-source AI keeps making custom vocaloids more accessible.","twitterPost":"A writer recorded his own voice for two days to train a DiffSinger model, planning to convert it to his late wife's voice with RVC. Open-source AI vocal synthesis keeps getting more personal.","threadsPost":null,"newsletterBlurb":"A CloseBox writer documented building a personal DiffSinger vocal synthesis model from his own singing, with the goal of converting it to his late wife's voice. The project shows how open-source AI tools are lowering the barrier to custom vocal synthesis.","attributionJson":"[{\"source\":\"GameBusiness.jp\",\"url\":\"https://www.gamebusiness.jp/article/2026/08/04/27608.html\",\"title\":\"自分と妻の歌声をAIボーカルシンセ「DiffSinger」に進化させた。もうボイチェン不要。アルゴリズムベースの自動作曲ソフトとも統合（CloseBox）\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":6047,"outputTokens":581,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1786314592,"createdAt":"2026-08-09T21:32:02.000Z","publishedAt":"2026-08-09T21:36:44.000Z","updatedAt":"2026-08-09T21:32:02.000Z"},"cluster":{"id":"c_ffd838ee9a595f87f7c2309b","canonicalTitle":"自分と妻の歌声をAIボーカルシンセ「DiffSinger」に進化させた。もうボイチェン不要。アルゴリズムベースの自動作曲ソフトとも統合（CloseBox）","representativeArticleId":"a_b811e4cb4c70032d6fda6d7d","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[\"DiffSinger\",\"SoulX-Singer\"],\"studios\":[],\"people\":[],\"type\":\"news\",\"domain\":\"music\",\"is_roundup\":false}","contentType":"news","status":"written","firstSeenAt":"2026-08-03T21:00:03.000Z","lastSeenAt":"2026-08-03T21:00:03.000Z","updatedAt":"2026-08-09T21:32:02.000Z"},"attribution":[{"source":"GameBusiness.jp","url":"https://www.gamebusiness.jp/article/2026/08/04/27608.html","title":"自分と妻の歌声をAIボーカルシンセ「DiffSinger」に進化させた。もうボイチェン不要。アルゴリズムベースの自動作曲ソフトとも統合（CloseBox）"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":["DiffSinger","SoulX-Singer"],"studios":[],"people":[],"type":"news","domain":"music","is_roundup":false},"keyFacts":null}