Phonemic problem in making my own dictionary for dope sheet

[Visping]

I would like to use dope sheets when animating my character. But as my native language is not english I would like to make my own dictionary in swedish.

I think I can just replace the dictionary.dic with a swedish dictionary.

But there is a problem: I do not recognize the way the phonemes are written. What phonemic alphabet is used in Animation Master?

I can tell it is not the IPA by International Phonetic Association. Does anybody know of any documentation that explains the way phonemes are to be written in the Dictionary.dic to work properly in Animation Masters dope sheet?

 

phoneme 'f/oU/n/i/m

phonemicist f/oU/'n/i/m/I/s/I/st

phonemicize f/oU/'n/i/m/I/,s/aI/z

phonemics f/@/'n/i/m/I/ks

[Rodney]

Hash Inc uses the Preston Blair phoneme set.

I'm not sure that will fully clue you in to the breakdown in the dictionary file though.

 

Some write ups by Gary Martin discuss the minimum set that you'll want to work with:

 

http://www.garycmartin.com/mouth_shapes.html

 

http://www.garycmartin.com/phoneme_examples.html

 

[Rodney]

According to Martin Hash:

 

There are ".dic" files in other languages. It comes from an American Sign Language website. We've never tried it but we chose this technique because many languages are supposedly available.

 

[Visping]

Thank you Rodney for your reply.

I am familiar with the Preston Blair phoneme set, so that is not the problem.

The problem is that I can not find any similar way of transcribing the pronunciation of words when I search the Internet.....

Animation Masters dictionary presents the words in a list. After each list is a transcribed version of each word of how it is pronounced.

I could not find any other dic-files on the Internet having phonetic transcription after the words at all.

To be able to make my own dictionary I need to understand how to write phonetic transcription from scratch. As you can see from the example below words seemingly sounding alike (first part of the words) has different phonetic transcription. That is why a guide or some sort of documentation would help me very much...

phoneme 'f/oU/n/i/m

phonemicist f/oU/'n/i/m/I/s/I/st

phonemicize f/oU/'n/i/m/I/,s/aI/z

phonemics f/@/'n/i/m/I/ks

[Rodney]

I don't think some of the extra symbols are used currently in A:M but they are useful when trying to actually say the words.

Example:

 

'f versus f

The apostrophe appears to designate which part of the word is emphasized.

This won't show up in the dopesheets usage of the f phoneme though as both 'f and f will be assigned an f phoneme.

 

Similarly, the comma appears to be a lowering of the emphasis when added before a phoneme.

 

The / character separates the phonemes.

 

@ is interpreted by A:M the same as the AI phoneme as is the & character... although they are pronounced differently in spoken language.

 

The underline character appears to be used when joining two words by an underline character.

This tells A:M that they are two words although they occupy the same word entry.

 

In short, if you only use the basic phonemes without the special characters you should be fine (and you can always add those special characters later).

More importantly, every phoneme used in the breakdown must be a phoneme that exists as a pose in your model/character or else A:M wont be able to assign it automatically via the dictionary.

Of course you will still be able to assign any pose that exists in the model via the dropdown menu when selecting the option to 'Add Single Phoneme'.

Words found in the dictionary will be a blue color while those not found will be red.

 

For those looking in it may help to replace the word Phoneme with Pose each time you read it.

Phoneme simply applies to a pose create for the purpose of articulating lipsync.

The poses driven by the dopesheet can be any pose that assigned a wordname.

(although I don't think A:M will accept spaces between words in a wordname, Hence the reason for the underline special character in the dictionary)

[Rodney]

To make this a little bit easier here are the basic phonemes/poses that comprise the Preston Blair set.

Any dictionary defined with a breakdown of these phonemes will animate the basic lipsync 'correctly'.

For additional extension to the set you can create your own but understand that other models without those poses/relationships/phonemes assigned wont drive any poses via the dictionary.

 

So the short of it is that with the basic Preston Blair phoneme set we have 9 basic values to choose from that form all lip positions.