Singer to play via keyboard - what's your take on it?

Probably some of you have tried some newer AI tools in form of singers already?

I didn’t yet,…
but with checking what new features come with the update of the DAW I use (Cubase). there’s now such a thing, reminding me of
ā€œPlogue Alter Egoā€? Alter/Ego
which I tried a few years ago.
A demo here: https://youtu.be/83JMEjKfSHc?si=0cQNY_mCUaMPefw4

So you play your keyboard, then you type some text to the notes and this way you let a text to speech engine sing.

The description for that new tool included in Cubase says:


Yamaha Omnivocal is a text-to-speech (TTS) singing synthesis technology, available in beta in Steinberg Cubase 15, that creates vocals from text and musical notes. It is a professional-grade tool for music production, not a general-purpose or simple TTS converter like some online services. The technology is designed for creating natural-sounding vocals within a digital audio workstation (DAW) environment, allowing users to generate singing voices with a simple interface.


No mention of AI, but I didn’t research deeply.
The thought of AI somehow invokes a negative feeling, but on the other hand, in this case, I am creating notes and lyrics myself and just someone lends me his/her vocal.
Still feels a bit strange, but when I just threw it on something I quickly created, I was baffled. :open_mouth:
Somehow expected this from the the good sounding demo, but playing a keyboard, quickly imagining whatever lyrics and then, have someone sing for you - and rather convincingly feels strange and I felt like it let my creativity increase quite a bit at that moment.

Here’s that song- ā€œconstruction-siteā€ alone…

I will explain what strange things I did here for those who might be interested because these sounds are rather untypical for me:
  • I basically just wanted to test the recently added pattern drum editor which I used in some software maybe 20 years ago once or twice and found fun at the time.
    Don’t jugdge my electronic drums too hard, they were just a quick clicking some patterns together.
  • Then I added a virtual acoustic bass. I should use this more often, because it sounds quite good and I love a double bass. Oh - I was in experimental phase, so I tried some step-filter plugin and some distortion on it.
  • Of course - this is a guitar site and my main fun is the guitar - two tracks of guitars - my own playing (you hear some string squeeching somewhere), also some effects used (would have to look up what I did there)

today, opened the above (without even remebering what it was, because I accumulated quite some of ideas or tests which I never finish) just to try out this vocal feature:

I was also thinking about where to put this post.
Though it includes audio, I thought it rather belongs under discussion. than AVOYP. It’s just a ā€œwhateverā€ included music file just to demo that vocal tool.
And then I didn’t put it into the ā€œAIā€ category because I felt that AI isn’t the main topic here, rather Text-to -Speech with probably help of some AI I guess.

1 Like

I would have placed it in the Gear and Tools Talk/Hardware,Software and Recording

1 Like

Just Chatting forum should be fine (which means it will probably get moved). The lines are getting blurry with the use of musical tools, AI included. I’m talking Trio +, drum loops, etc. There will be more DAW integration moving forward as well.

2 Likes

The title of your post, along with the content and discussion, suggests to me that a thread in the AI category would be the most appropriate place. AI has already become integrated with many platforms/programs and is rapidly becoming impossible to avoid.
It would also fit quite happily in tools/software, as Stitch suggests or, if it were a fully-fledged song, in AVOYP.
I can’t see it being ā€˜moved’, as the just chatting section functions as an ā€˜everything else’ bin.

1 Like

Oh well… maybe my quick demo above was too strange (instruments quickly put down).
Until now, you repliers only replied to whether to move the subject or not.

Or maybe I shouldn’t have mentioned AI at all, because I think that most of us - me included have some strange feeling in the guts about it - and Yamaha doesn’t say a lot about their engine and they don’t mention AI at all in the decription above or their user manual.

Here’s a video with a short overview:
https://youtu.be/tidtwO1tM3I?si=vUY7VWiEOpuysdj4

So I recorded some more guitar and sung myself… And suddenly playing the keyboard is more fun than before to me, knowing I could give it a voice and have that voice sing the lyrics I type in. I t’s a bit like having some other singer right here.
Me, I can only sing a male voice - and for sure I want to work at it, because I am aware of my limitations, but here, I can easily sing along with a female singer without having to ask and disturb someone else.

And this is strange: While using it, suddenly lyrics come more easily to me. Not GREAT lyrics, but at least some that are ok somehow :slight_smile:

1 Like

As said elsewhere already, I think it can be a useful tool for songwriting and developing and testing ideas. And if it helps you and makes, what is otherwise a bit more difficult, easier - all the better :slightly_smiling_face:

1 Like