Search This Blog

Tuesday, December 20, 2011

The Future of Music

Hatsune Miku - Virtual Singing Star
In my lifetime its hard to believe where music and audio have gone.

I was born in the late 1950's.  At that time music was recorded directly from human performances - often on acetate disks or single track, mono tape systems.  There were no computers, multi-tracks, or other devices.

People fell in love with the sounds of voices: Elvis, Sinatra, and so on.

But today all of this is different and, no doubt, will change more rapidly than anyone can imagine.

For one, there is now something called Vocaloid2 - created by Yamaha.  Vocaloid2 is a system for creating artificial singing.  It can be loaded with special sounds created by a singer to create a full singing voice.  This was developed around 2004 but really wasn't up to the job until recently.

Vocaloid2 is even rumored to support the extraction of these syllable.  The voice of a Japanese singer, Hitoshi Ueki who died in 2007, was recently used to create a Vocaloid2 library.

Today the Vocaloid site boasts links to Vocaloid3 (see this).

Below is a Vocaloid2 English version of "Amazing Grace" - note that the song is sung with a Japanese accent. 

This singer does not exist.

And while certainly not an Americal Idol-ready performance its hard to imagine that this won't find its way into music, like commercials, very, very soon where production cost is an overriding concern.

Here's another Japanese anime star "Hatsune Miku" - who is not real - singing a song on YouTube:

Again, she's not real, but she has lots of fans - some who do not believe she is not real.

I can still recall "Who Framed Rodger Rabbit" which was a modern mix of real and animated characters.  This was the forerunner of much of the virtual movie effects you see today (and sadly, the original Disney technology for this such as seen in "Mary Poppins" is rarely remembered today).

Very soon plug-ins for music software will allow producers to create singing voices that will sell in the commercial market.

There will be no longer be a need for singers.

Virtually all other instruments are synthetic today - guitars, drums, string sections, pianos - so why not singers.

Recently I was reading several threads on various music forums.  Many "professional musicians" claiming that they are being squeezed out of the business in various ways:  ASCAP/BMI requiring local bars to pay fees which ultimately mean the bar stops having live music, various pay issues related to "union scale" work, newcomers "undercutting" established acts, and so on.

These virtual singers require no pay at all and will work tirelessly for free forever.

Hatsune Miku doesn't care how hard she works.

She's just anime.

According to this Wired article soon this technology will allow Elvis and Sinatra to do a duo of a song written after both of their deaths.

No doubt fans will love it.

Sure it will be creepy... but hey, its Elvis and Frank!

Another article (here) shows how "hits" in the UK are being analyzed by computer to determine what makes them a hit.  Using this type of knowledge Hatsune Miku's creators will be able to make her songs more like "hits."

The article also describes MRI-based scans of children that show how their brains react to potential hits in an identifiable way - according to the article the children's "...ventral striatum — the brain’s reward region — was predictive of a song’s future sales."

Gives new meaning to the exploitation of children.

The future of music is technology because no one will want to drive to a bar and listen to live music - the risk will be too great:

- Pulled over for a DUI.

- You might not hear the song you want.

- You might get cold or wet.

Instead you can stay home with your 52" 3D LED TV and some nice hash or liquor and watch virtual singers do exactly what you like.

These singing voices will make their way into games as well so not only can you get these singers to sing songs you like but to perform them as you see fit.

I can see the PS/4 "Rock Band" version of this now autotuning your voice and changing it so that you can sing duets with yourself or sing like Elvis.

Turn the knob - now you're Sinatra - now you're Tom Petty - now you're Patsy Cline...  (Small syllables of peoples voices are potentially not copyrightable as long as their lengths are short or they come from the public domain.)

Is that Elvis?

Is it so close no one can tell?

Elvis never sang that song...  it can't be copyright infringement.

Pretty scary... huh?

No comments:

Post a Comment