This is an experiment with Alt Rock and Synth V. This is a rough draft of harmonic structure and AI vocals. I will replace the female voice with a male voice when it is for sale.

I need to know if I should continue with this or start something new.

Comment as you like. I am not sensitive.

https://soundcloud.com/planobilly/indie-test-track-mixdown?si=d23a71ae2b9a41aa8c89cb09a633f071&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing

Billy

That voice sounds amazingly like Liz Phair.

Yes, keep going please.

It's good. I'd continue.

Billy, as I am sure you know when we met BB and Robert Johnson down at the crossroads, that BB done told you: Billy the blues ain't about what you play, the blues is about what you don't play.

And if the levee is gonna break, you better not be anywhere anywhere near Orleans, cause the ghost of Stevie Ray is a voodoo chile, and you might get more black magick than you bargained for.

But if you need to get it in, just keep in between the ditches and don't bend over when they go to handcuff ya.

You'd be better off sitting in Yellow Mama if you know what I mean.

Well, I hope that about covers it.

Good piece of a song you got brewin' there. Sounds like the devil is mixed up in it, but you don't never seemed to have cared about that, and that's how you got as crazy as you already is is what I reckon.

LOL...well, David, I have known that I am crazy for quite some time now. I have been justly accused of playing the devil's music on more than one occasion.

What comforts me is that I know I am in good company...lol

Billy

Def pursue this.
My sole suggestion would be to drop "I don't want no" for a phrase that's less colloquial as the synth voices don't do slang/street very well.
"I don't want a..." would sound better from the rather "correct" sounding voices the synths offer.

Hi rayc

Yes, for sure, the lyrics do not match the voice. I will change the lyrical context to a subject matter more suitable to this voice.

These Synth V voice databases appear to only deal with a minimal number of musical styles. There is a "rock" voice that is in the works. Who knows how it will turn out?

The way the software works is more than a little difficult to use. Sometimes I find that what I had in mind for a vocal line will not work. To some extent, I guess all songs generated with software become "software" songs.

I was surprised that this voice had a limited vocal range, around two octaves. It does not sing on pitch out of the box and requires fixing with Melodyne, autotune, or both. The hope is that this technology will rapidly improve.

Billy

I like the song, but you need to work on your vocal rhythm.
Right now you almost always have one syllable per beat, that in itself sounds robotic, even if a real(TM) singer would sing it.
If a word has 2 syllables, that doesn't mean they are both the same length when sung.
Record yourself singing, adjust the rhythm and volume changes according to it and your vocaloid will sound MUCH better.
When you breathe, your synthetic friend needs to do that also.

Those vocaloids have a sweet spot where they sound best, just like real singers.
Don't request things they can't deliver.
And stop using Melodyne!

Thanks, B. D.

There is much to learn about using Synth V. My comment about range was based on the only two voices I currently own. One of the first things I did was have them sing "Joy To The World" which uses all the notes in a major scale. That gave me a range of around two octaves. I found this unusual because even an average good vocalist has a three-octave range.

The issue of using pitch correction software, I guess is subjective. I am pretty sensitive to things being out of tune. I was also surprised at the number of vocal notes that were off-pitch.

If there is a better way to control pitch issues, please let me know.

I was impressed with your songs due to the lyrical content and the video. I was also impressed with how they were accurate depictions of real life. That distracted me from analyzing pitch issues.

With the number of songs that are being generated in today's world and the wide variations in musical style, there is much that is not "standard." I think it is unreasonable to expect that software-generated vocals should sound precisely like human-generated vocals. Would we even want exact human replication of the vocal? These are very subjective questions.

One of the issues with a guitar riff-based song is coming up with something different enough not to be derivative of something previously done. I was also experimenting with having the guitar, bass, and vocal be in unison at different octaves. That may be a good idea, or perhaps not. Again, subjective stuff.

I do think it is possible to rewrite the guitar riff to incorporate your ideas of variation in syllable length. Also, as I mentioned, I will rewrite the lyrics to fit the singer both from the enculturation aspect and the rhythm issue you mentioned.

Again, thanks for the ideas.

Billy

The vocal sounds great but like my struggles, it's very difficult to line up that vocal well. Not a macro for that yet!

Keep on Truckin'!

so far so good. I don't know that I would feel like it had to be a male singer.... it seems to be working as it is currently. Of course the mix needs to be dialed in but you knew that already.

Keep on trucking. Oh yeah... and be careful about signing anything in blood, especially if it's yours.

Sounds like it's worth continuing.

I'd focus on completing the song and worry much about the vocal later.

However, you might try modifying the Gender and other parameters until you decide if you want to use Ninezero or another voice.

It's out!!! I didn't know the rocker voice was released, Ninezero. Thanks for the tip David!

EDIT: Picked up Ninezero. He also comes with a bunch of small .WAV clips with his voice. Alphabet letters, numbers, him counting, an emcee folder of phrases, shouts, screams, growls. One of his .WAV clip samples below.

NineZero Voice Sample (MP3)

There's also an upcoming English voice called "Jun" that should be released in late April 2023.

See: https://audiologie.us/products/synthesizer-v-ai-jun-voice-database-english-digital-download

Not a very gritty voice, but from the demos it seems that he's able to get a harder edge than Kevin.

Yeah, Jun kinda seems like a blend between Kevin and NineZero. Thanks for sharing!

That NineZero clip scared & scarred me. I can imagine it being used and the sort of song that would use it: Damagingly awful.

It should have scared you, he's an Aussie!