Here is the pinyin list! The files must be saved with the pinyin names to work with the base oto. Some of the phonemes might actually break filenames, but if you figure out a way to encode the filenames in phonemes and fix the base oto to reflect that, I will happily link to that as an option!
Here is the phoneme list. You might be able to set up OREMO to have it save the names from the pinyin list with the phoneme list as comments, but I was never able to get that to work with my English lists before I found the magic of using words. (I still couldn't get the comments to work, they just weren't needed anymore!)
(Note: I have no idea if I got the syllables like bo, po, mo, fo correct. They're in their own section under 'o', but have the same phonemes as those under -uo.)
I need to make a few edits to the base OTO. Not many, just a few. But I'm waiting until some more people record this list! Why? Because I'm doing this on my phone, but I need to use my computer to access the base oto :P
First, we need a disclaimer. I am functionally monolingual. Even if I was fluent enough in my second language to not be monolingual, Japanese and Mandarin Chinese have almost nothing in common when it comes to actually speaking it.
However, I love phonemes. I studied them a lot as a kid because I wanted to make the most complete UTAUs possible. All of the phonemes used come from this pinyin chart. In this tutorial, pinyin will not be used or mentioned beyond reference charts. I trained myself to read phonemes a decade and a half ago, and reading phonemes was always the easiest way to tackle something you don't have the most knowledge of… it even was the easiest way to do English before my super cool word lists! If you are learning Chinese and can read pinyin, you can stop here and just use the regular list.
The second disclaimer is that hobbies are meant to be fun! If you want your UTAU to sing in Chinese, you don't need to spend weeks fretting over if your ü sounds strange - you can jump right in! No one will send you to jail for accidentally saying b as b. (Even after hours of work, I still accidentally wrote b instead of p several times!) If the accent bothers you, you can always try again. That will let you pinpoint what bothers you specifically and what you want to change. If the accent doesn't bother you, then you just made a Chinese bank that you're happy with… and that's awesome!
And one more thing before we start… Other than changing a few things to make this compact, this theory creates the most compact CVVC Chinese (or CVVChinese) bank possible. However, this is still the same size as a Japanese VCV bank with very similar numbers of configurations. It is a good idea to record a Japanese VCV at least once before recording this to see if you can sit through recording for this long. Not going to lie, my first Japanese VCV took me six hours to record. Now I can do it in about forty minutes! But it's a very tiring forty minutes. Please practice recording by making a Japanese VCV unless you can fluently read pinyin! If you can fluently read pinyin, then it makes no difference whether you do JP VCV or CVVChinese first!
Changes I had to make
This is X-Sampa, except for two things. If there's anything that isn't X-Sampa other than those two things, I made a mistake.
The converter I created has the same changes, so the output can be used without modification with the exception of one thing. (Note: you must strip the text of tone marks so that it's plain Latin characters and surround each syllable with spaces. "Beijing" will return "Beijing", but " Bei jing " will return [p e I_^] [ts/ i N].
The one thing the converter will get wrong is "=". I could not make "=" work as an alias, so I replaced it with "--" in the OTO.
The two things that the converter will get right (but are not correct when compared to X-Sampa) are "\" and "`". The "\" is replaced with "/" because backslash has a tendency to break things. "`" is a little silly, though. I had to replace it with "'" because I could not find ` on my tablet's floating keyboard.
Consonants
Let's start with a chart that shows the pinyin vs the phonemes.
The fact that it looks like there are fifty variations of ts is scary! At least to me, it is. But once you figure out what's happening, it's less scary.
The Ones You Know
I'm sure you noticed a few phonemes that are pretty familiar! m, f, n, l, and s are exactly the same as in English.
There are two more in this section. They are not the same as phonemes in English, but they're close enough that I think anyone without interest in learning Chinese doesn't need to learn the difference. (It's 100% okay to make a Chinese bank just to make plug and play covers for fun, or to make one to share with your friends who do know Chinese!) "u_^" and "i_^" (notice they are lowercase) are almost the same sounds as in "we" and "yes".
These look familiar (aspirated consonants)…
Technically, we have all of the phonemes in this section in English except for once!
These phonemes come in pairs. There's k and k_h, p and p_h, t and t_h, and finally ts and ts_h. (We technically don't have ts as a phoneme in English, but we do use it in things like the word 'cats'!)
I included a chart above, so I can't pretend that k isn't written as g in pinyin. This was a smart choice in pinyin because there is no 'g' in Chinese.
With as scary as Mandarin phonology looks, other than 'ts_h', we have all of the phonemes in English!
The difference between k and k_h is aspiration. In the Chinese banks I've seen, the difference is very stark with ○_h being held out and forceful. Aspiration is when you force a burst of air out when you say the consonant. It is less drastic in English, but we still have that distinction! You can hold your hand over your mouth and feel the difference when you say these pairs:
pore, spore
core, score
tore, store
There is no pair to demonstrate this directly with ts and ts_h, but the theory is the same.
If you are absolutely incapable of saying unaspirated versions of these consonants without putting 's' in front of them (thanks a lot speech therapy), I do have some kind of technical advice. English doesn't have aspirated voiced plosives. So saying 'd' will get you halfway there to 't'. To make d into t, all you need to do is to make your throat stop vibrating (or voicing) while saying it. (Or, if it's more fun and easy, you can just go with 'd' as long as your 't_h' is extra forceful! UTAU is supposed to be fun, not stressful!)
Oh look! Tails!
This section is actually fully affected by my changes to X-Sampa to make it work for me. "'" means palatalization. "`" means the cute little tail. I didn't know how to type "`" with my floating keyboard layout, so I changed it to "'". Just remember! There's no palatalization in Chinese - just little tails that I had to write funny.
The other change was swapping \ for /. Thankfully, I don't know any phonemes that actually use /, so I don't think it is as much of an issue.
Now what do I mean by tails? This is 's'. This is 's' with a cute little tail: 'ʂ'. You can also have 'z' with a cute little tail: 'ʐ'. Those are "s`" and "z`", but banks using my base oto have them as s' and z'. ts' and ts'_h are just s' with a t or a t_h in front of them!
All that said, what does the cute little tail do? It means that it's a retroflex consonant! I don't know what that means at the time of writing. If you want to really learn them, you can start by learning about the unvoiced versions and the Mandarin 'r' sound.
It will not hurt anyone if you just pretend that s' is just the sound in "she" as opposed to a completely different phoneme. "ts'" would be like "cheese", "ts'_h" would be like "more forceful cheese", and "s'" would be "she". I really suggest that you study and all that, but if you're just having fun, why not?! I really, really do suggest trying to get z' right. Not because it's important that you not use the wrong phoneme (like r as in red or Z as in Asia), but because I got it right once while staring at the instructions and felt pretty darn cool.
Now, for the cool cat tails.
ɕ is the IPA version of s/. It looks very cool like a cat or fox looking away from us. That extra long cool cat tail means it is a voiceless alveolo-palatal sibilant fricative. You should probably read this tutorial on it. It's a sound that doesn't exist in English at all. With my untrained American ears, one sounds thicker and one sounds thinner. Thankfully, though, the cute tail and the cool tail are never used for the same finals. Even if you were actually speaking, I would assume people would get your point if you just used the American "sh" for both. (If you were learning Chinese, you need to know the difference, but this is for fun!)
There's s/, ts/, and ts/_h. Again, you could use she, cheese, and forceful cheese and it would probably be close enough!
There's also x.
'x' didn't fit into any other section. It is "h" in pinyin. As a voiceless velar fricative, most of the time places will say "x is the ch in loch" as if all the Americans reading it don't say loch as lock. It should be fine to just use the English 'h', but x is more throaty.
Vowels
When looking at vocal synth phoneme inventories for Chinese, it looks like there's five hundred different vowels! There's obviously aren't five hundred, but the way Chinese lists treat finals as vowels (see Chinese vocaloid phonetics) makes it seem like there's a lot more than there are. Here's a chart:
They're in English!
Luckily, there are plenty of Chinese vowels that are also in English!
a is the same 'a' sound as in sky. e is the same 'e' sound as in may. Why specifically make the distinction that it is the 'e' sound as in may? That's because there is E! E is the same 'e' sound as in bed.
(Note, in English the vowels in "be" and "sue" are long and are very slightly diphthongized at the end. This is for fun, so it doesn't matter if your vowels are long - just a fun fact!)
@ is the schwa, or the uh sound in what. U is "u" in put. u is the "oo" in boo. i is the "ee" in bee. So simply:
a = sky (no i at the end)
e = way (no i at the end)
@ = what
U = book
i = bee (short version)
u = who (short version)
E = bed
Something that seems really strange is that it seems like English VOCALOID phonetics as we understand them are… wrong? The phoneme "A" sounds like the phoneme I believed to be "Q". Q is the rounded version of A, but not once in my life have I ever said caught, cot, or taught as rounded! So uh,
A = cot
O is quite simply put O: in English VOCALOID phonetics, only not long. I'm an American affected by the caught-cot merger, so I don't know how to explain this one. In my English lists, I include O: only so that I can crossfade it into O@. It's worked so far, so I'd say O is like the "oh" in "more".
O = caught in the UK
They're in… German?
I have a section in my spreadsheet marked hard. It has 7, M_^ V, y, y_^, and 9.
ü is a cute happy face, but it's also the German vowel that's pronounced as 'y'! For my fellow Americans, 'y' is the rounded version of 'i' as in 'bee'. When you're saying 'bee', make your lips round and you should have it!
Now that you've figured that out… (don't worry if you haven't, I took German in high school in like 2010 and I still haven't figured out how to make it sound right) it's time to look at y_^. It's y, but it has a _^, just like u_^ above. _^ means something that is usually a vowel is now a consonant. I guess since i_^ is basically the sound in yes, y_^ is the sound in yes with round lips. It's not too common, so I don't think anyone would fault you for using the sound in yes instead of y_^.
y_^ is always followed by 9. That's not a letter! It is only preceded by six initial consonants, so it's not entirely vital to get it perfect… unless you're singing The Moon Represents My Heart. Thankfully, it's not too complex to explain! It's just a rounded version of E (which is as in 'bed')!
7 is also not a letter. Its IPA symbol, ɤ, also looks nothing like a letter. It is a close-mid back unrounded vowel. The only way I can describe it would be to say "U" as in put, then unround it. There's no unround version of "U", so this could be close enough. Make it a little closer by opening your mouth slightly more without moving your tongue. It is always followed by U_^, which is like U, but not a vowel. Listening to the examples on the pinyin chart, it kind of sounds like a more different "uh" or Japanese "o".
M_^ V only shows up together. If you know Japanese phonetics, then you know that M is the Japanese version of "u". If you don't, M is the oo in boo, but unrounded. Because it has _^, it is a consultant instead of a vowel, it's like w… but unrounded. V is actually the same V from VOCALOID phonetics! Therefore, it is the uh sound.
It's R.
So, most places just lie to you. The IPA symbol for American R's isn't r - it's ɹ. It isn't even r in X-SAMPA! It's r\!
Now, I got all excited about this, but the IPA symbols on the pinyin chart disagree with these special i's being ɹ or ɻ. Or r\= and r\`=. (Basically, the equal sign means that it's a full syllable as opposed to just a consonant - the opposite of _^ (the equal is – in the oto file).)
I dismissed the Wikipedia IPA help page for Mandarin at some point for having something the pinyin chart was more correct about. However, I think the pinyin chart is more right about these not literally being syllabic r, but I have no choice because that chart doesn't use a real IPA symbol that I can search to find the correct X-Sampa symbol.
The best thing to do is to listen to native speakers and copy them, but hey. Wikipedia says it's r and cute tail r. What's the difference between retroflex and normal when applied to this phoneme? I really don't know!
There is one more R. This time the pinyin chart agrees it is ɻ, or r\'. (Of course, r/' in the base oto.) This is what it sounds like according to Wikipedia. Originally, this was going to go into the next section as this is not a full final, but instead half of a final.
Final finals
These are pretty easy compared to everything else!
Is n actually n?
In Deadbyte's version of his bank's OTO file, he edited the n that occurs in finals to be something else. He told me a lot of Chinese people don't even know there's a difference between final and initial n. For that reason, I left the converter to use n in the final position also. Especially in this context, I think just pretending all n graphemes are n phonemes.
The other final that is an N grapheme is the "ng" in ring and sing. As a note, for a long time I thought "ring" was pronounced "r I N g", but it's not. It's just a more different n kind of sound.
Aye and Ow
Finally, we have I_^ and U_^ as in boy and how. They're not exactly the same as the English aU, @U, aI, and OI as they are syllabic versions of I and U, but it's pretty close!
I think that's all!
This is a rough draft I wrote over the course of about six hours! I would love feedback and how to improve this guide.
If you want to get deeper, listening to native speakers is always a great idea. But if you just want to get close enough, this should get your voice singing to Mandarin USTs ASAP!
Configuration tutorial and usage tutorial will be written in the near future!
No comments:
Post a Comment