some remarks about transcriptions
User: timobaumann
Date: 6/14/2008 2:49 pm
Views: 10622
Rating: 26

Hi Ralf,

I'm currently processing (and at least partly reviewing) the transcriptions so far acquired. 

I've noticed, that you are often not correctly paying attention to  "Auslautverhärtung". Especially, I have changed all participle forms ("schlafend", "rennend", ...) to not end in /?nd/, but to end with a /t/. Also, I've mostly changed /?n/ to /n?/ in these cases, unless the preceding consonant was already /n/. I think this besser matches how these are pronounced.

"schlafend" as [?la?fn?t]
User: ralfherzog
Date: 6/16/2008 8:39 am
Views: 213
Rating: 49

Hello Timo,

I have just read your remark about the transcription.  In the future, I will transcribe "schlafend" as [?la?fn?t].

I will read the article about "Auslautverhärtung" again.  Please tell me about the (future) mistakes that I submit.  I am planning to submit much more.  And I don't want to make the same mistakes again and again.

Greetings, Ralf 

difference between U+0251 = ? and U+0061 = a
User: ralfherzog
Date: 7/24/2008 9:54 pm
Views: 1624
Rating: 18
I have a question. It is about the German word "vergab".  The proposed transcription is "f???g?p".  But why isn't it "f???gap"?  What is the difference between U+0251 = ? and U+0061 = a ?
Re: difference between U+0251 = ? and U+0061 = a
User: timobaumann
Date: 7/29/2008 8:45 am
Views: 181
Rating: 16

Hi Ralf,

the phonetic difference is, that [a] is a fronted vowel, while [?] is a back vowel. The difference being, that the highest point of the tongue is either in more towards the mouth opening or more towards the throat. Then again, both are open vowels, which means, that the highest point is much lower than for example or [u].

German (and most other languages) do not distinguish [a] and [?] on the phonemic level. This means, that there are no pairs of words that only differ in these vowels. 

Thus, we should not distinguish [a] and [?]. I propose, that we generally use [a] for German words.

Cheers, Timo

"not distinguish [a] and [?]"
User: ralfherzog
Date: 7/29/2008 9:42 am
Views: 306
Rating: 15

1. "not distinguish [a] and [?]" - [a??nfe???tand?n]. Smile And thanks for the explanation.

2. I have another question: It is proposed to transcribe the German word "Gänge" with [gæ??].  What is the difference between the phone [æ] (= U+00E6) and the phone [?] (= U+025B)? Maybe it would be the best decision to drop [æ]?

3. Sometimes, there are several pronunciations possible for the same word.  For example, the German word "einig" does have the pronunciations [?a??n?ç] and [?a??n?k].  Can the dictionary acquisition project handle both pronunciations when I add the same word a second time by using the button "Wort hinzufügen"  in the left corner of the screen?

Re: "not distinguish [a] and [?]"
User: timobaumann
Date: 8/19/2008 9:04 am
Views: 3690
Rating: 14

Sorry for not answering in such a long time, Ralf.

2. Depending on where you're from, there might be a difference between [æ] and [?]. But I agree, that we can just as well drop it.

3. That's exactly the purpose of "Wort hinzufügen", to either add an additional form (for example with verbs) or to enter a second pronunciation for the same word form.

I'll definitely speed up in a few weeks, once I am back to my office.

Cheers, Timo

German dictionary like VoxForgeDict
User: ralfherzog
Date: 8/19/2008 10:32 pm
Views: 161
Rating: 16

Hello Timo,

I hope that you will speed up because there is a lot of work to be done. 

Recently, I was able to import the VoxForgeDict into Simon (after downloading it from the speech corpus).  I would like to import one of your German dictionaries into Simon.  This should work if the lexicon has a similar format like the VoxForgeDict.  So the format should be similar to the following example:

EXAMPLE         [EXAMPLE]       ix g z ae m p ax l
EXAMPLES        [EXAMPLES]      ix g z ae m p ax l z

We should try to find a solution to transform the German dictionary (PLS, IPA, UTF-8) automatically into an SAMPA-like format (ASCII). Currently, I am experimenting with XML/XSLT/XPath.  But I don't know how to transform IPA symbols into SAMPA symbols.  Maybe you could write a Perl script?

Greetings, Ralf