| | 1 | = How to import a FreeTTS Cluster Unit Voice into Mary = |
| | 2 | |
| | 3 | This tutorial assumes that you already have a Festival |
| | 4 | cluster unit voice converted into FreeTTS format. If not |
| | 5 | first build a cluster unit voice with Festvox and convert |
| | 6 | it into FreeTTS format. |
| | 7 | |
| | 8 | Now you are ready to import your voice into Mary. The installation |
| | 9 | directory of Mary is referred to as `Mary_Base`. The directory currently |
| | 10 | containing your voice data is referred to as `old_voice_dir`. |
| | 11 | |
| | 12 | |
| | 13 | 1. '''Where to store the voice files''' |
| | 14 | In Mary, cluster unit voices are filed under `Mary_Base/lib/voices`. |
| | 15 | Make a new directory for your voice data in this directory. |
| | 16 | The directory name and the name of your voice should be |
| | 17 | the same. In this new directory you should |
| | 18 | file all your voice data, except the config file. config files |
| | 19 | are stored in `Mary_Base/conf`. |
| | 20 | |
| | 21 | 2. '''Converting the text file''' |
| | 22 | The FreeTTS tools have constructed a large text file |
| | 23 | containing all the data needed for synthesis. This file |
| | 24 | is normally named `<voicename>.txt` and stored under |
| | 25 | `old_voice_dir/FreeTTS`. For the conversion, this file has |
| | 26 | to have the same name as the directory you created in step 1. |
| | 27 | Copy this file to your newly created directory. Now you have to |
| | 28 | convert it into the Mary format. Open a shell or another command line |
| | 29 | interpreter. Go to the Mary base directory and type |
| | 30 | {{{ |
| | 31 | ant clunit_voice_bin -Dclunit_voice=<voicename> |
| | 32 | }}} |
| | 33 | Mary now converts the text file into bin format and stores it in your |
| | 34 | new voice directory. |
| | 35 | |
| | 36 | 3. '''Copying information about Unit origin''' |
| | 37 | When testing your voice it can be interesting to see where the |
| | 38 | selected units came from. When running Mary in debug mode this |
| | 39 | is shown in the server log. All you have to do for this is copy |
| | 40 | the catalogue file of your voice into the new voice directory. |
| | 41 | Normally, the file is named `<voicename>.catalogue` and you can |
| | 42 | find it under `old_voice_dir/festival/clunits`. |
| | 43 | |
| | 44 | 4. '''Creating example text''' |
| | 45 | If your voice is a general domain voice, you can skip this step. |
| | 46 | For limited domain voices, Mary can show example sentences in the |
| | 47 | Expert Interface. This gives users an overview of what your voice |
| | 48 | can synthesize. To define the example sentences, create a new file |
| | 49 | named `<voicename>.examples` in your new voice directory. The encoding |
| | 50 | should be UTF-8. Write the sentences in this file, one sentence |
| | 51 | per line. You can add comments by beginning a line with `***`. |
| | 52 | |
| | 53 | 5. '''Writing a config file''' |
| | 54 | In order to make Mary aware of your voice, you have to create a |
| | 55 | config file. |
| | 56 | |
| | 57 | 6. '''Phoneme conversion''' |
