While we are all awaiting the sure-to-be-exceptional Garritan Choir product, I have found an interim approach that has worked for me. I work with vocal ensembles and need to put practice CDs together so parts can be learned quickly. Here's the approach:
1. Notate my music in Finale including lyrics, with each vocal part on a separate staff.
2. Export the chart as XML.
3. Import the XML chart into Harmony Assistant (with the Virtual Singer addon).
4. Select the type and range of voice I want for each staff.
5. Select US English as the language (there are several others) and, if necessary, tweak some of the phonemes to achieve the syllable sound I want.
6. Place the virtual vocalists on the virtual soundstage to achieve the spacial effect I'm looking for.
7. Export the resulting product as a WAV file and burn the CD.
I wouldn't lip-sync to the resulting track at Carnegie Hall, but it serves my purpose quite well. Some of the available voices are recordings of real singers, others are artiificial. There are MANY facets of the voices that can be configured, including depth of vibrato and pitch change rise times. Using this process I am able to quickly develop a practice CD with the part to be learned panned almost entirely left and the remaining voices on the right channel. My vocalists like being able to learn their notes AND words while driving their car, etc.
I have no financial interest in Harmony Assistant, (to prove it I won't even include the website ) and I'm not related to the fellows in Toulouse who developed this product, I'm just a satisfied user. The product is shareware at less than US$100 and has more capability than I will ever use. If anyone is interested, I can provide a sample.
Just so we're clear, as soon as Gary's choir patches are available for purchase I'm sure I will abandon HA for GPC (Garritan Personal Choir ), but until then, this works for me.