Difference between revisions of "Vowel Synthesis"

From PiWiki
Jump to: navigation, search
(Created page with "Jonas Beskow https://www.youtube.com/watch?v=xv4PlpUTf_c http://www.speech.kth.se/~beskow/papers/fon98veiron.pdf http://www.speech.kth.se/wavesurfer/formant/ <-- can't get thi...")
 
 
(6 intermediate revisions by the same user not shown)
Line 1: Line 1:
Jonas Beskow
+
Jonas Beskow
https://www.youtube.com/watch?v=xv4PlpUTf_c
+
http://www.speech.kth.se/~beskow/papers/fon98veiron.pdf
+
http://www.speech.kth.se/wavesurfer/formant/ <-- can't get this working on OSX because of TCL problems
+
^ source code has 4 formant frequencies for many vowels (>20)
+
  
STRAIGHT
+
https://www.youtube.com/watch?v=xv4PlpUTf_c 
https://www.google.co.uk/webhp?sourceid=chrome-instant&ion=1&espv=2&ie=UTF-8#q=straight%20speech%20synthesis
+
Uses Matlab
+
  
praat
+
http://www.speech.kth.se/~beskow/papers/fon98veiron.pdf 
http://www.fon.hum.uva.nl/praat/manual/Source-filter_synthesis.html
+
  
Csound
+
http://www.speech.kth.se/wavesurfer/formant/ <-- can't get this working on OSX because of TCL problems 
https://en.wikipedia.org/wiki/Csound
+
 
?
+
^ source code has 4 formant frequencies for many vowels (>20) 
 +
 
 +
 
 +
STRAIGHT & WORLD
 +
 
 +
http://indiana.edu/~acoustic/s522/gettingStartedV40_006b.pdf
  
 
http://ml.cs.yamanashi.ac.jp/world/english/
 
http://ml.cs.yamanashi.ac.jp/world/english/
 +
 +
http://ml.cs.yamanashi.ac.jp/straight/english/index.html
 +
 +
https://github.com/Sleepwalking/world-octave -- Octave fast impl (using C++ lib)
 +
 +
https://github.com/sotelo/world.py -- Python2 bindings, which uses https://github.com/r9y9/WORLD -- this looks like a tidied up version of the original
 +
 +
 +
praat 
 +
 +
http://www.fon.hum.uva.nl/praat/manual/Source-filter_synthesis.html 
 +
 +
 +
Csound 
 +
 +
https://en.wikipedia.org/wiki/Csound 
 +
 +
 +
 +
MISC
 +
 +
https://ccrma.stanford.edu/~jos/fp/Formant_Filtering_Example.html
 +
 +
http://www.festvox.org/festival/

Latest revision as of 11:33, 28 October 2015

Jonas Beskow

https://www.youtube.com/watch?v=xv4PlpUTf_c

http://www.speech.kth.se/~beskow/papers/fon98veiron.pdf

http://www.speech.kth.se/wavesurfer/formant/ <-- can't get this working on OSX because of TCL problems

^ source code has 4 formant frequencies for many vowels (>20)


STRAIGHT & WORLD

http://indiana.edu/~acoustic/s522/gettingStartedV40_006b.pdf

http://ml.cs.yamanashi.ac.jp/world/english/

http://ml.cs.yamanashi.ac.jp/straight/english/index.html

https://github.com/Sleepwalking/world-octave -- Octave fast impl (using C++ lib)

https://github.com/sotelo/world.py -- Python2 bindings, which uses https://github.com/r9y9/WORLD -- this looks like a tidied up version of the original


praat

http://www.fon.hum.uva.nl/praat/manual/Source-filter_synthesis.html


Csound

https://en.wikipedia.org/wiki/Csound


MISC

https://ccrma.stanford.edu/~jos/fp/Formant_Filtering_Example.html

http://www.festvox.org/festival/