Skip to main content
July 21, 2010
Question

Is there a reference of Loquendo tags (\item=...)?

  • July 21, 2010
  • 6 replies
  • 8543 views

In the demo site of Loquendo I noticed special tags such as "\item=cough_01" that produce sounds. They work in Captivate, is there a full reference of these sounds available?

Also I noticed that Umlauts (ä, ü etc.) are spoken correctly by the Loquendo demo, but not in Captivate even though its the same narrator. It seems as if Captivate converts these (ae, ue etc. ) before handing them on to the TTS engine?

This topic has been closed for replies.

6 replies

Participant
November 22, 2017

All the phrases are located as gilded GDE files in each installed voice directory of your Loquendo installation folder. (C:\Program Files (x86)\Loquendo\LTTS7\data\voices\)

For example, the phrases of Dave are in C:\Program Files (x86)\Loquendo\LTTS7\data\voices\Dave\DaveGilded.gde

I parsed these dictionary files to JSON and released them on GitHub a couple years back. GitHub - jochemstoel/loquendo-phrases: Loquendo LTTS7 phrases in JSON format.

I understand that this topic is 7 years old but the topic starter was looking for a reference list of 'Loquendo tags' and nobody ever answered the question.

August 31, 2010

The problem with umlauts is solved, this is the solution I received from Ashwin Bharghav (thanks again):

Steps:

1. Launch the speech management dialog and add a TTS

2. Select the German speaker “Stefan” and edit the TTS with the following tag

\@TextEncoding=utf8

3. Add another TTS and Add the text which you wish to convert (the one with umlaut characters) ( See the ScreenShot)

4. Generate speech.

I tested it for both German and French.

September 7, 2010

Regrettably I found this solution to be unreliable. Whenever there are changes to the TTS texts, Captivate tends to loose the coding correction again. I.e., if I edit the TTS text for just single frames and generate the audio new, then the Umlaus are pronounced as nonsense again.

So this is still quite buggy, hopefully the problem is corrected with a patch.

Known Participant
November 2, 2010

Hi,

Does anyone have any updates on this issue? We're now seeing it with German (using the Stefan voice).

Thanks,

Joel

Inspiring
August 30, 2010

Oops! Posted to the wrong thread. Feel free to delete this comment, admins.

August 6, 2010

Hi robert-sfl,

Please have a look at the blog post : http://blogs.adobe.com/captivate/2010/08/editing-text-to-speech-text-for-loquendo-voices-in-captivate-5.html

Thanks

Ashwin Bharghav B

Known Participant
August 8, 2010

Great write-up! Thanks very much for that information--very helpful.

July 21, 2010

Evidently this is a coding issue. Some Umlauts are pronounced as complete nonsense. Perhaps Captivate converts the TTS-text from UTF8 to ANSI when it hands it over to the speech system. And it looks like nobody ever tested non-english TTS before releasing the program.

Tom_Tantalus
Inspiring
July 21, 2010

In both of these topics I am very interested too.

Known Participant
July 22, 2010

The complete list of tags would be great, but a way to customise the dictionary (like you could witht he previous voices) would be even more beneficial!

Any word on how we can customise the dictionary? I read somewhere that the dictionary is held within a lexicon file (*.lex) but I have no way of reading this file...

Help would be greatly appreciated.

July 22, 2010

We saw that the data is stored in a DBL file (e.g. German.dbl) which is an SQLite database format. This list contains pronounciation data for words in alphabetic order.