Erm - I'm really sorry I didn't want to open up a new thread.

 

Could someone merge this with the old "simon goes alpha" thread?

 

-- bedahr 

--- (Edited on 6/26/2008 1:43 am [GMT-0500] by bedahr) ---

Hi bedahr,

No worries, you can leave it here, this is good news!

Here is the link to the original thread: simon goes alpha

Ken

--- (Edited on 6/26/2008 10:35 am [GMT-0400] by kmaclean) ---

Hi Ken!

Ok... Sorry again!

 

@ all German speakers:

The video tutorials are now done and online. You can find them all linked in this wiki article: http://www.cyber-byte.at/wiki/index.php/Videotutorial_DE

Btw: during the making of this video, simon did not crash a single time and the demonstration at the end of the last tutorial is done in a single-shot (so it really recognized the commands without a hitch).

 

@ all non-German speakers

As I'm going on vacation on saturday (until next saturday), I won't be able to translate the videos till I'm back. I'll see that I put the basics on the wiki before I leave. Sorry!

 

-- bedahr 

--- (Edited on 6/27/2008 12:52 am [GMT-0500] by bedahr) ---

As I am on vacation starting this saturday (28. 06.) there will be no technical support (at least from me) until the 5. 07. 
 
Sorry! 
 
-- bedahr

--- (Edited on 6/27/2008 2:09 pm [GMT-0500] by bedahr) ---

I am back home and completely relaxed :)

 

So if any questions arise - please ask and I will gladly try to answer them.

 

-- bedahr

--- (Edited on 7/5/2008 4:10 am [GMT-0500] by bedahr) ---

Hello bedahr,

While you were spending your vacation, I downloaded and installed Simon (screencast). Everything worked fine. I saw that you included a file with the name "simon_C.qm".  Simon was installed in the English language without complications.  Great.

I have a question: I saw that Simon supports SAMPA.  Are there plans to support the IPA directly?

Currently, we are collecting German IPA transcriptions with the dictionary acquisition project.  In the long term, the result will be a German GPL dictionary (SAMPA or IPA) that should work with Simon.

Maybe you (or someone from Simon) is interested in submitting transcriptions?  We could use some help to increase the size of the German pronunciation dictionary as fast as possible.  

Greetings, Ralf

--- (Edited on 2008-07-06 6:32 am [GMT-0500] by ralfherzog) ---

Hi!

 

"While you were spending your vacation, I downloaded and installed Simon (screencast)."

Yes I saw that and I am reading your "Testing simon" blog with huge interest.

I noticed that you are having trouble getting it to work correctly. If you need any help, please feel free to ask me anytime.

 

"I saw that you included a file with the name "simon_C.qm"."

Yes. Thanks again for the bugreport!

 

"I have a question: I saw that Simon supports SAMPA. Are there plans to support the IPA directly?"

Actually simon does not support anything. It does not need to. As the model is generated without adding any externally transcribed data, you are free to use whatever you want for your pronunciation dictionary. You could even invent your own - simon couldn't care less :).

 

The only limitation for this is that the HTK toolkit uses only ASCII (afaik) and we already ran into some problems regarding numbers (HTK phonemes can not be numbers - so for example the standard X-SAMPA phonemes "2" or "6" or "9", etc. are out of the question) .

I guess it wouldn't be to happy about IPA... 

 

But simon already includes a mechanism to transform IPA to SAMPA dictionaries (it uses this translation matrix with custom modifications http://www.theiling.de/ipa/) for the Wiktionary-import (german Wiktionary uses mostly IPA). This would probably work in this scenario too. 

 

At this point let me add that I am extremely honored that simon is even thought off when designing a german dictionary.

This really is a great project of yours and would provide an extremely benficial resource for simon in the long term.

(Actually I presented the project at a conference today and one of the questions was if and why we don't use a (german) GPL dictionary which I had to reply to: "Because there isn't any".)

 

Of course I would like to help but sadly I don't now a thing about how to proberly transcribe words - let alone transcribe with the IPA (I can't even proberly read it). 

But one thing I can do is to promote the project. If this is ok with you I will link to the transcription page from the wiki (I'll have to talk about the team if we could even put something on the homepage), the kde-apps page, etc.

 

Once again: hats off for the effort - I know how hard it is to get an open source initiative about linguistics off the ground.

 

-- bedahr 

--- (Edited on 7/7/2008 1:12 pm [GMT-0500] by bedahr) ---

Hello bedahr,

Why don't you include the files "[...]\julius-4.0.2-win32bin\bin\mkfa.exe" and "[...]\julius-4.0.2-win32bin\bin\dfa_minimize.exe" directly in the Simon distribution?  It is possible that their license would allow that.  But of course, before you do that, you should consider to contact them first.

In my opinion, it would be really good if someone from the Simon team would submit IPA transcriptions on a more or less regularly basis.  

You mentioned one problem that I didn't know about: "the HTK phonemes cannot be numbers".  I will take a look into the HTK book to learn more about that.  I hope that in the long term we will be able to employ the IPA directly without machine-readable mapping.  But this seems to be not so easy.  So obviously it is some kind of workaround to use US-ASCII without numbers.  But to be honest: I would prefer to work directly with the IPA (using UTF-8).

I don't know how to solve all these issues.  But in the end, the German pronunciation lexicon should be compatible with Simon.  I hope that in a few months or so, there will be a German GPL pronunciation dictionary available that can be distributed directly with the Simon package.

Greetings, Ralf

--- (Edited on 2008-07-10 6:13 pm [GMT-0500] by ralfherzog) ---

Hi!

 

"Why don't you include the files[...]"

It is my strong belief that runtime dependencies should almost never be shipped directly with the software.  

When the software has matured enough we might consider some kind of "Quickstart"-Bundle where this approach would be appropriate.

In the mean-time, however, I would find it really strange to force parts of a 3rd party project on the user who wants to download simon. Parts, which we can neither keep current (we don't sync our release cycles to Julius) nor support or maintain.

 

"In my opinion, it would be really good if someone from the Simon team would submit IPA transcriptions on a more or less regularly basis."

I agree. But again, I don't know any IPA. The only other active project members are Franz Stieger (teacher; responsible for organisational stuff and financing) and his son Mathias Stieger (responsible mainly for the homepage) - and neither of those two know any IPA either.

I am sorry but submitting IPA transcription is not much of an option right now.

But again: Would you like me to at least try to promote the project a bit? Maybe we find some people who know the IPA and submit a few transcriptions...

 

"You mentioned one problem that I didn't know about: "the HTK phonemes cannot be numbers". I will take a look into the HTK book to learn more about that.[...]"

Next week, Franz Stieger has a meeting (sadly I can't attend) with a student of the technical university Graz (specifically the speech laboratory) for which I compiled a list of questions. The character set and allowed "signs" for use with the HTK is one of these question. I should know more after the meeting.

 

"I hope that in a few months or so, there will be a German GPL pronunciation dictionary available that can be distributed directly with the Simon package."

That is an ambitious goal! Let's hope it works out!

 

-- bedahr 

--- (Edited on 7/11/2008 12:35 pm [GMT-0500] by bedahr) ---