1

Im searching for resources and ways to do this on the internet, but I just can find any links.

The only place I can find that can do this is THIS LINK where you can enter a word and it gives you the phonemic spelling of that word in SAMPA.

This guy posted something in the line of xsampa, but doesn't give any clue on how he went about getting those classes etc. Do you know if this is even possible in C#, and where I can find resources about this?

How your translation SHOULD look like to work with JULIUS.

Example:

abatement   @ b @i t m @ n t    1   0
abates  @ b @i t s  1   0
abating @ b @i t i N    1   0
abattoir    { b @ t w A: r\ 1   0
abattoirs   { b @ t w A: z  1   0
abbess  { b E s 1   0
abbesses    { b E s @ s 1   0
abbey   { b i   1   0
abbeys  { b i z 1   0
abbot   { b @ t 1   0
abbots  { b @ t s   1   0
abbreviate  @ b r\ i: v i @i t  1   0
abbreviated @ b r\ i: v i @i t @ d  1   0
abbreviates @ b r\ i: v i @i t s    1   0
abbreviating    @ b r\ i: v i @i t i N  1   0
abbreviation    @ b r\ i: v i @i S @ n  1   0
abbreviations   @ b r\ i: v i @i S @ n z    1   0
abdicate    { b d i k @i t  1   0
abdicated   { b d i k @i t @ d  1   0
abdicates   { b d i k @i t s    1   0

Should one first convert it to Sampa then replace the words to X-sampa compatible with Juluis, or Directly use some mapping to convert it to X-Sampa

Community
  • 1
  • 1
Ruan
  • 3,969
  • 10
  • 60
  • 87

1 Answers1

2

It sounds like you need a dictionary of English pronunciations. One commonly-used one is the CMU Pronouncing Dictionary (Wikipedia), which lists the pronunciations of 133746 words. It is freely available (public domain) as a plain-text file.

It uses its own weird phonetic transcription, but it won't be difficult to replace each phoneme with the X-SAMPA equivalent, provided you don't care about lexical stress. Example:

stack overflow
S T AE K . OW V ER F L OW2 .

can be converted to X-SAMPA

/st{k Owvr=flOw/

using a lookup table for each phoneme.

The dictionary doesn't contain all English words, but there is also a tool to guess the pronunciation of words not in the dictionary.

Mechanical snail
  • 29,755
  • 14
  • 88
  • 113
  • That sounds like a plan, but where would i get the lookup table for X-sampa? And also, i need to replace some characters in the X-sampa converted words because the program i use (Julius) cant handle all the characters of X-sampa (and yet it uses X-sampa, don't ask me why..), so ill need to run them through 2 lookup tables.. – Ruan Oct 29 '12 at 05:50
  • @Ruan: I don't know of any pre-made table, but you can make one by cross-referencing https://en.wikipedia.org/wiki/Arpabet with https://en.wikipedia.org/wiki/X-SAMPA, and using only characters that Julius understands. – Mechanical snail Oct 29 '12 at 07:31
  • The reason "Julius" uses X-SAMPA, but not all of it is that X-SAMPA is used to describe phonetics for all languages of the world, whereas I suspect that "Julius" is only for English pronunciation. CHEERS – happy coder Jan 12 '13 at 18:10