Hi All,
I've created a tool to generate voice files, using the a variety of TTS providers (such as Ivona.com and Google Translate.)
The tool allows you to automatically generate all the voice files used in ersky9x. It takes around 2-3 minutes to generate the full set (on my internet connection here, it may vary)
The tool downloads MP3 files, and locally converts them to 16kHz, 16 bit WAV files suitable for use in ersky9x. Both MP3 and WAV files are saved.
The tool allows you to open a 'psv' file containing folder, filename and phrase to be spoken. You can also create new files. You can modify the variables on each line, and save the file over the top of the existing file, or with a new name. I have included a sample Voices.psv file that includes all of the phrases in the pack provided by João
You can specify the voice to use (or language for Google Translate). I quite like the Australian accent, so I select that from the drop down box .
If you create a psv file with German text and select German from the drop down box, the tool can create German voice files for example.
Limitations of the tool are:
- Ivona.com voices - There's a restriction of 50000 units a month, where each voice file generally equates to 1 unit. We haven't hit that limit so far, but if we do, then the Ivona voices will stop working for the rest of the month.
- Google Translate - Can't select voice. You're stuck with whatever the default voice is for the selected culture.
- Google Translate - Can't adjust quality, speed or volume.
- Google Translate - Google may kill this at any time. It uses an undocumented API to call their TTS engine, and they may break that without warning. If they do I will endeavour to keep the tool updated to keep it working, but no guarantees.
- The tool will overwrite any existing file in the directory with the same name, without warning
- Select Browse Phrases File (Ctrl-O) to open an existing psv file, or enter new phrases directly into the blank lines
- Select Browse Output Directory (Ctrl-P) to choose where to save the audio files. The files will be saved under two directories under the selected folder; One for MP3, one for WAV, and under that each voice file will be saved with the Folder and Filenames specified
- Add, or modify the folder, filename or phrases in the list of phrases below.
- Select a TTS provider and voice using the drop down boxes.
- Preview a line to hear how that voice will sound.
- Select Go (Ctrl-G) to start the downloading process. As each file is downloaded, the preview button next to it changes to Play, which you can use to listen to the voice from within the tool. Select Stop (Ctrl-H) to stop the download process, or let it run to completion (the progress bar indicates overall progress)
- If you want to modify any of the phrases, edit the phrase in the list, and preview again to hear the change. You will note that the Play button changes to Preview after you edit the line. You can select Go (Ctrl-G) again to have the tool download only those missing files.
- If you have modified your phrases file, or created a new phrases file, you can save the file with the Save Phrases File (Ctrl-S) button, or save the file with a new name using the Save Phrases File As... (Ctrl-A) button.
- You can move lines up and down, by selecting the lines (click on the row header and using the buttons.
- You can add new rows above / below selected lines, or delete lines
Download from GitHub