Tomi
Image Unavailable
Image Unavailable


Image by UpTown-P (VoiceFace design/UTAU appearance)

Western Name: Tomi (Thomathy Chaston)
Eastern Name: トミ
Kana Pronunciation: トミ

Gender: Male
Age: (Supposedly) 16 as a base guideline

Release Date: June 20, 2021
Group/Team: VoiceFace (For this version)

Official Site


Technical Details

Voiced by: UpTown-P
Managed by: VoiceFace & UpTown-P

encode0.png alias0.png config0.png


Voicebank Distribution

For character information related to the original Tomi/FNV Tomi, please visit here. All voice databases produced by VoiceFace are available on the website for download as well as here.

Tomi has a five voice databases that are distributed under the VoiceFace website; VCV ACT3, STRIKETHROUGH & BLOODMOON for UTAU and two AI vocoders made for Diff-SVC.

All UTAU voice databases produced by UpTown-P and VoiceFace are free for download, as well as free for commercial use as long as the TOS/TOU are followed. Failure to follow Tomi's TOS/TOU will result in termination of your right to use his voice for commercial purposes


【*CORE_DATABASES (VCV_JP)*】

「 VCV_JP 3.0 (Alternative tone) 」
A 3 pitch VCV voice database that was recorded to be an alternative tone to the 1.0 & 2.0 VCV databases, it has a feminine vocal range with pitches recorded in E4, G4 and B4 (falsetto.) It was also made to be used by more experienced users.
It does not come with glottal stops, however it does include end breaths as well as vocal fry.
It is encoded in Romaji, and has hiragana aliases. It can only read hiragana (VCV only)

Recommended configuration settings (Resampler, wavtool, rendering options ect.)
» Engine/resampler: tn_fnds_PB
» Wavtool: ?
» Rendering options: Y0H0B0C99g+6eo

VCV_JP 1.0 Download


【*APPEND_DATABASES (CVVC_JP)*】

「 STRIKETHROUGH_JP 1.0 」
A 2 pitched CVVC database with a masculine voice, the pitches recorded are D#3 and F#3
It is encoded in Romaji, and has hiragana aliases. It can only read Hiragana (CV/CVVC only)

Recommended configuration settings (Resampler, wavtool, rendering options ect.)
» Engine/resampler: TIPS/tn_fnds_PB
» Wavtool: Default wavtool for UTAU
» Rendering options: Y0H0B0C99g+5

STRIKETHROUGH 1.0 Download

「 BLOODMOON_JP 1.0 」
A 3 pitched CVVC database with deep and husky tone, it is specifically designed for alternative rock and rock music genres. This database is configured by MoonFighter
It is encoded in Romaji, and has hiragana aliases. It can only read Hiragana (CV/CVVC only)

Recommended configuration settings (Resampler, wavtool, rendering options ect.)
» Engine/resampler: ?
» Wavtool: Default ?
» Rendering options: ?

BLOODMOON 1.0 Download


【*AI_DATABASES(Diff-SVC)*】

「 Tomi_AI_LITE 」
A Diff-SVC database comprised of English and Japanese training data, It was trained using the official Diff-SVC repo as well as UtaUtaUtau's repo. The data provided is comprised of singing samples only and uses pretrained models like Opencpop and Nyaru for better stability in voice. The data model was trained at 24khz, has 40 minutes of data and 100k steps of training.
The voice is young and feminine in tone while having a higher falsetto range and a boyish lower range, it is good for music genres like metal, indie, rock and techno. The range is roughly G3 - F5

Recommended render settings
» add_noise_step : 800 or higher.
» Recommended to use gt_mel (image-to-image rendering method)

Tomi AI LITE Download

「 Tomi_AI_REGULAR 」
A Diff-SVC database comprised of English and Japanese training data, It was trained using the official Diff-SVC repo as well as UtaUtaUtau's repo. The data provided is comprised of singing samples as well as speach samples. It uses pretrained models like Nehito, LIEE and a multilang speaker model made by lilyyy for better stability in voice and a wider range of language. The data model is currently being trained at 44.1khz, has 1h20m of data and 135k steps of training at the time of writing.

Similar to Tomi AI LITE, the voice is young and feminine in tone while having a higher falsetto range and a boyish lower range, it is good for music genres like metal, indie, rock and techno. The range is roughly G3 - F5

Recommended render settings
» add_noise_step : ?
» Recommended rendering method: ?

→ Private database and will not be released

「 Tomi_AI_FULL 」
A Diff-SVC database comprised of English and Japanese training data, It was trained using the official Diff-SVC repo as well as UtaUtaUtau's repo. The data provided is comprised of singing samples as well as speech samples. It uses no pretrain models so it can be used for commercial use freely as long as TOS is followed. The data model is currently being trained at 44.1khz, has 2h of data and 200k steps of training at the time of writing.

Similar to Tomi AI LITE, the voice is young and feminine in tone while having a higher falsetto range and a boyish lower range, it is good for music genres like metal, indie, rock and techno. The range is roughly G3 - F5

Recommended render settings
» add_noise_step : 1000
» Recommended rendering method: ?

Tomi AI FULL Download


Terms of Use

R-18 Content Allowed? See Additional Terms of Use
Commercial Use of Voicebank Allowed? Permission Not Required
Commercial Use of Character Allowed? See Additional Terms of Use
Do these terms apply to derivative characters/voices? Yes, but derivatives must be created with permission
Click here to view the terms of use for this UTAU.
Click here to view commercial license info for this UTAU.


Voice Sample(s)


Character Details


Height: 183cm

Weight: 168lbs

Previous affiliations and history.

Based on the popular game, Friday Night Funkin, Tomi's origins come from a project called Friday Nighters (Verse)/FNV/FNer. It was made in 2021 as a collaborative effort from a friend group with many characters corresponding to each individual within the group. FNers was fully disbanded as of early 2022, however 2 of the main creators still actively use their respective characters frequently.

In Late October of 2022, Tomi found place in a small scale vocal synthesis collective project called VoiceFace. Tomi still has place in FNV, but is now under the partial ownership of VoiceFace (with collaborative efforts from FNV & UpTown-P.)

Personality and bio.

In concept, Tomi is a demon that possessed the body of a life-sized art doll, his body is made out of fiberglass and canvas, his joints are similar to a ball jointed doll. The only parts of his body that aren't synthetic are his horns, eyes, tail and internal organs.

His personality is said to be shy, introverted, and cold to people he doesn't know, but can become easily clingy or jealous of people he likes. He is noted to have ASD and anxiety.

However, all official personality traits can be discarded in favour of user's personal preference or used as a base guideline for users to built off of.

Relations

They have no official relationship set in stone, but are considered best friends and are often paired together. A duo pack containing Tomi AI & Yuno NIO MASC AI was considered for released, but plans fell through.

They have no notable relationship set in stone, but are considered friends.

  • Astrid - Fellow VoiceFace member

They have no notable relationship set in stone, but are considered friends.

  • Nemo - Fellow VoiceFace member

They have no notable relationship set in stone.


Image Unavailable


Image by UpTown-P (FNV appearance)

Image Unavailable


Image by UpTown-P (AI version)

Image Unavailable


Image by UpTown-P (Mid summer appearance)

Image Unavailable


Image by UpTown-P (First appearance)

Image Unavailable


Image by UpTown-P (October 2021 appearance)

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License