Who needs a choir?

Date: September 2, 2010

Who needs a choir anymore? RequiemLight sounds just like one:

Requiem Light is based on a variety of new recording- and sampling techniques some which have never before been integrated in choral sample libraries. The library can literally sound like a real choir and includes both full choir (SATB), divisi groups (males/females) and two solo singers (soprano and tenor) - all recorded in 24 bit / 44.1khz with 1 microphone position (pre-mixed from Requiem Pro (5.1) mics. The library contains true (polyphonic) legato intervals from piano to forte, sustains from piano to forte, staccato with repetitions, ultra forte marcato, BPM (host-synced) legato based latin chants recorded at different speeds across entire interval of choir, tons of choral effects including whispers, shouts, clusters, demonic chants, sweeps, consonants without tone, claps, snaps and over +50 different types of effects.

See more here.

Reply >>

Craig Jennings on September 2, 2010 2:58am

Is there a function that allows them to replace a lyric ensemble? The demos were background / sonority sort of things. Anything that had text had pretty specific text.

Stephen Bigger on September 2, 2010 10:11pm

There is also some software called "symphonic choir" which allows input of specific unique text with a variety of styles, ensemble size, and other choices. I am thinking of using that for rehearsal tracks for my choir. Still have to research that so please don't think this is an endorsement.

Edward Palmer on September 3, 2010 11:18am

Some have complained that some of the music presented to choir directors for consideration is not acceptable and one might even call it "faux" or fake.

So there should be a welcome from the purveyors for the opportunity to hear their work done by a "faux" or fake choir. Why Not?!

Seriously, it does sound like a fun thing and a good tool.

But is this choracide?

Joshua Oppenheim on September 3, 2010 7:48pm

Interesting I listened to all the samples. I'm quite sure many of us have already heard this type of "software singing" in visual media before. It's somewhat impressive in terms of what it is able to accomplish. It actually works somewhat more effectively in full instrumental textures, but listening to the unaccompanied stuff, I'd say it's pretty obvious that it's not a real choir. Then again, the average person might have a harder time discerning that, but I'd still say it's an eventual conclusion. Two big giveaways for me:

1. There doesn't seem to be any computational value able to generate the varying degrees of timbre changes resulting from pitch range.

2. Too perfect. While we all rehearse to achieve high degrees of "sameness" and "accuracy," something seems a bit inhuman when it's actually achieved to such a consistent degree.

Interesting.

Joshua Oppenheim on September 3, 2010 7:53pm

After finding some more samples (what am I doing on a Friday night) I see that they've intentionally tried to create some "human-ness" by adding some intentional "mistakes" such as entrances/releases in the diction.

Reply >>