Omnivocal EULA

The beta EULA is quite open ended apart from the copyright restrictions (other artist’s music or melody) and the ‘moral’ aspect. Based on the text I don’t see how Yamaha could pursue any infraction in commercial use of the beta version where original material was involved.
But, once a proper GA EULA is drafted, it will be interesting to see if there are restrictions applied as per Vocaloid, given the competition is offering pretty wide open commercial usage.

1 Like

to say the software cant legally be reverse engineered is clear and acceptable. It needs to be just as clear to say “ a user can freely use the sounds generated in this software in their own unique musical sound compositions without restrictions”

1 Like

Hi all,
We asked Yamaha to clarify this part of the EULA:

Copyrighted data, including but not limited to MIDI data for songs, used by or obtained by means of the SOFTWARE, are subject to the following restrictions which you must observe.

Data received by means of the SOFTWARE may not be used for any commercial purposes without permission of the copyright owner.

This section only refers to the input data you use with Omnivocal, such as copyrighted MIDI files.

You can absolutely use Omnivocal for commercial work, but you can’t, for example, download the MIDI version of “My Way,” run it through Omnivocal, and publish it as your own intellectual property. That’s the part the EULA is addressing.

Yamaha will review the wording to make this clearer.

5 Likes

Hi Matthias, I have a question related to this post: many music distributors — CD Baby, TuneCore, etc. — are starting to block tracks that contain AI-generated voices or instruments, as we all know by now. Does this also apply to Omnivocal? In other words, if I use an Omnivocal voice in one of my tracks, do I risk it not being published? Do you know anything about this?
Thanks, Lucio

This post was the first I’d heard about that, so I did a little quick research to try to dig into some specifics.

CD Baby has two specific places where they address these restrictions:

and:

https://support.cdbaby.com/hc/en-us/articles/23638450756237-Understanding-Production-Sounds

In the former, they say, “Even if the A.I. generator you are using permits commercial use of what you made, there’s no way to ensure unique sounds or rights compliance.” This would appear to be a “safe” (as in “C.Y.A.”) stance given that, despite some settlements (e.g. Warner), there is still ongoing litigation (e.g. Universal, Sony), against companies like Suno. In the latter, they add, “Even if you contributed original sounds to the recording, and only part of the recording is AI, we are still unable to distribute this content.”

TuneCore appears to bury a more general policy in an article relating to one specific AI technology that license use of an AI version of Grimes’ voice for use in original songs and recordings subject to credits and a royalty split:

https://support.tunecore.com/hc/en-us/articles/16428915033492-Distributing-collaborations-with-GrimesAI

Within that article that is mostly specific to GrimeAI use, they say, , “TuneCore will not distribute any works that are 100% AI-generated, but we are in support of the use of AI technology that enhances human creation, assisting with artists’ productivity and increasing fan engagement and artist revenue.” That is a little nebulous.

The other key distributor, DistroKid does, however, explicitly allow generative AI recordings, subject to the AI company’s terms and a few other specific restrictions (e.g. no impersonation):

https://support.distrokid.com/hc/en-us/articles/41182362733715-Can-I-Upload-Music-Made-With-AI-Tools-to-DistroKid

All this said, there is nothing in any of these that I’d interpret to affect Omnivocal since Omnivocal is vocal synthesis, not generative AI. You’re playing and/or programming the notes in your recording. Assuming you’re doing the same with the rest of your recording, and that you’ve written the song yourself you aren’t generating the song with AI. Omnivocal simply provides an instrument that has a human-sounding voice that can also translate its notes into words.

1 Like