Synthesizer V

Synthesizer V is singing voice synthesis software made by Dreamtonics Co., Ltd., first released in beta in August 2018 and officially in December 2018. The latest version, Synthesizer V Studio 2, launched in March 2025. Users enter MIDI notes for a melody and type lyrics for each note, then choose a voice from Dreamtonics' library; the software sings the part back, complete with breath, expression, and dynamics. The engine combines concatenative synthesis with a neural network-based singing model, working from real vocals recorded and licensed from human singers rather than generating voices from scratch. Interface language support spans English, French, German, Japanese, Korean, Portuguese, Spanish, Russian, and both Simplified and Traditional Chinese, while singing synthesis itself supports English, Japanese, Cantonese, Mandarin, Spanish, and Korean, with French, German, and Brazilian Portuguese available in beta. A cross-lingual feature, added in December 2020, lets a voice sing or rap in a language it wasn't originally trained on. Synthesizer V Studio runs on Windows, macOS, and Linux as a desktop application, plus a web version introduced in January 2020. It's offered in free "Basic" and paid "Pro" tiers, with individual singer voicebanks sold separately as demos, "Lite," or full versions. It is used by music producers, composers, and hobbyists to create vocal tracks without a human singer. One limitation is that voicebanks are licensed per-singer, so building a varied vocal library requires purchasing each voice individually.