[LAU] [LAD] seeking "fresh" way to process/shape human whisper in real-time

[Jeremy Jongepier]

On 12/22/2010 09:05 PM, Atte André Jensen wrote:
> On 2010-12-22 20:57, Julien Claassen wrote:
>> Hello Ico!
>> why not go for praat? Praat offers many voices related functions. You
>> can use the easy, funny bits, like changing speaker/gender or I believe
>> trace different melodies, if there are any to find. You could I think
>> move the formant filters. A very interesting tool to work with voices.
>> It has a GUI, but you can also use a simple script language, if you want
>> to do something clever, based on algorythms. The results are very
>> living, very close to a real voice, if you want them that way. I liked
>> it, because it didn't sound so obviously synthesized.
>
> Sounds like something I *have* to try out! Fired it up, looks scaring,
> seems it's exclusively for offline processing. Couldn't even find a
> "load .wav" anywhere.
>
> Will dig in, though. Thanks for the review!
>

A little OT.

Praat is cool, just for being one of the very few, if not the only open 
source project that is being developed at the faculty where I work. I 
even think it's the only department of my faculty where they use Linux 
on a daily basis (apart from two of my direct colleagues and myself). So 
props to them :)

Enough OT.

Best,

Jeremy