Vocal Synthesis fun: First Steps with Virsyn Cantor
So, Cantor has a.. I’m a little confused about this.. but it seems to be a 30 day or 20 hour trial.. seems I’ve been stuck with the 20 hour one! In any event.. I’ve downloaded and installed said software.. and started playing around with it. So allow me to talk about it….
In the instructions it explains that Cantor is not meant to be a ‘vocalist replacement,’ but a means for the exploration of vocal synthesis.. which is interesting..
Err, I should slow down to explain the back story.. or at least kinda what I’m talking about here… Cantor is a vocal synthesizer. You can find a video demonstration of an older version of Cantor here.. to see for your self. What a vocal synthesizer is is… well.. it creates vocals for you.. the old “singing computer” trick. I’m interested in it.. because I can’t sing.. and the notion of putting vocals in my music is kind of exciting.
How the first steps feel
When you first jump into this here Cantor and its new fangalled vocal synthesis.. once you kinda figure you’re way around the program a little bit (I’m still in the middle of that process) you realize that its time to learn something totally new.. which is to say vocal synthesis. Well.. maybe that’s over stating it a little bit.. but you are jumping into the world of “how do I program a vocal performance.”
The work flow goes something like this:
- Program you’re melody line
- Type in the words you want it to sing
- Arrange “Phonemes” for desired result
- Screw around with volume levels over time.. for each note
- Screw around with “tension levels” for each note over time
- Screw around with pitch bend and vibrato of each note over time
- Screw around with a few other parameters over time
- Concern you self with other elements of the production
That’s more or less how it works.. . I mean if you were to actually achieve a realistic vocal.. it would be via how you program these performance parameters. This basically means you need to start paying attention to the human voice a whole lot more.. and sorta develop your ears.. and go from there. So long story short.. my work features some robotic type singing.
Ok, so what else is there to know about this subject?
Some of you might be wondering what “Phenomes” are. Well.. if you take a look at all languages.. the component sounds are “Phenomes.” When animators go about animating mouths.. they’ll often look at the Phenomes.. So, when you “arrange the phenomes” you’re basically saying.. over a long sustained word sung.. when the vocal should move from one word sound to the next.. got it?
There’s other features in here to.. like the ability to create you’re own voices.. and other stuff I haven’t quite gotten to figuring out.
Ok, what’s it like in practice?
For starters I’m feeling a little like it might be time for them to release a new version.. cause this version has not only been around for a while.. it appears a little clumsy to use. Now I say this without yet fully understanding it.. which isn’t fare.. but so far it’s a bit of a pain in the ass..
You spent a lot of time crafting your vocal.. a phrase or two.. then you want to move on to the next phrase or two.. and it’s just.. kinda clumsy in how you move around..
Still.. it’s way better then what I’ve been messing with thus far.. (all of which has been free tools)… It’s relatively clear that it should integrate with my usual productions pretty well.. much better then the other process I’ve been trying out.. which are all horribly time consuming… And there is something terribly cool about programing a vocal performance.. My verdict on it comes down to.. “I think I’m going to have to play with this a bit more.. and see how it goes.” At the end of this time period.. I should have a better sense of how well it fits into my work.. my process.. and if what I’m now perceiving as flaws.. really are flaws.
After that its just a matter of looking at other similar programs.