Music / Synth V

Synthesizer V, usually shortened to SynthV, is a music synthesizing program developed by Dreamtonics. The program's name is based on it being the company's fifth attempt at a voice synthesizer program.

As with other programs of this type, producing a song only requires typing in notes and corresponding lyrics, although the user can fine-tune the results using a set of parameters or breaking the notes down into individual phonemes. The program can also be plugged in to any DAW to make the mixing process simpler, especially when working with original songs. Finally, it is possible to import .vsqx files exported from Vocaloid or UTAU, with varied results.

A demo of the program can be downloaded for free from their official homepage here. On June 26th, 2020, Dreamtonics unveiled the second generation of the program called Synthesizer V Studio. This edition and its two versions, Pro (a paid version) and Basic (a free version), was on July 30th, 2020.

In October 2020, another generation was announced known as Synthesizer V AI, which instead utilizes deep learning artificial intelligence rather than the traditional tuning method. AI was first introduced to the program on February 18, 2021, and the software has been receiving periodic updates since then, now including AI-based tuning, cross-language synthesis, different Vocal Modes, High Dynamics Voice Models, the ability to do 'retakes', and even rap synthesis.

In June 2022, Synthesizer V AI Gold was announced, a series which promises to work with existing popular vocalists "with the sole goal of creating the perfect voice synthesizer for serious music production".

There is an official forum where you can discuss peripherals around the program, share your songs made with it, etc. The forums aim to be trilingual with English, Chinese, and Japanese posts, though you are allowed to write posts in any language as long as you tag the language(s) used.

Provides examples of:

• Accent Upon The Wrong Syllable:
  • A potential problem with homonyms, since the vocals' dictionaries have one pronunciation programmed per word. The user can convert input words to phonemes and edit them to avoid this problem, though.
  • Importing .vsqx files from Vocaloid almost always causes this, due to the programs using different methods to split up words per note and write out phonemes.^note To be specific, Vocaloid automatically splits words into syllables ("going" would be "go-" "ing") while Synthesizer V's words manually have to be split ("going" could be "go" and "ing" or "going" "+" depending on user preference).
• Alternate Company Equivalent: To Vocaloid in particular. (While Dreamtonics have re-released the UTAU Renri Yamine for SynthV, the developers have no plans to implement custom voicebank creation like for UTAU.)
• Ambiguously Human: Eleanor Forte is referred to as "being an epitome of the latest technology" on the official site, strongly implying she is an android (or virtual singer even in-universe). Solaria and Asterian might be celestial beings, but Eclipsed Sounds leaves it up to interpretation.
• Blank Slate: Downplayed. Vocals do have stated interests and/or personalities, but no backstories. There is nothing preventing the same producer from casting vocals in different roles, either.
• Canon Immigrant:
  • Yamine Renri is an UTAU vocal, and her vocals in SynthV are explicitly rebuilt off of her voicebanks in that software.
  • Beijing Photek S&T previously developed a Vocaloid called Xingchen^note also known as Stardust, and have rereleased her as Xingchen (or Xingchen Infinity).
  • Kotonoha Akane and Aoi were originally released for VOICEROID, and were unable to sing at all unless you had the plugin KotonoSync. The same goes for other VOICEROID-originating characters like Maki and Seika.
  • Xia Yu Yao is also an UTAU vocal.
  • The most famous UTAU vocal of all, Kasane Teto, became a SynthV voicebank in 2023.
  • Haruno Sora was previously released for both Vocaloid and VOICEROID before her SynthV voicebank in 2023.
• Colour-Coded for Your Convenience: When you use several voice tracks in the same project, they are given different colours by default to make it easier to differentiate between them when not selected. (However, the user could avert this and give all tracks the same colour. Tracks that are duplicated will be the same colour.)
• Colour Coded Elements: The vocals developed by Beijing Photek S&T have colour-coded designs inspired by their element (such as Chiyu having red hair and clothing and Haiyi having blue, ocean-inspired attire).
• Cover Version: A good number of the official demos are covers made with the original artist's permission.
• Expansion Pack: In March 2022, the Vocal Modes function was released, which allowed for voicebanks to have different emotional performances. For example, Feng Yi's "Opera" Vocal Mode changes her tone to sound like a Peking Opera singer, while Kevin's "Soft" Vocal Mode changes his tone to sound softer. The Vocal Mode function is only available for SynthV Studio Pro users, however.
• Improbably Female Cast: Before 2022 there were only two male voices (Genbu and Muxin) compared to the much larger cast of female voices. Since then there have been more male and even androgynous voices to even out the ratio, though there are still a greater number of female voices overall.
• Ooh, Me Accent's Slipping: Eleanor Forte was described as American English before her release, but her Lite voicebank has a slight British accent when rolling her r's. (This can also be invoked by editing phonemes manually in a word.)
• Punny Title: The V in Synthesizer V can either be read as a shortening of "Synthesizer Voice/Vocals" and as the roman numeral, given that this is Dreamtonic's fifth attempt at such a program. The latter interpretation also explains the program's save file format being .s5p.
• Robo Speak: Downplayed. The program uses legato effects, vibrato, and phoneme transitions by default to make the vocals sound more realistic from the start; however, producers who don't fine-tune + mix the vocals properly will still get results that sound emotionless. It gets downplayed even further with the AI banks, which manage to sound even smoother.
• Iconic Item: Eleanor Forte's sealing stamp-themed microphone.
• Vocal Dissonance:
  • GENBU, which is most evident in this demo once his full image is shown: the arrogant pose and haughty facial expression in his profile contrasts with the soft, beautiful voice from his beta version. His description on the homepage implies this is an Intended Audience Reaction.
  • Eleanor Forte narrowly averted this before release, according to an VocaloidNews interview:
  Kanru: I was kind of afraid that her voice and design wouldn’t match. She had a larger head-to-body ratio and a more mature voice until we pulled them closer on each end in early August.

---