LaoDing Posted January 9, 2016 at 11:48 AM Report Share Posted January 9, 2016 at 11:48 AM Anyone familiar with Praat? It's an amazing linguistics tool, and best of all it's free (search it). It's great for isolating tones and seeing tone variation according to acoustics, and it can be used with audio files. The only bother is I'd like to isolate the pitch contour lines from the backdrop spectrograms in order to get nice, clean pictures. Can this be done? Quote Link to comment Share on other sites More sharing options...
Flickserve Posted January 10, 2016 at 02:02 AM Report Share Posted January 10, 2016 at 02:02 AM Yes. There is a previous thread on it that describes its use in detail for Mandarin. Somebody posted some jpegs without the spectrogram so it can be done. Quote Link to comment Share on other sites More sharing options...
Yadang Posted January 10, 2016 at 07:17 AM Report Share Posted January 10, 2016 at 07:17 AM Here is the thread Flickserve mentioned. In that thread, this link to an article in the blog Sinosplice is mentioned, which talks about how the diagrams in textbooks depicting tone contours aren't really helping students because they're not very accurate when compared to tones in real speech. In the Sinosplice article, this link is given to learn how to set up Praat to see tone contours. 2 Quote Link to comment Share on other sites More sharing options...
LaoDing Posted January 14, 2016 at 04:27 AM Author Report Share Posted January 14, 2016 at 04:27 AM Thanks for the help Yadang, really appreciate it, but I still could not find how you isolate the blue pitch contour from the spectrograph background in the Sinosplice article. You could do it with an image editor, but what a lot of work! I found SpeakGoodChinese2 to be extremely buggy and crash prone on Mac OS X but when launched it runs as 'Praat.' So as a implementation of the Praat program it's obvious that the contours can be isolated quite nicely. But the program, for me, is so unstable that it's practically unusable. And those synthesized voices really have to go! SGC seems to have the capability to load audio and vocabulary lists, but every time I tried it crashed. Pratt is robust and has lots of powerful features for any language. I'll keep looking for a way to isolate the pitch contours. It's no doubt really easy but I'm just missing it. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and select your username and password later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.