Synthesizing singing
So this is from the late 90s ... http://www.cs.princeton.edu/~prc/SingingSynth.html
Why isn't it filmed? (We can synthesize photorealistic images, but singing synthesis ... still seems to be in very primitive stages.)
What exactly makes it difficult to synthesize chanting?
http://www.interspeech2007.org/Technical/synthesis_of_singing_challenge.php <- still seems primitive.
a source to share
I feel like we're falling into a supernatural valley for sounds easier than images. While our brains perceive poorly formed images relatively well, they do not accept poorly formed sounds unless they appear natural. Anything that doesn't sound perfect sounds creepy, and this creates a very strong barrier to real-world applications. This is good for announcements and phone services, but we are far from fully synthetic singing.
On the other hand, changing the actual voices is done daily, both live and in the studio. Without Autotune, all the "gangstas" and "ladies' gaga" there will be doing work more suited to their real talent.
a source to share