Synth V Choir - “Try To Remember”
I recently acquired a few more voices, like Sheena, Hayden, etc, so I thought I might try an acapella choir with 8 different voices singing together:
The different voices have different accents, particularly Sheena and Mai, but there are differences in all of them, and this makes the choir sound fuller. Every voice is using its own AI, which also gives variety to the choir to make it sound like a lot of different people, not double tracking the same few singers.
THAT Thomas - is an exceptional piece of work . . . . is it your composition?
even more kudos if so . . .
Those harmonies are wonderfully “advanced” so I guess your Musicianship says
it all . . .
The Choir is spookily “together” - by which I mean “tight” . . .!
As a long time singer myself, I KNOW that Choirs are not usually able to sing with
such precision - much as they’d like to - so that is the only observation
I would make.
I’m not sure how well “Humanisation” might be done in SVP to loosen it up a
Parametric, I did not write the song, or make the choral arrangement. The song itself was written in 1960, by Tom Jones and Harvey Schmidt, and this choral arrangement (with very interesting harmony) was written by Gene Puerling in 1971. Puerling is considered the master of complex jazz vocal harmony, so I thought I would test out Synthesizer V on something as difficult as this.
Yes, the timing is TOO perfect, as is the perfect intonation of the singers. It sounds very human in voice quality, but a choral singer like you would say that it would be difficult for real singers to do this. However, if you listen to the original recording HERE you will find that the Singers Unlimited came very close to perfection. I modeled the timing of this version by Synthesizer V on their original recording.
It would be easy to “humanize” this so that the timing of the words are not perfectly together, and I have done that in the past on other songs, with great results. The way to do this is to use Melodyne Studio, and use the setting to randomize the timing (and/or pitch) so that the words do not always strike exactly at the same time. I might do this to this track in the future, but this was just a test I made for fun.
Still a nice piece of “work” . . . .
You are clearly a long way down the road of using SVP and getting it to sound
convincing. I would certainly be interested to hear a “humanised” version.
THAT would be the icing on the cake - for sure
I like the sound of this choir ! It might be worth doing some more of those and making an album in that style. One difference I notice from the original, I think your singers are entering words from below the pitch and rising up. It’s quite heavy in places, and with that close harmony might be causing some problems.
Yes, you are right ianwwest. With synthesizer V It would be easy to fix the “scooping” up to notes if I wanted to. Synthesizer V voices are programmed mainly for solo singing, not choral style, so the AI for each (in SING MODE) will automatically do that because that is normal in solos. But if you wanted a cleaner attack you can go into manual mode and draw some of that out. Believe it or not, this version was entirely automatic, I let each singer stay in Sing Mode and let them all do the scooping.
Here are a few more choir tracks I made with Synthesizer V in the past:
A Nightingale Sang in Berkeley Square
London By Night
Somewhere Over the Rainbow
Their Hearts Were Full of Spring
Can’t Help Falling in Love