PG Music Home Forums General Discussion User Showcase After You've Gone (Synthesizer V)

After You've Gone (Synthesizer V)

After you've gone
Who'll stop to see the sun (See the sun)
Sinking low?

And after you're gone (After you're gone)
Who's left to tend the flowers (Who'll tend the flowers?)
Growing by the garden wall?

After you've gone
And everything has changed (Everything has changed)

I never dreamed you'd ever go
But here I am without you
Lonely as a child

An empty bed, a silent room,
Shoes no one can fill (Shoes no one will fill)
I'd give the world to have you back
No, I never will

Now you're gone
And life goes on (Flowers bloom)
The flowers bloomed without you there
They watched the sun (They watch the sunshine)
Like nothing's changed
One more day

Now you're gone (Now you're gone)
Now you're gone (You're gone)

I've tried to put your things away
But i just can't get started
Never letting go (Never let go)

Your folded clothes upon the bed
Shoes that sit there still (Shoes that wait but no one comes)

I'd give the world to have you back
I know I never will


This started out a _ASPEN.STY (Aspen Songwriter Pop) demo - the "Mary Tyler Moore" add2 sound of the piano got my attention - and went through a bunch of chord changes. I decided to experiment with irregular phrase lengths. I created the melody in BiaB - probably not something I'll repeat, as there's no punch in for MIDI, and the undo feature is erratic at best.

After hearing Shigeki Adachi's excellent mockup of 君の子供 (Translation: Your Child) done with a Vocaloid, I decided to try mocking up the lyrics using Synthesizer V.

I imported the MIDI into the Synthesizer V editor, and started working on the lyrics, and eventually got something I was pretty happy with for the first portion. It occurred to me that I could make the song a before/after mirror, and so the second half was a straightforward write.

Obviously, a strict rhyme scheme wasn't my priority.

The male vocals turned out much better than I expected - I really like the breathy vocal. So I figured that I'd just stick with that for the final version of the song.

The female vocal wasn't quite as successful, but with some tweaking I got it in the ballpark of where I wanted it.

The voice I used - the free Elanore Forte voice - is ostensibly a female voice, but there are options for modifying the gender.

But because Elanore Forte is only recorded on a single pitch, it doesn't fare as well in the upper register as other Synthesizer V voices. When they release a good English voice, I'll be first in line.

Writing the harmonies was fun - pretty much the same process I use in my DAW.

Getting the vocals to be clear was a bit of a challenge, and I'm not entirely happy with it. I may have another go at some places. For example, "Who'll tend the flowers" and "Flowers bloom".

It's always better if you don't have to read the lyric sheet in order to understand.

I added a cello for some variety, and decided to try to keep it simple and finish it up. I tried the multi-riff, but BiaB didn't want to play nice.


As always, comments are welcome!


Technical Details:

~683:Bass, Electric, PopHalfNotesPush Ev 120
1533:Piano, Acoustic, Rhythm PopPowerPushArp Ev 120
2935:Guitar, Acoustic, Rhythm FolkPulse8thsFintan Ev 110
3768:Guitar, 12-String Electric, Rhythm ModernPopBrent Ev 120
1856:Cello, Background PopCountry Ev 085
NashvilleEven8^4-a,b:Sidestick, HiHat

Synthesizer V (vocals)
Neutron 3 Elements
Lurssen Mastering Console

the work you're doing with vocaloids is fascinating to me David. It's different than what I normally hear, so I have to consciously suppress the rush to judgement that "different" is bad somehow. Funny how subjective preferences influence our ability to hear instrinsic value in music.

The thought occurs to me that there is cinematic value in your work. Its just a matter of time before some producer will be developing a project about an artificial life form who is exploring her self expression through art and music, and you'll be one of a handful of composers who has a body of work to submit.

The thought also occurs to me that a synth vocal backed up by analog instruments is kind of the opposite of Moody Blues and other similar groups where synth instruments back up the human voice. It would be kind of amusing to do a moody blues cover song while reversing all the instrument/singer roles.

But it would also be interesting to hear a fully synth production: a vocaloid choir performing alongside an entirely synth orchestra. The post-human players.

Please be patient with us while we wrap our heads around the future.

Then again, there's nothing wrong with saying that you can't get past how robotic and inhuman the voices sound, and you really can't enjoy the song. If that's the case, that is.


There are already producers claiming that music represents the "self-expression" of social media constructs.

I don't think we're that close to self-actualized computer programs composing music to fill some existential void.


Well, you've already got that with Hatsune Miku to some extent.

I remember back in the day talking to a local music store owner who was explaining that Madonna had a doubly-redundant hard drive system because all the vocals were lip-synced to samplers, and they needed to be sure there would be no catastrophic failures.

So people are already accustomed to vocalists performing and not actually singing in concert.

Given the cost-cutting measures that have been taken in more recent years, people are already used to the use of using backing tracks instead of having a band.

I'd have thought that there's no need to even represent the band in a Vocaloid concert.

However, the Vocaloid concerts I've seen on YouTube have all had an actual band present. Now, the band is off in the wings with the Vocaloid in center stage, draped in darkness.

But they're there, which shoots down that idea.

I also noted that there's a lot of sing-along at the concerts.

So Vocaloid concerts might actually be the exact opposite of what I imagined. Instead of people going to watch a concert, they're actually there for a massive sing-along/karaoke with a live band, but with a much more sophisticated bouncing ball.

Or not.

David,

Very simple lyrics and melodies that capture the hearts of listeners.
I've never heard of the vocal by Synthesizer V.
It has a different taste from Yamaha's Vocaloid.
It's very difficult to add expressions to Vocaloid songs, but it's also rewarding.
I also enjoy it all the time.
I thought that this work also had an attractive expression.
Enjoyed my listen a lot.

Best Regards.

Hi David,

I was impressed how you 'created' those vocals with Synthesizer V. I hadn't heard from it, looked it up on the internet and learned a bit more.
That was when I got impressed. When I heard your song I had a strange feeling about that female voice, than I read what you did.
Very interesting. I knew about vocaloid, but didn't like it mutch (till now), but you had to arrange more with the synthesizer V as I understood.
Who knows where that leads to. I work mostly with Melodyne to create the female vocals. That is a lot of work too, but I think less than what you did.
The song itself was not really my taste, however not bad at all!
But I like creativity and that is what you showed for sure!

Hans

David and SV...

A good your-gone-i-can't-let-you-go write. A point that might get missed in the rush to judge the "singer" (Sthee-V?).

I could understand the lyric without reading. I thought there was a decent balance between sounds-like-synth-voice and sounds-real..

The music (production) is really well done. And well mixed. Musical.

However (I think you have said in the past "there's always a 'but'...")...
I could not help but think you should be working more in a "modern music world" - where Sthee-V would be more at home. Your BAND is a great classic singer-songwriter sound. (And very well done, I might add, again). IF your musical bed was made up of more current Pop ideas - claps-and-snaps, synth hits/repeats, echo-y massively delayed guitar arpeggios, etc. - Sthee-V would seem the logical choice for the vocals.

That being said... your work with the Vocaloid is truly astounding. Cannot imagine how much work that must be. You are to be commended.

I enjoyed several listens.

fj

David, I know you’ve been working on synthesized vocals for a long time. They’ve definitely come a long way. The weaving in and out of the harmony vocal was very well conceived. The song is very melodic and arranged well. The mix is excellent and the drums sound especially good. The only thing I find missing is emotion in the vocal and I understand that is something that may never be achieved by a non human voice. Nevertheless this is an extremely creative and innovative piece of art. Bravo,Tom

Thanks, Shigeki!



I think it does a better job that the Vocaloids.

It "basic" version is pretty full featured, and the free version it comes with an English female voice, as well as a Japanese male and female voice.

I've got some Vocaloid voices, but I've only got the "tiny" Vocaloid editor, so that's pretty limiting for me. The "basic" version of Synthesizer V allows one backing tracks and two voices, with no limit on the length. The full version of the editor is half the price of the Vocaloid editor.

One of the features that's coming out this year is neural-network trained automatic tuning. The demo that I've heard has been very impressive, so I'm looking forward to hearing more of that. It's still synthetic, but pretty expressive. For obvious reasons, they're focusing on Asian language voices first.


Thanks! My biggest worry was that having a synthetic voice would detract from the song, but I thought emotionally it was a good match.



Thanks, I'm glad to hear it!

Hi, Hans.

It's basically the same as working with a MIDI piano roll editor. Most of the default values are pretty sensible, so it sounds fairly good out of the box.


No, it's as easy to use as Vocaloid.

With any vocals, I'll go back and micro-edit them in the DAW. I'll slice up all the syllables normalize all the voiced portions, and adjust the plosives and sibilants. This sort of thing is normally handled by compressors and de-essers, I'm just used to doing it manually.

I'm just trying to make the vocals as even and clear as possible. This leveling is can be done automatically in the latest version of Melodyne.

There really aren't problems with the voice. It's more a job of making it work in the mix. In some places, it might boost the sibilant to make it more clear.

For example, there was a spot where - to my ear - "dream" sounded more like "tream", so I played with the volume envelope. There was another spot where the background voice sounded like it was saying "ton" instead of "sun". That turned out to be caused by a guitar strum hitting at the same time, so the fix was to reduce the guitar strum - the vocal was fine.

But I think Synthesizer V is pretty easy to use. The main complaint is that the only English voice is the "basic" version. In addition to being multi-pitch (giving a better upper range), the non-free versions of the voices also feature "alternate" phonemes, which can be helpful.


I suspect it's the other way around.

It's an entirely different product, but Realivox Blue is supposed to be getting a full-phoneme upgrade sometime this year. The recording's already done, Mike just needs to find the time to get the work done. Depending on what your background vocals need to do, this might be a viable option.

The main issue I have is that since the Elanore Forte voice is based on a single pitch sample, it gets more and more synthetic sounding as it goes up in pitch.


I appreciate your honesty! Not all songs appeal to everyone, and there's no need to pretend that they do.

Thanks for taking the time to listen and comment!

I appreciate you mentioning that!


Thanks! I could probably have "tuned" the vocal to have more emotion, but I was trying change things up a bit in my process and focus on getting something done, and not finesse it to death.


Thanks again!


All well there should be!


I fully agree. Honestly, I hadn't really intended the song to showcase the synthetic voice at all. My computer is in the middle of the living room in a busy household, and I wanted to get working on the vocal. I figured (correctly) that using Synthesizer V would be a good way to get the work on the lyrics, and expected that I'd replace the vocals at some point.

Then I started writing the counterpoint lines, and that sort of became integral to the arrangement. That made keeping it all synthetic a better choice. The last time I tried mixing "real" vocals with synthetic vocals I wasn't that happy with the results. So I figured there would be less disparity if I kept it all synthetic.

And I was surprised how much I liked the breathy "male" vocal. I think it's got a bit of a Colin Blunstone vibe.



I don't want to give people that impression.

It's pretty much the same as working with a MIDI piano roll. And it's pretty cool to be able to draw in the notes, assign lyrics, and have them sung! It allowed a lot more freedom in working out the harmony lines.

For someone who's used to "just" singing into a microphone, doing something like this is a whole bunch more work.

But in my house, I have to wait until past midnight to do any recording. So doing vocals on the computer is much simpler - and more practical.

Getting an expressive performance from these programs is what requires real work.


Thanks again! I really appreciate it.

Hi, Tom!

Thanks for the compliment. But it's more that the software has improved.


I pilfered most of the ideas from The Indigo Girls.


Glad to hear it!

As usual, credit goes to my son on the mix. I'll do the arrangement and what I think is the final mix, and then he'll "fix" it. I used a lot of Neutron in this, but he ended up overriding a lot of the automated choices. And while I originally tried using Ozone on the final mix, I decided that Lurssen sounded better for this particular song.

I was worried that the synthetic voice would distract from the song - especially given the lyrics - but I decided to take a chance anyway. Sort of a trial balloon, so to speak.

Live vocals are always preferred, but I won't rule out synthetic voices if that's what it takes to get the song out the door. Unless people really complain, of course!

I was quite intrigued reading your description on how your created your production--sounds a bit complicated.

Wow! I really like the results. The artificial voices didn't bother me at all. I thought they added a subtle surreal dimension to the song. Beautiful melody and harmony arrangement. I don't think the voices would have worked as well without the strong melody.

A+ lyrics.

All in all, an extraordinary composition and production.

I enjoyed the song and the technical and artistic choices made in its creation. It's a very sad piece, and I think the quality of the female vocal expresses well her "absence". Really well done, David.

David:

This is really interesting . . . you’ve done amazing work on creating the synth vocals—very impressive. I wouldn’t know where to even start on something like this. And the harmonies on this are excellent.

The song itself is really pretty—a sweet write. Really enjoyed listening to this—in fact, I listened to it twice. I’d be curious to hear a more traditional version of the tune.

Killer stuff.

My best to you,

Deej

I guess that we're all used to our own workflows.

Once I got the everything set up in BiaB, I was able to import the melody and backing MIDI track into the Synthesizer V editor. I did the vocals there, and then exported them out to my DAW, like I usually do. So it didn't seem that complicated at the time!


I'm glad to hear the vocals didn't get in the way of the song.


Thanks!

Thanks for having a listen!

I appreciate the feedback.

Hi, Deej.

I'm happy with the harmonies as well.

Sometimes things work out much better than the should, and the harmony part were one of them.


You mean with a "real" vocalist? I'd like that as well.

Thanks for the encouragement!

Hi David,

Those synthesized vocals sounds much better than I thought they would to be honest, perhaps a testament to your skills in that area.
The male vocal especially has some nice moments.
Good melody and harmonies, that probably helped them along

An interesting listen, listened twice actually. Enjoyed!

Excellent write, David. I'm still getting used to the synthesized vocals. They have come a long way, and the mail vocal you programmed did sound more realistic, as you pointed out. On this song the vocals didn't get in the way of putting your song across, and that's a good thing

Arrangement is nice. At first I was thinking the positive feel of the arrangement was at odds with the sad lyrics. But then the 2nd half of the song cheered up and makes it work. Sort of has a Carpenters feel (and that's a very good thing.)

Enjoyed this one.

Steve

Man, that's a lot of work, David, and I really really like the result. The song itself is a gem, and you present it in a very engaging way. The vocals sound cool, perhaps a bit too perfect, but you made them work very well. Yes, the female sounds a bit more alien, but it's definitely as cool as singing into a tube hooked to a guitar. I like the effect.