mjohnson/espeak-ng
1
0
Fork 0
mirror of https://github.com/rhasspy/espeak-ng.git synced 2026-10-31 05:35:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs: describe the Kirshenbaum (ASCII-IPA) transcription scheme.

Browse Source
This commit is contained in:
Reece H. Dunn
2018-03-17 17:27:44 +00:00
parent 1fa7c141f9
commit dc67d294c2
2 changed files with 216 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/phonemes.md
+20 -4
View File
@@ -1,5 +1,6 @@
# Phonemes
- [Phoneme Transcription Schemes](#phoneme-transcription-schemes)
- [Consonants](#consonants)
- [Other Symbols](#other-symbols)
- [Gemination](#gemination)
@@ -54,6 +55,20 @@ Instead, it is designed to be a transcription guide for authors of espeak-ng
languages and voices on how to specify phonemes so that the IPA and feature
transcriptions are consistent.
## Phoneme Transcription Schemes
| BCP47 Variant | Abbreviation | Transcription Scheme | Encoding |
|---------------------------------------|--------------|------------------------------------------------------|----------|
| `foncxs` | CXS | Conlang X-SAMPA | ASCII |
| `fonipa` | IPA | International Phonetic Alphabet | Unicode |
| [`fonkirsh`](phonemes/kirshenbaum.md) | | Kirshenbaum (ASCII-IPA) | ASCII |
| `fonxsampa` | X-SAMPA | Extended Speech Assessment Methods Phonetic Alphabet | ASCII |
| `fonzsampa` | Z-SAMPA | Zompist Bulletin Board (ZBB) SAMPA | ASCII |
1. `foncxs`, `fonkirsh`, and `fonzsampa` are private use extensions defined in
the [bcp47-extensions](https://raw.githubusercontent.com/espeak-ng/bcp47-data/master/bcp47-extensions)
file.
## Consonants
| | `blb` | | `lbd` | | `dnt` | | `alv` | | `pla` | | `rfx` | | `alp` | | `pal` | | `vel` | | `uvl` | | `phr` | | `glt` | |
@@ -189,10 +204,6 @@ consonants, but Wikipedia does not. This model uses the Wikipedia descriptions.
| `sml` | æ | | ɐ | | | |
| `low` | a | ɶ | | | ɑ | ɒ |
__NOTE:__ The `smh` vowels are more `cnt` than the other vowels. However, this
distinction is not needed to classify these vowels, so is not included in the
above table.
### Other Symbols
| Symbol | Features |
@@ -220,6 +231,11 @@ above table.
| `cnt` | center | `kirshenbaum` |
| `bck` | back | `kirshenbaum` |
__NOTE:__ The `smh` vowels are more `cnt` than the other vowels. However, this
distinction is not needed to classify these vowels so there are no features
for front-central and back-central. The `fnt` and `bck` features are used
instead.
### Rounding
| Feature | Name | Origin |
docs/phonemes/kirshenbaum.md
+196
View File
@@ -0,0 +1,196 @@
# Kirshenbaum (ASCII-IPA) Transcription Scheme
- [Phoneme Transcription Schemes](#phoneme-transcription-schemes)
- [Consonants](#consonants)
- [Other Symbols](#other-symbols)
- [Vowels](#vowels)
- [Other Symbols](#other-symbols-1)
- [Diacritics](#diacritics)
- [References](#references)
----------
The Kirshenbaum (ASCII-IPA) transcription scheme was created by Evan
Kirshenbaum and others between 1991 and 1993 for use in Usenet and e-mail.
Since then, the support for Unicode has greatly improved. An ASCII encoding
of IPA is still useful as it is often easier to type.
The Kirshenbaum transcription scheme defines a set of phoneme features that are
described in eSpeak NG in the [phonemes](phonemes.md) document. The `cainteoir`
and `espeak-ng` phoneme features defined in that document are not useable in
transcriptions using the Kirshenbaum transcription scheme. They are used here
where relevant to define the Kirshenbaum phonemes in a consistent way across
the different transcription schemes.
Broad (phonemic) transcriptions are enclosed in `/` characters (`/p/`), while
narrow (phonetic) transcriptions are wrapped in `[` and `]` characters (`[p]`).
A phoneme can be transcribed as a named phoneme (`p`), or as a set of features
wrapped in `{` and `}` characters (`{vls,blb,stp}`). A feature (typically a
diacritic) of a phoneme can be added or modified by wrapping the feature in `<`
and `>` characters after the base phoneme (`p<dnt>`).
## Phoneme Transcription Schemes
| BCP47 Variant | Abbreviation | Transcription Scheme | Encoding |
|----------------------------|--------------|------------------------------------------------------|----------|
| `foncxs` | CXS | Conlang X-SAMPA | ASCII |
| [`fonipa`](../phonemes.md) | IPA | International Phonetic Alphabet | Unicode |
| `fonkirsh` | | Kirshenbaum (ASCII-IPA) | ASCII |
| `fonxsampa` | X-SAMPA | Extended Speech Assessment Methods Phonetic Alphabet | ASCII |
| `fonzsampa` | Z-SAMPA | Zompist Bulletin Board (ZBB) SAMPA | ASCII |
1. `foncxs`, `fonkirsh`, and `fonzsampa` are private use extensions defined in
the [bcp47-extensions](https://raw.githubusercontent.com/espeak-ng/bcp47-data/master/bcp47-extensions)
file.
## Consonants
| | `blb` | | `lbd` | | `dnt` | | `alv` | | `pla` | | `rfx` | | `alp` | | `pal` | | `vel` | | `uvl` | | `phr` | | `glt` | |
|-----------------------------|----------|----------|-------|----------|-----------|-----------|----------|----------|-------|-------|-------|-------|-------|-------|----------|----------|----------|----------|----------|----------|-------|----------|-------|--------|
| | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` |
| `nas` | | `m` | | `M` | | `n[` | | `n` | | | | `n.` | | | | `n^` | | `N` | | `n"` | | | | |
| `stp` | `p` | `b` | | | `t[` | `d[` | `t` | `d` | | | `t.` | `d.` | | | `c` | `J` | `k` | `g` | `q` | `G` | | | `?` | |
| `sib`&#xA0;`afr` | | | | | | | | | | | | | | | | | | | | | | | | |
| `afr` | | | | | | | | | | | | | | | | | | | | | | | | |
| `lat`&#xA0;`afr` | | | | | | | | | | | | | | | | | | | | | | | | |
| `sib`&#xA0;`frc` | | | | | | | `s` | `z` | `S` | `Z` | `s.` | `z.` | | | | | | | | | | | | |
| `frc` | `P` | `B` | `f` | `v` | `T` | `D` | | | | | | | | | `C` | `C<vcd>` | `x` | `Q` | `X` | `g"` | `H` | `H<vcd>` | `h` | `h<?>` |
| `lat`&#xA0;`frc` | | | | | | | `s<lat>` | `z<lat>` | | | | | | | | | | | | | | | | |
| `apr` | | | | `r<lbd>` | | `r[` | | `r` | | | | `r.` | | | | `j` | | `j<vel>` | | `g"` | | | | |
| `lat`&#xA0;`apr` | | | | | | `l[` | | `l` | | | | `l.` | | | | `l^` | | `L` | | | | | | |
| `flp` | | | | | | | | `*` | | | | `*.` | | | | | | | | | | | | |
| `lat`&#xA0;`flp` | | | | | | | | `*<lat>` | | | | | | | | | | | | | | | | |
| `trl` | | `b<trl>` | | | | | | `r<trl>` | | | | | | | | | | | | `r"` | | | | |
| `clk` | `p!` | | | | `t!` | | `c!` | | | | | | | | `c!` | | `k!` | | | | | | | |
| `lat`&#xA0;`clk` | | | | | | | `l!` | | | | | | | | | | | | | | | | | |
| `imp` | | `` b` `` | | | | `` d` `` | | `` d` `` | | | | | | | | `` J` `` | | `` g` `` | | `` G` `` | | | | |
| `ejc` | `` p` `` | | | | `` t[` `` | | `` t` `` | | | | | | | | `` c` `` | | `` k` `` | | `` q` `` | | | | | |
| `ejc`&#xA0;`frc` | | | | | | | | | | | | | | | | | | | | | | | | |
| `lat`&#xA0;`ejc`&#xA0;`frc` | | | | | | | | | | | | | | | | | | | | | | | | |
1. `h` is transcribed as `{glt,apr}` in ASCII-IPA. It is transcribed as
`{glt,frc}` in this document to match the assignment used in the
Wikipedia IPA table, and the placement in other documents in eSpeak NG.
The IPA chart places them as being `{frc}` or `{apr}`, as does the main
Wikipedia page for the `[h]` phoneme.
1. The `` q` `` phoneme is incorrectly placed in the segment table (Appendix D)
of the ASCII-IPA document. According to the diacritics list (Appendix C),
the `` ` `` character denotes an `{ejc}` phoneme for a `{vls}` base
phoneme, not `{imp}` which is used for `{vcd}` base phonemes. The table
above uses the correct assignment.
1. The `L` phoneme has an alternate meaning in ASCII-IPA. This document only
describes the variant preferred by Evan Kirshenbaum.
### Other Symbols
| | `bld` | | `alv` | | `pla` | | `pal` | | `lbv` | | `vel` | |
|-----------------------------|-------|-------|-------|-------|-------|-------|-------|-------|----------|----------|----------|-------|
| | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` | `vls` | `vcd` |
| `nas` | | | | | | | | | | `n<lbv>` | | |
| `stp` | | | | | | | | | `t<lbv>` | `d<lbv>` | | |
| `afr` | | | | | | | | | | | | |
| `vzd`&#xA0;`frc` | | | | | | | | | | | | |
| `ptr`&#xA0;`apr` | | | | | | | | | | | `w<vls>` | `w` |
| `fzd`&#xA0;`lat`&#xA0;`apr` | | | | | | | | | | | | |
1. The `w` phoneme is transcribed as being `{vcd,lbv,frc}` or `{lbv,apr}`, and
`w<vls>` as `{vls,lbv,frc}` in ASCII-IPA. They are transcribed as
`{ptr,vel,apr}` in this document to match their definition as labialized
velar approximants in the IPA chart.
## Vowels
| | `fnt` | | `cnt` | | `bck` | |
|--------|-------|-------|----------|-------|-------|-------|
| | `unr` | `rnd` | `unr` | `rnd` | `unr` | `rnd` |
| `hgh` | `i` | `y` | `i"` | `u"` | `u-` | `u` |
| `smh` | `I` | `I.` | | | | `U` |
| `umd` | `e` | `Y` | `@<umd>` | | `o-` | `o` |
| `mid` | | | `@` | `@.` | | |
| `lmd` | `E` | `W` | `V"` | `O"` | `V` | `O` |
| `sml` | `&` | | | | | |
| `low` | `a` | `a.` | | | `A` | `A.` |
1. The `&` phoneme is transcribed as being `{low,fnt,unr,vwl}`, and `a` as
`{low,cnt,unr,vwl}` in ASCII-IPA. This does not match their placement in
the IPA chart according to the Unicode characters those ASCII-IPA
phonemes map to. As such, the table above reflects their IPA chart
positions.
1. The `&.` phoneme is defined in the segment table (Appendix D) of the
ASCII-IPA definition. According to the IPA charts, this is not a valid
phoneme. As such, it is not listed in the table above.
1. The `W` and `Y` phonemes have an alternate meaning in ASCII-IPA. This
document only describes the variant preferred by Evan Kirshenbaum.
### Other Symbols
| Symbol | Features |
|----------|-------------------------------|
| `R` | `unr` `mid` `cnt` `rzd` `vwl` |
| `R<umd>` | `unr` `lmd` `cnt` `rzd` `vwl` |
1. The `R<umd>` phoneme is transcribed as being `{umd,cnt,unr,rzd,vwl}` in
ASCII-IPA. However, the IPA phoneme it is representing is based on the
`{lmd}` vowel. The table above keeps the `<umd>` modifying feature to be
compatible with ASCII-IPA, but assigns it to `{lmd}` in this case to be
compatible with the IPA chart.
1. The `R` phoneme has an alternate meaning in ASCII-IPA. This document only
describes the variant preferred by Evan Kirshenbaum.
## Diacritics
| Symbol | Feature |
|--------|---------|
| `:` | `{lng}` |
| `!` | `{clk}` |
| `[` | `{dnt}` |
| `;` | `{pzd}` |
| `^` | `{pal}` |
| `<H>` | `{fzd}` |
| `<h>` | `{asp}` |
| `<o>` | `{unx}` |
| `<r>` | `{rzd}` |
| `<w>` | `{ptr}` |
| `<?>` | `{brv}` |
1. The `<o>` diacritic has an alternate meaning in ASCII-IPA. This document
only describes the variant preferred by Evan Kirshenbaum.
1. The `{ptr}` (protruded) feature is specified as `{lzd}` (labialized) in
ASCII-IPA.
1. The `{brv}` (breathy voice) feature is specified as `{mrm}` (murmured)
in ASCII-IPA.
The ASCII-IPA transcription scheme defines several diacritics that vary in
meaning depending on whether the base phoneme is a vowel or consonant.
| Symbol | Vowel | Consonant |
|--------|---------|-----------|
| `~` | `{nzd}` | `{vzd}` |
| `-` | `{unr}` | `{syl}` |
| `.` | `{rnd}` | `{rfx}` |
| `"` | `{cnt}` | `{uvl}` |
The ASCII-IPA transcription scheme defines several diacritics that vary in
meaning depending on the voicing of the base phoneme.
| Symbol | Voiceless | Voiced |
|---------|-----------|---------|
| `` ` `` | `{ejc}` | `{imp}` |
## References
1. <a name="ref1"></a> Kirshenbaum, Evan,
[Representing IPA phonetics in ASCII](https://web.archive.org/web/20160419125856/http://www.kirshenbaum.net/IPA/faq.html) (HTML). 04-01-1993.
Updated 30-04-2003.
2. <a name="ref2"></a> Kirshenbaum, Evan,
[Representing IPA phonetics in ASCII](https://web.archive.org/web/20160419125856/http://www.kirshenbaum.net/IPA/ascii-ipa.pdf) (PDF). 22-02-2001.
Updated 06-09-2011.