#WavTool

[utau-4-you]

Resampler/Wavtool Directory

Resamplers and wavtools are what make your voicebanks sing! Resamplers together with the wavtool take your samples and take them apart, pitch, edit, apply flags, etc. and then reassemble them to create the audio output for your voicebanks. They also are responsible for generating FRQ (or other types of frequency map) files so that they can quickly go through this process.

Resampler List

resampler- this is the default resampler that comes with UTAU

fresamp ver11 ver14 by Ameya- known for strengthening voices and reducing background noise. Can have a nasally quality on higher notes. (Mod heta note: I personally think fresamp11 produces higher quality renders than fresamp14)

world4utau (w4u) by Ameya - is very picky with voicebanks, reacts very strongly to background noise. Requires FFTW files to work. Comes with source code. w4u Zany ver. by Zany is an updated version. wn4u/worldend4UTAU by Zany ports a different version of WORLD0.2.0

phavoco by Ameya- Phase vocoder, adds a vocoder effect to renders. phaavoco is an updated version. There's also an archive of this resampler's download in case the UTAU site goes down put together by Mod heta.

Model4/M4 by Ameya- was a privately distributed through UTAU Days, an invitation only SNS. Beta resampler that was never publicly released due to instability.

doppltler by Ameya- Supports 16, 24, 32 bit .wav files, Supports mp3, OGG Vorbis and FLAC encoded voicebanks. OTO files should look like ".mp3.wav" in order to use them. The R flag reverses samples and can be used to create pseudo-CVVC. There's also an archive of this resampler's download in case the UTAU site goes down put together by Mod heta.

f2resamp by Ameya- A completely new resampler meant to be part of the fresamp collection. There's also an archive of this resampler's download in case the UTAU site goes down put together by Mod heta.

TIPS by ScientistB- usually works really well for softer voicebanks. Can make banks have a metallic quality and emphasize background noise more than other resamplers.

bkh01 by Zteer- With the right flags, you can get more realistic renders with softer voices but with stronger voices, it can create a lot of distortion, metallic and nasally qualities.

tn_fnds by Zteer- Uses looping samples to render files instead of stretching unlike most resamplers. Using the e flag will make it stretch. Similar to TIPS but doesn't seem to magnify noise as much. Requires FFTW files to work. tn_fndsX and tn_fnds_PB by Zany are variations of tn_fnds. young3 by Zany is a version of tn_fnds that uses Harvest.

EFB-GT by Custom.Maid- Works very well for soft voices and smoothing out noise while getting crispy consonants

EFB-PB by Custom.Maid- Likely works well for soft voice but requires EFB-GT to work (See below for why it be hard to use) (Mod heta note: I don't think I've ever been able to get this resampler to work in 12 years of using UTAU; you can get away with not downloading it.) Requires FFTW files to work.

EFB-GW by Custom.Maid- overseas users have had trouble getting this resampler to work. (Mod heta note: I don't recall being able to get this resampler to work in the 12 years I've used UTAU) Requires FFTW files to work.

WARP by Custom.Maid- Works better on lower notes but renders fast. Uses .llsm files. EFB-BS by Zany is a version of WARP that can use FRQ files instead of llsm files. EFB-BH by Zany also can use FRQ files but uses HARVEST.

moresampler by Kanru Hua- Based on libllsm. Flags can greatly change the voice such as breathiness, tension, growl, etc.

utaugrowl by _torifly- patches resamplers with growl capabilites and flags.

presamp by Delta- Automatically converts USTs between formats such as CVVC and VCV. You need to patch resamplers to use.

PyRwu by Delta- PyWorld based resampler.

What are flags

Flags are an editing feature that change the tone of the voice. Each resampler uses their own unique set of flags that you enter into the rendering options box in the project property or flag box in the note property. It should be noted that flags in the note property will override the flags in the project property. You can find out about all the different flags using the Flaglist by susrever. This document has the list of flags for most publicly available resamplers and what they do.

Resampler Plugins

Speedwagon by Custom.Maid- quickly creates FRQ (or whatever other file types) for resamplers.

Resampler Patcher by nmasao- improves resampler speed and performance, and also allows you to use multiple resamplers in the same UST. Uses flags to switch between the different resamplers.

Wavtool List

(Mod heta note: to be entirely fair, wavtools mostly change smoothness but you might not be able to tell the difference between them to be entirely honest.)

wavtool by Ameya- default wavtool. wavtool64 by Ameya ports it into 64bit.

wavtool2 by nmasao

wavtool4vcv by nmasao -made specifically for smoothing out VCV voicebank but also works for CVVC.

wavtoolex by Delta- included with presamp

moresampler by Kanru Hua

#utau #resource #utau library #resampler #wavtool

[domioblog]

「AIで音楽をより深く楽しもう！最新音楽アプリの紹介」

「AIで音楽をより深く楽しもう！最新音楽アプリの紹介」 #音楽 #music #ライブ #ギター #バンド #歌 #弾き語り #レコード #洋楽 #dj #歌ってみた #カバー #ボーカル #rock #song #dance #ベース #スタジオ

音楽をより深く楽しむために、最新の音楽アプリを紹介します。音楽にAIを取り入れたアプリが増えてきており、音楽制作から楽曲の発見まで、様々な場面で活躍しています。以下では、音楽制作に特化したORB Composer、楽曲の発見に特化したバンドナビ、無料のAI作曲アプリFIMMIGRM、音楽アプリのランキングTOP10、そしてAIによるトラック作成が可能なWavToolについて説明します。 ORB ComposerバンドナビFIMMIGRMWavTool ORB Composer Logic ProやCubase、Studio OneなどのDAWと連携して使うことのできるAI作曲ツールです。ORB…

View On WordPress

#AI #AIツール #ア���リ #ソフト #バンドナビ #音楽 #FIMMIGM #ORB #WavTool #創作 #作詞

[sundial-bee-scribbles]

surprise, bet you weren't expecting THIS

ok but real this is technically just a test. wasn't actually aiming to really make this lmao, but things lined up and i realized this sounded good actually so i was "hmm maybe i should post that." idk if i'll make a video for this so for now just take the audio ig.

VSQx by Hatsune Dan, distributed by Yumeko. Some minor tuning edits by me. Mixing also by me. Original song by wowaka.

#my audio lol #vocaloid #vocaloid cover #utau #utauloid #utauloid cover #utau cover #openutau #nurse robot type t #two faced lovers #wowaka #this is her sherry vb specifically; i wanna do smthn else w/ it but i was testing how she sounded like w/ files i already had #and this was one of the only ones where she actually sounded good 😭😭 maybe the different resamplers and wavtools are a piece of the puzzle #but still i'd need to mess around w/ her maybe. i got her cause i wanna do a specific thing w/ her but we'll see if i ever get there lol #i feel bad ugh i didnt actually do that that much here :'| feels like cheating. but if i keep trying maybe i'll get better some day #i had this vsqx just sitting in my files for a while lol... i have a lot of shit in there honestly that i plan to do stuff w/ but havent #i was actually intending to make a cover of this w/ iku b/c ily her (a slowed version tho b/c the song's sadder slowed imo) #we'll see if that happens tho lolll... rip to all the covers that never actually happen. ugh i dont really like any of my past covers :') #im stubborn so maybe ill figure it out SOMEDAY. but bruh how is it im better at making them swear at each other than like. sing #CAUSE YKNOW SINGING IS THEIR INTENDED PURPOSE. ugh. vsynths got me actin crazy #sorry if the mixing's kinda bad i need to read up on how to do better mixing... tried smthn a lil different here and i think it sounds okay #but who knows. idk

[raytaku]

So lately I've been dealing with this bug in utau where, when I'm using Olio's english vb, random notes will be a lot louder than the others when I haven't changed the volume of any of the notes

After a lot of work I Think I've narrowed it down to a bug in wavtool. Unfortunately I can't find Why this is happening or how to fix it (besides swapping out wavtool), and I haven't found a solution online

And I don't want to swap out wavtool bc I don't like how Olio sounds with the wavtool alternatives : /

#I wouldn't be surprised if wavtool was broken by windows updating bc utau Has had those issues before #perhaps I should make a utauforum account to help w troubleshooting lol

[utau-bowl]

Maria Fuwa AUSPICE+ Review

My life is getting busier, with my birthday, work, and a multi-country vacation coming up! That’s why I’ve been writing these so quickly. I reviewed Maria last year, and was asked to look at her again, with her newest voicebank. You all know how much I love her, so this one might be a little more positive than usual!!  Maria’s voice type and tone has mostly stayed the same, it’s the small things that have changed.

This review was requested by HIRA(formerly iosion). 

*Art by HIRA

Hear a Sample of Maria Fuwa AUSPICE

Bio

Name: 不破マリア (Fuwa Maria) Age: "20" y.o. Gender: Fluid Height: 5'5" (165 cm)

A pyromancer infamous for her unmatched prowess in robot matches. Maria is an extremely competitive person and prioritizes winning above all else. Surprisingly enough, she's no sore loser (or so she claims). She's rather "humble" in contrast to her aggressively competitive nature, but she would always demand a rematch when the opportunity strikes.

Official Site

Maria has a website here on Tumblr. Here is her voicebank page. Please read the TOS before downloading. The full version of AUSPICE voicebank is 945 MB when unzipped and has fully generated frq files. AUSPICE is multiple appends put together, so for a smaller file there are separate download options.

First Impressions

This Maria is intended to be the penultimate version of her. She comes with a set of normal, Soft, and Power pitches as well as Whisper and Screamo appends named Spirit and Howl. I also like the addition of different designs/icons for each append. It’s very cute.

Configuration

Maria is a VCV with each set(Normal, Soft and Power) recorded at A3, D4, F4, and G4. The G3 pitch from Ifrit is omitted, but her low range doesn’t suffer that much for it. Maria isn’t even meant to sing in tenor and lower ranges, so I have no issue with this. Spirit and Howl are monopitch and at F4 and C3 respectively. There are VC endings, and CV aliases. She’s sort of usable as a CVVC, but few would probably choose to use her that way over VCV. VC endings help for added realism though. She comes with glottal stops and vocal fry for all vowel combos, end breaths/inhales, and standalone breaths. The standalone breaths don’t sound very good in the program.

There tends to be a lot of room between different syllables and long consonants. She does have a westernized accent but not a talking tone like overseas UTAUs tend to have. To fix this, setting the consonant velocity to less than 100 sometimes helps.

With more than 20,000 oto lines, it also take a while to get into voicebank settings/view the oto. That’s more a fault of UTAU itself than Maria, though. The oto uses a usual 1:3 ratio of 80:240 for preutter/overlap. This also goes for the VC endings. The samples feel a bit cut off and it removes some of the uniqueness from the power append, but I can see why this was done.

Her power append is a more natural sort of power and not screaming/shouting like a lot of other UTAUs go for. It does cause the normal and power pitches to sound a little similar, but listening enough reveals a clear difference.

Spirit and Howl are interesting, the screamo append is recorded in a lower register which I’ve never seen before. They both require the use of tn_fnds in order to draw out their intended sound but TIPS is an okay substitute. Spirit is so quiet it’s often tough to integrate with the other pitches, even the soft ones.

My recommended flags: F0B0Y0H20C99P89

My recommended resamplers: wavtool4vcv(wavtool), TIPS, tn_fnds, doppeltler

Final Thoughts

Maria may have changed but she is still peak UTAU. She deserves to be used more, and I don’t say that a lot.

Got any other UTAUs you want me to review? Send an ask!  

#maria fuwa #fuwa maria #utau #utauloid #review #requested review #thanks for asking again nicely hira #request

[maks19770926]

WavTool - Нейросети для работы с аудио - WavTool Генерируют музыкальную композицию на основе текстового описания. Отлично подходит для создания простых фоновых мелодий для торговых центров, лифтов или презентаций. Кстати, под капотом у нейронки лежит языковая модель GPT-4, которая используется в небезызвестном ChatGPT. Предоставляет большой набор настроек, с помощью которых можно уточнить запрос. Готовые треки можно редактировать и сохранять на компьютер. СТОИМОСТЬ: бесплатно до 10 запросов каждые 8 часов. Если

[henb-y]

hi gamers heres a hot tip from henby about using the silly goofy free software UTAU-

initialized wavtool + resampler!! very useful if you’re setting up a UST for use and have a favorite wavtool + resamp that you’re gonna use. 

if you go into your options, then to path, you can set what your preferred wavtool + resampler is. so when you start a new project, it’ll just have them set up already! if you’re using a UST that has other tools saved, you just hit the initialize button and the program will set your initialized wavtool + resampler as the ones for that project!! ex:

the only thing to note when doing this is that when you’re creating a bank and get around to generating frequency files (frq), the initialized resampler is going to be the resampler that is generating those files. some resamplers aren’t supposed to do this, and thus will give a bad result! (ex: dopp is good for general usage, but bad at frq gen!)

#utau #vosynth #utau tip #I dunno if this is well known or not #but this is something that I use often! #so in case people don't know about it it doesn't hurt to share!!!

[hisuix01]

新DTM時代が爆誕！ AIにテキストで指示して作曲できる音楽アプリ｢WavTool｣が公開 | ギズモード・ジャパン

[blackfield]

[d23cozy]

[echanique]

[96theater]

something went Horribly Horribly wrong with the tuning on the hack writer ust i was working on so i have to do it. All over again and now it sounds Bad

#same resampler same vb and wavtool #just #years of academy training wasted #for some reason there was a bunch of redundant pitch information #like it would be like #pitchbend info: 299 #299 #and just keep adding about 200 or so of those 299s #i reset it but then it was only letting me do linear style bends soooo #new ust baby #i just copied the notes into another document #van posts #van does utau

[auspicious-voice]

UTAU Tutorials - How To Use USTs

Having discussed what USTs are and having installed some voicebanks, the next step is to naturally download some USTs and make the voicebanks you installed sing them.

In this tutorial, we’ll talk about how to fit USTs, as well as explaining what resamplers, wavtools, and plugins are and what they do.

Downloading USTs

You don’t really need to install USTs by the way. You just download them and extract them to a folder with file extractor software such as WinRAR. UTAU doesn’t have a dedicated UST folder, so I recommend making a folder within its root directory where you can put USTs into.

You can look up “UTAU UST download” and see what kind of USTs you can use.

Fitting USTs

Keep in mind that most USTs can come in many formats, such as Japanese CV or VCV, or in other languages. Depending on the voicebank you are using, you might want to use the appropriate voicebank so it can read the UST correctly.

Loading USTs

There are two ways to load a UST in UTAU:

Double-clicking the UST file

Opening and loading the UST file within UTAU

Once the voicebank is loaded, there might be a chance that the screen pops up as this:

The text may display in red because the required voicebank, resampler, or wavtool is missing.

Just pick whatever voicebank, resampler, and wavtool you currently have, and click OK to properly load the UST.

What are Resamplers and Wavtools?

The Tool 1 (append) is the wavtool, which specializes in stringing the rendered notes, and the Tool 2 (resample) is the resampler, which specializes in synthesizing the selected notes.

It should be worth noting that the default resampler and wavtool included with the installation of UTAU may not be a one-size-fits-all, and this depends on the voicebank’s quality. If that is the case, then I recommend that you look around and find other wavtools and resamplers that help make the voicebank that you’re using sound better.

Converting USTs

When converting USTs, you might want to consider the following:

The encoding of the voicebank (Romaji? Hiragana?)

The encoding and format of the UST (Romaji? Hiragana? CV? VCV?)

This usually applies to Japanese USTs. USTs in other languages is another topic I’d like to get into in the future.

Plug-ins and How to Use Them

Plug-ins are used in UTAU to make things a bit easier for you. To get iroiro2, download the plug-in we need here, and since most plug-ins come in archives, extract it to the plugins folder included in the UTAU root directory.

If you still have UTAU open, go to Tools(T) > Plug-Ins(N) > Reload(R) to reload the list of plugins. Go to the plugins option again and select the iroiro2 plugin. Depending on the plug-in name, it may show up as Japanese, so to change the language to English, go to Settings and click on the last checkbox. Close the plugin afterwards.

Since we are going to fit the UST, press Ctrl+A to select the entire UST, click on the iroiro2 plug-in, and the plugin’s name should show up in English with the following interface:

Now depending on what voicebank you’re using and what format the UST is in, the actions shown in the menu and what they do are rather straightforward. Use any of the actions as needed and click OK to make the changes.

Adding The Final Touches

Now that most of the UST is fitted to the voicebank, the last step is to clear any settings that may make the voicebank sound weird.

Again, press Ctrl+A to select the entire UST, right-click on a note and select Region Property.

The Note Properties box contains the settings for all of the notes in a UST. Here’s brief explanation on what the parameters do:

Intensity - the volume of the note.

Modulation - the pitch fluctuation of the note.

Preutterance and Overlap - these settings actually belong to the oto.ini of a voicebank, and aren’t really used that much.

Consonant Velocity (beta) - changes the speed of the consonants in between note transitions.

BRE - the breathiness parameter.

Flags - Other parameters that can alter how the voicebank sounds. See here for a comprehensive list of flags and what they do.

STP - the “starting point” of a note, measured in milliseconds (ms). Not used that much, but can be useful to emphasize transitions.

Clear all of the Flags, BRE, STP, Preutterance, Overlap, and Consonant Velocity (beta). Then, set the Intensity parameter to 100%, and the Modulation parameter to 0%. Click OK, and your UST is fitted. From this point on, you can make the voicebank you selected sing, or make further adjustments to the UST.

#tutorial #utau #utauloid

[utau-bowl]

Bitter Review

Hi again! Life has been busy lately but I still like UTAU after it all. I know this review is a little late, but I still really wanted to do it! Tasteloid is a really impressive and professional project, and I hope I can show how they have so much potential. I can’t wait for the release of their new boy next month!

*Art by Guillotama

Hear a Sample of Bitter JPN

Bio

Name: Bitter/苦い (Nigai)

Age: 20

Height: 5'3"

Gender: Female

Birthday: October 22 (Libra)

Item: A Switchblade named 'Mrs Butterfingers'

Creator/Voice Provider: Guillotama 

Likes: Vandalism, arson, petty theft, eggs, plushies, chaos

Dislikes: You probably

​A crude and abrasive young girl. Bitter sings in a harsh, deep, yet energetic and chaotic style.​ 

Official Site

Bitter has an official page on Tasteloid’s website. There’s a direct download on her character page. Her voicebank is about 230 MB when unzipped.

First Impressions

Bitter has an impressive, strong voice. It feels like a voice that was made to be in UTAU-and sounds crystal clear and even a little...realistic? Her overall tone is nasally and powerful, at both low and high ranges. She has a strong accent I was so shocked when using her for the first time because she’s kind of a dream voice for me. If this isn’t your favorite type of voice, Bitter might convince you to give this genre of UTAU a try.

Configuration

Bitter is a contralto/tenor with a range from G2-D5. She has five pitches at C3, E3, G3, B3, and F4(a falsetto).

The oto is one of the best I’ve seen in a CVVC - very consistent and well thought out! There are only a few extra samples, but a falsetto append, glottal stops, and vocal fry is there. I like these glottal stops since every vowel and N is accounted for. The vocal fry is a little subtle, but that makes it more versatile. A humming append is also there, with options to hum between vowels. She has no breaths though, except for one wav file with a series of breaths to be used during mixing.

The C3 humming samples are a little too pitchy, especially the mm a one, and they sound weird in the editor? These probably should have been re recorded.

The default resampler works okay with her, but it doesn’t quite bring out the best in her voice. Though it’s a little difficult to find the “just right” resampler for her. Playing around with wavtools might help too.

My recommended flags: g-3F3A-80H50

My recommended resamplers: moresampler, TIPS, vs4u, doppeltler

Final Thoughts

Bitter is one of the best UTAUs of 2021-a standout entry in the UTAU catalog with a decent number of extra features.

#utau #utauloid #bitter #utau review #review #bittersweet #tasteloid #long

[maks19770926]

WavTool - Нейросети для работы с аудио - WavTool Генерируют музыкальную композицию на основе текстового описания. Отлично подходит для создания простых фоновых мелодий для торговых центров, лифтов или презентаций. Кстати, под капотом у нейронки лежит языковая модель GPT-4, которая используется в небезызвестном ChatGPT. Предоставляет большой набор настроек, с помощью которых можно уточнить запрос. Готовые треки можно редактировать и сохранять на компьютер. СТОИМОСТЬ: бесплатно до 10 запросов каждые 8 часов. Если

[nick-the-nick-blog1]

UTAU REVIEW #1

Voicebank type: powerscale 

Recorded Pitches: A3-C#4-E4-G4-E5(from where the E5 is an optional belt pitch that is not prefix.mapped) 

Theorical range: G#3~D5 (The E5 can be used to extend that)

Optimal resamplers: fresamp11 and vs4u (B0F0 flags can be used)

Wavtool: wavtool4vcv

Gahata Meiji is a voicebank made to work in the UTAU vocal synth program.

It was recorded by the japanese user Saccu and illustrated by Souno Kazuki. Meiji is supposed to be a witch and from all her banks HUNTER is the one that represents her core design the best. Hunter is the most eclectic Meiji bank, in our point of view it represents a strong voicebank the best, all its recorded pitches come off well and sound strong even in the lower registers,giving her a sultry and mischivious voice.

Meiji tackles tuning very well, her samples most of the time suffer little to no distortion when adding exaggerated pitchbends making her fun to tune.However since the highest recorded pitch is her G4 when the song reaches notes past A4 tuning her too much may cause her voice too break.This can be used to your advantage making her voice to sound a little bit more emotional, but if the song needs more aggresive sounds she can’t handle them.

Mixing wise, she is medium-challenging due to her volume problems, she tends to come off very clean, she does tend to become too sharp if not equed properly. Her volume problems aren’t caused by the general volume of the sample but the actual frequencies of it, her middle frequencies being louder than her higher ones in some vowels.(especially her “i” in the G4 pitch) To fix this problem when mixing you should limit her middle frequencies to like 2/3 of the initial value. (we recommend using Maximus in Fl studio cuz it has a neat interface where you can see the waveforms)

Meiji MIX-O-METER:

Meiji TONE-O-METER:

Resampler wise fresamp0.11 should be the best since it gives her a little bit more strength making her range a little bit wider and also compressing her voice.The B0F0 flags can be used to increase this compression. Despite her site putting fresamp0.14 as the reccomended we choose fresamp11 because it has less glitches and works better on more users than fresamp14.

But we personally recommend vs4u if you are a tiny little bit better at mixing since this resampler makes her handle pitchbends better(especially vibratos)but at the cost of her samples having a little bit more volume problems.

Meiji Hunter is better used on pop, pop-rock and techno songs, Deco and Giga are producers that represent well the things where Meiji really shines.

She can’t take songs with alot of emotion because of her natural sass. Songs where the range jumps too low or alternates between very high and low notes might be a problem to her too.(kaikiri bear songs tend to do that)

PERSONAL COMMENTS ON THE VB:

Ramus N:

Personally I think Meiji might not have the best recording or the most well made OTO. However she is a really nice bank with a tone very characteristic, anyone wanting to use Meiji should, she is a challenging UTAU to use well, however she is definitely one that also started a trend in between Voicebanks, even Vocaloid being the inspiration to the Luka V4 EVEC system. So, I think she deserves a seal of approval even if set back by some technical details. 

Nick:

Meiji is the vb that made me an UTAU user 2 years ago to begin with.She is an UTAU everyone should use atleast once cuz her tone is really unique in this fandom and most importantly you can tell her character by only hearing her voice alone which is alot for an computer generated voice. She is kinda clunky to work with at first but now i find her a pleasure to use especially in the songs she can tackle with perfection.A tip from me is that if you want her to have a wider range i recommend adding her C5(from fantasy) and B4(from lilith) so if her G4 isn’t powerful enough you can replace it with her C5 or if you want a cool little falsetto or weak note in the high range you can replace it with her B4.

Example of usages:

Usage by Aoka on YT

Usage by Akem on SC

Proof of usage and experience with the vb by the reviewers:

Usage by Nick on SC

Usage by Ramus N on SC

#UTAU #review #gahata meiji