NEW DEMO I JUST MADE![Synth V Choir - A Nightingale Sang in Berkely Square]
Using all four English voices (Solaria, Natalie, Kevin & Asterian) singing in a jazz choir style.
Let me know what you think. I’ve learned a lot in the past few days, and realize there is a lot under the hood of this program.
Simply thank you for sharing! I really love this, congrats Thomas!
Solaria, Natalie, Kevin & Asterian
I bought Kevin, but I didn’t know there were other English singers.
Where can I buy them?
Hi ThomasS, You did a great job!
I’m starting now and I’d like to know how you made it. If you started from a pre-existing midi and then optimized it, or if you entered all single notes on parts. Thank you
As a lover of Manhattan Transfer, that was really an inspirational piece. I have so much work to do to make Synth V sing like that. The arrangement was very nice and your tuning is pretty good. Thank you so much for sharing.
Here’s how I made it:
Reading the sheet music, I played the midi file in my DAW (Cubase) WITHOUT any click track. I tried to make it as free (rubato) as possible to suit the style - leaving pauses for breaths etc. Sometimes I played just the rhythm of the notes, or sometimes the melody and rhythm, and then copied the rhythms of one part to the others, and changed the notes so that all parts had the same rhythm (note on and not off)
I had four tracks (Soprano - Alto - Tenor - Bass) in a midi file which I imported into Synth V. It doesn’t matter what the midi tempo is, but mine was 120, so I made sure that Synth V was also set to 120. It could be anything as long as the midi file and Synth V are the same.
I then copied the tracks to 2 voices for each of the four parts. These are the voices I used:
Soprano - Solaria & Natalie
Alto - Natalie & Solaria
Tenor - Kevin & Natalie
Bass - Asterian & Kevin
I changed the tone of each voice to suit the part and be different when the same singer was used twice. So Solaria had a deeper sound when she sang alto, and Natalie a deeper sound when she sang tenor, and Kevin a deeper sound when he sang bass, etc. There are actually 3 Natalie’s but each is different, brighter for Soprano, normal for Alto, and deeper for Tenor.
I then tweaked the vibrato so it wasn’t too much, and very little on the strike of notes for about the first half, but kept the slides and approaches to notes, because that is what sounds realistic.
ThomasS, Thank you very much for answering me in such a clear and detailed way!
I would say that you did a great job and the result can be heard. I think it is brilliant that you used 2 singers for each voice. One last cuoriosity: how much time did you spend to make it?
It took me maybe five or six hours, but that is not a fair judge of the program because I was learning how to use it (only a few days after buying it.) I have learned many things since then that would make the job faster.
Using just one singer per part did not sound like a choir, and if I doubled the same singer on the same part it didn’t sound good either. So I finally discovered that each part should be sung by two different singers, and experimented with different tone settings to make them more diverse. Now that I know this, I remember the settings and combinations, so the next one would be faster. Of course, if there were more English voices (at least one more male and one more female) I could make a choir with three singers per part (12 in total instead of 8) which would sound even better. When more English voices are available I will do this.
I experimented with different mixes, and finally settled on panning the voices hard-left and hard-right for each part, so that the difference in tone and character was most obvious. That took some time to discover, which I can do automatically in future.
There are processes that I have now assigned to keyboard shortcuts which will save many hours, that I first did with the mouse. For example going in and out of drawing mode (to manually draw the pitch curve) is now assigned to keys. Also “scale pitch deviation” to reduce the vibrato to the size needed for choirs is now on one key. “Removing short silences” is also on one key, and I select a group of notes in a phrase and set it to maximum to eliminate all gaps between notes quickly. And so on, once you know what processes you will use many times for a job you should assign them to single keys.
I also found a great trick, using Melodyne, and hit the settings “random rhythm variation” (assigned to one key) which very slightly alters the rhythm of notes so that they do not all strike at the same time, which sounds too perfect. I did this a little bit on the demo you heard, but it was hard using the mouse and menu but now I can do it more and faster with a keyboard shortcut.
ThomasS, thank you for sharing your work experience! It’s good to talk about it here.
Your working method impressed me so much that I would like to redo the same song you did, considering that I really like it. I hope to succeed…