How I create Arpasing USTs

How I create Arpasing USTs

This is a basic tutorial on creating usts using the Arpasing method designed by Kanru Hua. I am aware there is already a tutorial on creating usts for this method, and you can refer to that for things I don't explain here, but this is to cover things not necessarily covered in the other tutorial available.
The methods explained will generally revolve around the usage of a tool named "LyricParser (B.15) which can be downloaded here.
The tool requires Java to run.
(For those who can't click the hyperlink)
http://www.mediafire.com/download/nrnaelhi9qlc64p/LyricParser_(B.1.5).zip





I explain two methods for creating USTs that I use below. I rarely use the Arpasing Assistant.

A method I have started using is this: Create a VSQx first. If you are familiar with the Vocaloid software yourself then this shouldn't be the difficult section.
  1. Import vocal midi into Vocaloid
  2. Insert lyrics (Leave the dashes and etc!)
  3. Export as .VSQ (Only .vsq, NOT vsqx)
  4. Import .VSQ into the UTAU application
  5. Export as .UST
We have our .UST at this point. Looks exactly the way it does in Vocaloid with the words and dashes in the notes. At this point we're going to use an external tool called "LyricParser (B.1.5).jar" which I will link as a resource.
Lyric Parser is an amazing tool that will take usts (Especially ones exported into .usts from Vocaloid) and parse them in the VCCV English format with it's dictionary of 5000 words. Don't worry! This won't be a problem in later steps.
  1. Open LyricParser (B.15).jar
  2. Find, locate, and import your .UST
  3. You will see words that the application doesn't recognize and it will ask you to type in syllables, I DO NOT RECOMMEND DOING THIS. Press cancel on every prompt.
  4. Your .UST will export in the folder where LyricParser is.
You now have a UST created in the VCCV method.
Now here's the clarification here; Arpasing is almost exactly like VCCV, but with simple phonetic changes and much smaller. You can replace the VCCV phonemes with Arpasing phonetics. Here is a basic conversion chart for VCCV to Arpasing provided by Kiyoteru: https://arpasing.neocities.org/resources/czampa-arpabet.html

It is at this point that you will most likely come across words that were NOT parsed. Example, the word "crawl" is not in the phonetic dictionary with LyricParser. You will have to manually splice the word "crawl" by clicking and dragging the note with the word back, editing what appears on the rest ahead, and keep doing so.

The word "crawl" would be "[k r] [r aa] [aa l]"

Note: You can use the Arpasing Assistant for notes that don't get parsed into VCCV, but I mean I don't do it. Creates too many notes and things can be off-time.

You have a vocal midi with all the notes aligned correctly but do not have Vocaloid to import it into. This will work just fine but may be a bit less user friendly than with the Vocaloid Editor.
  1. Import your vocal midi into UTAU
  2. Write the words of the song in each corresponding note, but split syllables per note. If you had "helping", you would put "help" in one note, and "ing" in the one after that. Same process for words with 2+ syllables.
  3. Save .UST
I talked about this in the upper procedure, but I'll say it again. There is an application that can parse .usts with words in them into the VCCV style, it's called LyricParser (B.15) and it contains a 5000 word dictionary. We will use this to convert the majority of our new ust into VCCV.
  1. Open LyricParser (B.15).jar
  2. Find, locate, and import your .UST
  3. You will see words that the application doesn't recognize and it will ask you to type in syllables, I DO NOT RECOMMEND DOING THIS, IT WILL NOT EXPORT. Press cancel on every prompt.
  4. Your .UST will export in the folder where LyricParser is.
You now have a UST created in the VCCV method.
Now here's the clarification here; Arpasing is almost exactly like VCCV, but with simple phonetic changes and much smaller. You can replace the VCCV phonemes with Arpasing phonetics. Here is a basic conversion chart for VCCV to Arpasing provided by Kiyoteru: https://arpasing.neocities.org/resources/czampa-arpabet.html

It is at this point that you will most likely come across words that were NOT parsed. Example, the word "crawl" is not in the phonetic dictionary with LyricParser. You will have to manually splice the word "crawl" by clicking and dragging the note with the word back, editing what appears on the rest ahead, and keep doing so.

The word "crawl" would be "[k r] [r aa] [aa l]"

Note: You can use the Arpasing Assistant for notes that don't get parsed into VCCV, but I mean I don't do it. Creates too many notes and things can be off-time.

Besides the nature I take upon creating USTs, lets talk about some other things involving Arpasing.




  • Consonants should last 60 milliseconds. [- k], [t -], [- r]
  • CC (consonant consonant) notes should not last any more than 120 milliseconds, if not exactly.
  • You don't need over 15 duplicates in your oto. [t21], [ah21], etc.
  • In an oto, the overlap almost never needs to be modified (According to my experience!)




This will be updated if I discover new methods, need to make changes and fixes (likely, knowing me), or just need to clarify things.

Special thanks to Mango-Tama.

This guide is in no way perfect or professional. If you want another guide refer to Kiyoteru's tutorial. http://utaforum.net/resources/arpasing-ust-tutorial.281/
Author
Damien
Views
520
First release
Last update
Rating
0.00 star(s) 0 ratings