To One In Paradise (Synthesizer V)
Based on the poem by Edgar Allan Poe

Paradise
You were my paradise and all I wanted to find
Island of green in the sea of love
Fountain and my shrine

My days in trances
And in all my nightly dreams
Wherever your grey eye glances
Wherever your footstep leads
Lost in an ethereal dance by eternal streams

I lost you
My life's one light
A dream too bright to last
All that's gone (All that's gone)
It's lost in the past (It's lost, past)
It's lost in the past (Lost in the past)
Living in the past (Living in the past)

Paradise (Paradise)
You were my paradise
And all I wanted is find (You were all I want to find)
Starry hope that was overcast (Hope was overcast)
Dream no longer mine (Dream no longer mine)

No tree shall bloom (Trees shall not bloom)
No eagle soars (Shall not soar)
Silence holds the solemn sea to sandy shore (Language holds the sea to the shore)
Where light shines no more (No more)

All my days in trances (My days in trances)
And all my nightly dreams (And all my nightly dreams)
Follow your glances (Following your glances)
Where your footstep leads (Where your footstep leads)
By eternal streams (By eternal streams)


The Band

Bass, Electric, JazzFunkPoppy Ev16 110
Piano, ElectricVintage, Rhythm BluesyPopMike Ev 100
Guitar, Electric, Rhythm Pop8thsZaneHigh Ev 120
Guitar, Electric, Rhythm Pop8thsZaneLow Ev 120
Organ, Rhythm CountryBluesRockMike Ev 110
Piano, Acoustic, Rhythm Soul60sA-B Ev 110
Sax, Tenor, 2-Beat Rock Ev 110
Organ, Rhythm CountryBluesRockMike Ev 110
Vocal Aahs, Rhythm Pop3-part Ev 085
RealDrums:NashvilleEven16^5-a:Snare, HiHat , b:Snare, Ride


Details

In keeping with my goal of getting songs out the door, all tracks are BiaB and vocals again sung by Synthesizer V.

I though I should do an upbeat ELO-type song, so I worked out a nifty chord progression based around a stepwise bass movement in the piano, and dropped it into BiaB.

I then started looking for an appropriate ELO-ish style. But I liked how the _FLYAMP.STY demo sounded, even though it was more like Aztec Camera than ELO. I rendered a backing track and loaded that up Synthesizer V, where I worked out a melody. The song was a bit short, but I figured I could always add another verse or instrumental if I needed it.

That was the easy part. A couple of pages of dreadful lyrics later, I realized I wasn't getting anywhere. I thought that perhaps I could get some inspiration, so I started checking out ELO song lyrics. I saw this lyric from Strange Magic:

Oh, I'm never gonna be the same again,
now I've seen the way it's got to end

That got me thinking about expulsion from Paradise - one of the classic "that ended badly" stories. Coleridge's poem Kubla Khan and Poe's To One In Paradise seemed like they might have some interesting material, so I grabbed what I thought might be appropriate portions of the poems, and started putting things together.

Line by line, all the bits of Kubla Khan fell away, and with some effort I managed to shoehorn a semblance of Poe's words into the song. I didn't really want to do To One In Paradise since there's already a version by The Alan Parsons Project, but that's where I ended up. Best laid plans, and all that.

As usual, the harmony - the fun part! - was added after everything else was mostly done. This time, I start out without harmony so there was room to build.

I then searched BiaB for additional instruments, with the intent of putting in a short guitar solo. But nothing seemed to fit, so I used a sax instead. Poor sax, not much space for a solo, is there?

I didn't spend a lot of time "tuning" the vocals, but I ended up putting in a lot of work changing words and word lengths to try to make the lyric more understandable. Still, "Silence holds the solemn sea to sandy shore" doesn't exactly come tripping off the tongue.

I did an initial mix, and my son adjusted the mix and further trimmed back the instruments for clarity.

Do the lyrics match the upbeat melody? Not at all.


As usual, any sort of feedback is welcome!

David,

An excellent Pop soundscape.

I'm not familiar with a lot of ELO, so cannot commetn on that. But I will say...the opening chord progression reminded me of "Steal Away" (Robbie Dupree) - one of the ALL TIME GREAT Pop songs...

The mix is excellent. And the vocalist "settles in" well after the initial "ah...vocaloid..."

Re: "Do the lyrics match the upbeat melody?"
I thought they were a bit "esoteric" for such a Pop melody. But that is to be expected for "based on Poe".

Well done. Keep 'em coming!

fj

Wow, who needs singers?

I was wondering about that vocal sound, then I read the explanation.
I'll have to look into the Synthesizer V.

moto

David, I really enjoyed listening to this work! And I really appreciate all the background information, especially the connection to the lyric source, which I read after I listened to the track. I love the vocal harmonies. One of the great things about this forum is that new ideas come from listening to the work of others, and that is certainly true here. You’ve given me some new things to think about! Also like to use that chromatically creeping fifth tone in the melody. Thanks again for this contribution!

Well that was interesting.
A little mechanical but I'd like to play with that synth.
Maybe for backing vocals. Can you humanise it? I'll give it a try.
Thanks for posting, David.

Vic

That's really really cool, David. Quite amazing. Like painting with music. That counter-part harmony is perfect. A fun listen for sure!!

Very clever, David!

Vocally I like it best when the two "singers" are singing together and off each other, that worked quite well I thought.

Great sounding track too, and nice mix.

Enjoyed!

The synth vocals a making huge leaps forward...the harmony was particularly good.
Cool.

David,

Solid composition! Lyrics work very well. Personally, I would prefer more bass, especially for the genre you chose. Besides that, fun listen! Those vocalizers are definitely getting better!
Being a fan of ELO I can definitely hear influence.
"She does the things you do. But she is an IBM" ELO (c)

Correct me if I am wrong.... As I understand there are three main uses for vocal synths in music.
#1 Composition / draft - instruction to live singer
#2 Actual production to achieve certain articulation and vocal tonality, or some backing vocals. But I would think, unless a producer wants to achieve "certain" synth feel that only vocalizer can give, why not hire a real singer? Does not make much sense, unless "producer" is on extremely tight budget.
#3 having fun.
-----
Please do not get me wrong, I am not being negative. I do believe this technology is very useful, I am sure it is benefits people who are mute, visually impaired and many other areas. Just trying to understand general usage, or passion if you may, behind it.

Thank you for sharing.
Misha.

Hi, floyd jane.

Thanks, glad to hear it!

Hopefully next time I can get a "ah, not vocaloid" for you!

Thanks, captain!

Interestingly enough, the version of Synthesizer V I'm using is the free version, and it's a lot more functional (and better sounding) than the Tiny Vocaloid editor I'd been using. While the Tiny Vocaloid editor only allowed a single voice limited to 16 bars, the Basic version of Synthesizer V allows for two voices and unlimited song length.

Plus, the Tiny version of the Vocaloid editor is no longer available, so it'll cost around $200 USD to get into the game.

There's currently only one English Synthesizer V singer available, but there are enough parameters to adjust that you can get a fairly wide set of voices from that single one. For my purposes (demoing a song), that's sufficient.

Hi Tom.

That's the best I can hope for.


While it's best to have real vocals in the end, I find it a lot more helpful to do a harmony mockup in Synthesizer V instead of using a MIDI flute - especially if the line is going to have different lyrics.

It's also good for checking how well the parts "work". I know that synthesized vocals suffer from poor intelligibility, and this song in particular has issues. Still, I found it helpful to be able to adjust the timings and words so they didn't interfere as much. Doing that while recording (or afterwards) is much less convenient.

Obviously, the downside is that you can end up writing stuff that's out of a specific singer's range.


Yeah, that's a cool sound! It shows up in some of my favorite songs - it also pops up in the chorus of America's "Sergeant Darkness", for example.

Thanks for stopping by!

I'm using the free version of Synthesizer V, and it's available for the Mac, Windows and Linux, so the barrier to entry is pretty low.

The only English voice is only recorded at a single pitch, so as it moves out of the "best" range, it tends to get a bit nasal. However, changing the tension parameter helps with that.

There's a lot of options for "humanizing" the sound, but I've found that you need to be careful - it's a very short distance from making the voice more "human" to making it sound like a bad singer. I'm still figuring out where that line is.

Although you can edit all the parameters to the micro detail on the timeline, I've found it's better to start with a higher level approach. For example, quite often a singer will embellish a note instead of singing a note straight. So adding embellishments (say, a 16th note grace note to a syllable) can help make it sound less robotic.

I think I mentioned that I didn't spend a lot of time "tuning" those parameters on this song, although I did spend a time replacing phonemes so there was a more American pronunciation. There's also an option to adjust the relative duration of each phoneme at the syllable level, which is also really helpful for making words sound more natural.

I'll be happy to answer any questions about it that I can.

Thanks for stopping for a listen!

I'm glad you liked it, thanks!

Playing with synthetic voices can be constraining, but in some ways it's liberating because it also feels like painting with music.

Mine too!

The problem for me is that it's really tempting to put in too much of it. When I get around to buying the program, I won't be limited to "only" two voices. But for now, I think it's a good limitation that's keeping me from going overboard.


Thanks!

I appreciate you taking the time to stop and listen.

Hi, Ray.

Yeah, I think this software works much better than Vocaloid. Still synthetic, but better than before.

I'm glad you liked harmony. In some ways, it's a handwaving cheat, where I can say "Just listen to the pretty harmony, so you won't notice that the lyrics aren't very intelligible."

Thanks, Misha!

I generally trust my son (a bass player) to find the right balance on the bass in the mix. If I even hint that it should have more bass, you know he's going to turn it to 11!


I'm glad some of it got through, despite the lack of a cello section. (And nice reference!)


Obviously, there are as many uses as people want there to be, but I think you've hit on the primary ones.

Last year, I'm not sure I put out any songs. That wasn't my intent, so this year I decided I'd try to change that. That meant removing as many barriers as possible, even if that meant lowering my standards. As the phrase goes, "Perfect is the enemy of good."

One of those barriers is recording vocals. I think I've mentioned before that my house isn't really conducive to recording, with my computer out in the open in the literal center of the house, with a house full of people busy doing their own stuff.

So being able to put together the vocals while someone is playing video games just feet away is really a bonus.

I think of these songs as as drafts. Having a "real" voice on the vocals is always better, but at the moment, I'm trying to write more songs. So - for now - if I have to decide between getting a better vocal or moving on to a new song, I'm going to choose working on new songs.

Obviously, that's subject to change.

I have more than a passing interest in vocal synthesis. I've been working on a synthetic singer of my own for a number of years - but I don't think it's in any way preferred to a real singer. I'm just using Synthesizer V as a tool to get demo songs out the door.

If I had to sing the vocals, that would likely add another 2-3 days of time working on the song. I've generally got to wait until after midnight before it's quiet around here, and then the vocals need to be worked on because... well, I'm slow.

I also don't have access to a female vocalist, so Synthesizer V gives me a color that I normally don't have access to. I've mixed real and synthetic vocals before, and I think it's more jarring than having all synthetic vocals.

I could ask someone to sing the vocal, and I suspect there would be people on the forum that would be gracious enough to help out. As I mentioned, I may very well revisit some songs for that reason.

But I don't think I have a "passion" for Synthesizer V, in the sense that I prefer using it over a real vocal, or am showcasing it. I think of it as a tool like a MIDI instrument (sorry for that analogy, Mario) - it gives listeners enough of a flavor to hear to what I had in mind, but is hopefully not much a distracting replacement for the real thing.

So I appreciate people's kindness in listening to these songs despite them being sung with synthetic voices.

Thanks for asking!

Hi dcduny,

Pleasant orchestration. Nice second voice. It's very soft and pleasant to listen to. Well done

Kindly regard
Derochette
alisas JaniJackFlash

Hi David,

You based this nice, but also a bit peculiar song on a pom of Edgar Alan Poe!
I had to read his bundle of stories in English when I was young. I loved the tales, but the words were so difficult to understand.
I took the translated version of it to get through. Illegally of course, but it helped me much! I still love his tales, but I never read his poems, strangely enough.

It took me a moment to get used to your way of singing here, but then it became more and more interesting.
And what you did with Synthesizer V is also very good. The female vocals sounded very realistic, although they were used in a surreal context. I'm thinking of trying this too.

So thumbs up for your work, David!
Hans

I have to admit that I'm not familiar with Poe's works, although I have listened to the first Alan Parsons Project album and then later Eric Woolfson's Poe musical songs. But you have nicely captured the mood, and at least the title gives upbeat impression.

Definitely sounds bit like an Aztec Camera song, at least in the chording (Safe In Sorrow) and the instrumentation (drums, piano - other Dreamland songs) and the counter-melody (Sister Ann). Not only because sounding like one of my all time favorite bands, this is really nice pop song with interesting chord changes and very pleasant melody, including the enjoyable counter-melody. The arrangement is lush and the solo sax fits well.

Janne

David,
Thank you for explanation. No doubt, synth vocal technology has a significant potential. Miracle for composers to get stuff, as you said: "out of the door" I knew you had been working synthetic singer for some time, that is why I wanted to ask directly to get fuller picture and a sample of philosophy behind usage
Thank you.

I'm going to be 100% honest. That synthetic voice kinda jars with me, sadly. Pity, because the song sounds incredible otherwise ... really great. But that voice jars me and wakes me up from the "light trance" I usually go into when hearing a good song.

That said: I realise the limitations of synth voices - and it's not like I'm Freddy Mercury in the slightest when I sing, either, haha ... so:

(apart from the above mentioned) I really, really enjoyed it. I also really, really "envy" the clarity of mind that you have in your approach, as shown by the way you detailed things above. And the song itself is so well done ... there's a certain amount of complexity in it, in many ways, and you handle it effortlessly. So well written. Sincerest congratulations!

Thanks, Derochette.

I appreciate you taking the time to listen and comment!

I'm glad to hear it kept your interest!


There's an AI version of some of the Synthesizer V voices, where it can apply pitch fluctuations that it's learned by listening to an actual singer. Unfortunately, that option isn't available for the English version.

But there's a work-around, where you can switch the voice from the English voice to an AI voice (such as the free Saki AI (Lite) voice), select all the notes, apply the Auto Pitch Tuning (now that it's got one of the "AI" voices), and then switch back to the English Eleanor Forte (Lite) voice. It'll keep the values it's written in the Pitch Deviation track.

It's not perfect (I just figured it out last night), but it does a fairly good job.

Thanks for commenting!

Hi, Janne.

I've always liked Tales of Mystery and Imagination, but I think in Poe, Steve Balsamo really knocks it out of the park.


Dreamland is great, but High Land, Hard Rain was probably my favorite album back in the '80s. I played it constantly.

Thanks for your kind comments about the song!

One of the frustrations with my own software project is that - at this point - the results are actually pretty good, but not what I'd want to use on a song. So I'm not sure I'll ever be able to deliver a usable tool with what I've got. But I'll still probably keep working at it, because it's a technical challenge.

For most people (myself included), I think some voice lessons, good microphone technique and pitch correction software will produce results miles beyond what voice synthesis can do.

I appreciate your honest response!


Thanks!

I keep having to remember that the key to clarity is to remove what doesn't need to be there, no matter how nice it sounds. There are a number of songs where, on listening again, I really regret not taking more out. And on this song, my son chopped out a bunch of things after I did my initial edit, which gave even more space for the remaining instruments to come through.

So I'm glad that came through in the mix!

Hi David .
What you have created here is very impressive. I was interested in Vocaloids a few years ago but the software was very expensive and all of the voicebanks were Japanese so English pronunciation was a problem ( no Jeff Lynne in sight lol ) and so my enthusiasm waned . Now your excellent post has appeared which has piqued my interest again. I am working my way through the UTAU Manual as we speak , trying to get a handle on it all . I can see so much potential with it going forward. Well done putting this production together.
cheers
Gary

Hi, Gary.

Thanks!


Vocaloids tend to have accents, and the resynthesis process only magnifies that. The Dex and Daina voicebanks focus on a more American English accent, and I was toying with the idea of buying the full editor, but I wasn't terribly impressed by the demos, and the price of the full Vocaloid editor always priced me out of the market... even when I had that much money to spend.

UTAU is cool, especially since you can build your own voicebanks. But working with voices requires a lot of low-level tweaking. One nice thing about Synthesizer V is that you have sliders for each phoneme at the syllable level. That makes it a lot easier to tune a word, and even work around errors in the voicebank.

You may be aware that Synthesizer V actually grew from Kanru Hua working on tools for UTAU. He developed his own resampler (resynthesis engine) for UTAU called Moresampler, which is different from most most resamplers available for UTAU. Most UTAU resamplers use crossfades and standard pitch transposition techniques to reassemble the phonemes from the source audio files.

Moresampler performs spectral analysis on the harmonic and inharmonic audio, as well as the vocal pulse. The resynthesis doesn't need to refer back to the original audio. As a result, Moresampler is a lot more flexible in resynthesis, both in in changing the pitch and duration. It also doesn't suffer from as many of the sort of artifacts you get from a lot of pitch shift/time stretch algorithms.

Kanru also came up with a his own reclist (phoneme recording list) method which he called "Arpasing", named because he used the ARPAbet phonemene set instead of SAMPA phonemes that Vocaloid and UTAU tend to use. (The word "Arpasing" confusingly refers to "ARPA" in "ARPABET", not to "parsing").

Because his reclist uses English words, the resulting samples tend to be more realistic than if the phonemes were recorded in isolation. On the other hand, Synthesizer V sounds a bit more "spoken" than "sung".

Anyway, at some point Kanru rightly decided that it would be better to branch away from UTAU and create his own application from scratch. If nothing else, Synthesizer V is a lot simpler to install than UTAU.

Rather amazing. AI is here. He/she sound pretty good. I really like the song. Lovely melody. Good use of the band too.

Thanks for taking the time to stop, listen and comment!

David, I definitely hear the Jeff Lynne reference in the chord progression. The mix sounds great. Poetry is a excellent idea for a starting point. When the two voices intertwine it is so well done that I consider that the highlight of the piece. Well done again, Tom

Thanks!


I appreciate you pointing that out. Those harmonies are probably my favorite part.

David,

Another wonderful vocaloid song!
Very pop!
The cliché of I - I#5 - I6 - I#5 is doing a very good job in this song.
Enjoyed my listen a lot.

Best regards.

Shigeki Adachi

Hi David ...

What a great song! The vocaloid was well executed - especially the harmony. And I, too, was a fan of ELO. It was a little different from most of the pop rock music offered in the 70s. I can only imagine the work involved with created a near-0human sounding voice. I applaud your effort and results.

I've always like that chord progression you use in the opening : tonic followed by the augmented, followed by the 6th and then reversed to tonic resolution. I've heard in a few songs from the 60s and 70sd but not too often since. The Dave Clark Five song "Because" from 1965 comes to mind.

Nicely done and a pleasant listen. B est to you ...

Alan

We, again, are not gonna join the synthesized vocal fray It makes for a good "do it all yourself" demo.

The melody was very nice as were the harmonies and counterpoint lines.

The backstory can stand alone as a good and interesting read.

Us rubes out here in the deep south provinces have not read much Poe other than the obligatory high school exposure. However, your shoehorning of his poetry into this lyric resulted in a fine lyric. The rubes missed out on ELO also but that notwithstanding we much like and enjoyed your arrangement.

Guenter from several years back and Shigeki often posted vocaloid based songs but haven't of late so it's cool to again hear and learn about this technology!

J&B

David, that sure is a catchy tune! Loved the use of the diminished chords in this, they really bring such great vibe to the song! Thanks, T

Brilliant! The lyrics shine and the melody is outstanding! And I just love the Saxophone - superb choice. Superb choices at every level!

Hi, Shigeki.

I'm glad you enjoyed the song. I'm sure to re-use the I - I#5 - I6 - I#5 progression in the near future.

I'm curious if you have any thoughts about how Synthesizer V sound in Japanese. Most of the current work on the Synthesizer V voicebanks is being focused on Asian languages, because that's where most buyers will be.

Thanks!

Hi, Alan.

Thanks! Obviously, I didn't end up going for the "real" ELO sound, but I won't rule it out in the future.

I don't want to give the impression that using Synthesizer V is particularly difficult. It's pretty much the same as using an music program with a piano-roll interface, only you can add words to the notes (and then tweak the sound). But it's quite usable out of the box.


I'd used it on another song I did a number of years ago. It's a very cool sound.[/quote]

Thanks for stopping to listen and comment!

Heh. It would be a massive step down for you.

And yes, these songs really are "demos", even if that's as far as they get.


Your background isn't that far from mine. The main difference is that I'm a big fan of the Alan Parsons Project, who started their series of concept albums back in the '70s with an ode to Poe titled Tales of Mystery and Imagination.

More recently, Eric Woolfson (one of the two producers of the Project) did the musical Poe - you can find the video here, and it was also released as a studio album. Think of it as the musical Cliff's Notes version, although - as expected with any adaption - it veers a bit from the historical reality.

Good to hear from my favorite "rubes"!

Thanks, Torrey!

Thanks, Lawrence.

The sax wasn't my first pick, but it was exactly what the song needed there.

It's really nice to be able to get pro-level music from BiaB on demand. It's almost an embarrassment of riches.

Liked the lyrics. Liked the sound of the backing tracks. Nice mix.

Fun to see what you can do with synthesized vocals.

Great job!

Really like this David!! Very melodic, very well arranged and produced!! Was there an Alan Parson's song with that title.....they did a Poe inspired record and the title is very familiar to me. Very cool stuff!! Take care. Greg