Jon Stockill wrote:
Those demos are based on festival 1.4 - the prerelease of 2.0 includes a
synthesis module called multisyn, which is a great improvement on the
older modules. http://flightgear.stockill.org.uk/testing/atis.wav
contains the synthesised text of an atis transmission using one of the
Thanks for the clarification and the example - it does indeed sound
already quite usable, particularly taking into account that your example
file did not even make use of any customized language model !
Now, I am really tempted to look for online ATC streams in order to have
festival speak to me even more aviation-like ;-)
I only wonder, how long the actual calculation of a new language model
takes - any ideas ?
Flightgear-devel mailing list