completely A.I. generated songs with singing are here (Udio/Suno)

[enroe]

https://suno-ai.notion.site/FAQs-b72601 ... 11c73ade77

Suno claims that If you are a paying subscriber then you "own" the songs you generate.
While at the same time Suno has no guarantee that the songs generated using Suno are subject to "copyright" protection

These are all legal questions that are by no means clear. Suno is
simply deciding this for itself - but for the first model lawsuits, Suno
would have to show which songs the AI ​​learned from. Because every
neural network needs a lot of learning data. In other words, songs that
Suno can learn from.

"AI-generated" means: pieced together from the vastness of the
internet. The AI ​​looks for all possible puzzle pieces and set pieces -
and puts them back together again according to a learned algorithm
(neural network). In other words, the AI ​​ultimately regurgitates what has
already been there. One could object here that every human musician
actually does this too.

I assume that the major companies will file a lawsuit. The legal
dispute will be very exciting.

Look here:

When you enter 'beminem' into Suno as a prompt, says Ed Newton-
Rex, it produces music that's strikingly similar to that of superstar
Eminem.

Newton-Rex, a classically trained composer, has tested the platform
extensively and found numerous examples in which the songs
generated by Suno are strikingly similar to well-known hits in terms
of style, melody, harmony or instrumentation.

Suno blocks AI prompts with artist names. However, small spelling
mistakes in the prompt are enough to outsmart the system. One song
was conspicuously created with the lyrics of Abba's Dancing Queen
and "70s Pop" as the prompt.

Look here and here.

So, the last word has not yet been spoken here.

In the long run, however, other platforms will emerge similar to Suno -
and ultimately shake up the entire music world.

[painpotato]

I would recommend everyone interested in this topic check out Stable Audio which is another generative AI music provider just like Suno AI/udio,but the difference is that Stable Audio was trained on a licensed dataset from the AudioSparx music library consisting of over 800,000 audio files.
To me the generated music from Stable Audio is ok but not stunning at all like what Suno/udio did.
And that makes me wonder how Suno/udio actually did to make such difference.

[Tiles]

The difference is imo the hand picked training data. Stable Audio trained their AI most probably with everything. You name it, it's a dataset from AudioSparx. And this includes low quality content. Udio and Suno might have focussed more on high quality content here.

[osiris]

Udio is better IMO, because they've taught the AI breath. I never breath sounds on anything Suno did. I also think it's funny how much time I've spent trying to control breath sounds on singing and now breath sounds make software generated voices sound more real.

[igorius]

Suno currently has a referal program. If you register with this link: https://suno.com/invite/@mus_ai_c and generate 10 songs (you can use the free version), you get 250 extra credits.

[Ratibbla]

I would recommend everyone interested in this topic check out Stable Audio which is another generative AI music provider just like Suno AI/udio,but the difference is that Stable Audio was trained on a licensed dataset from the AudioSparx music library consisting of over 800,000 audio files.
To me the generated music from Stable Audio is ok but not stunning at all like what Suno/udio did.
And that makes me wonder how Suno/udio actually did to make such difference.

I wonder if they (Stable Audio) give enough free credits. I'd rather try it enough times first before paying for AI. Considering that they do not give amazing results.

---

Could it be that the main my essay writer ai will become chat gpt? Perhaps yes, if it is well configured and tested a hundred times. But everything needs to be checked.

[osiris]

The problem I have with this is, how do human singers learn to sing? They listen to other people singing, so how is this different? As long as you're not telling it: sing like Adele and it copies here voice. I don't know about other services, but Udio will chunk any kind of instruction that says : Sing like, or in the style of. I will note this for free accounts. I don't know what the paid tier has. I had an email from Gaudio saying they were taking all their files down and come get what you want to save.

[enroe]

In the 60s and 70s, music had a completely different reputation
than it does today. At least that's what our grandparents tell us: If
you heard a great song on the radio back then, you only had one
option: you went to a record store and bought the single or the LP.

The albums of the music heroes had an almost godlike status -
and almost everyone bought the LPs of their music heroes. Of
course, many recorded the charts and radio shows on cassette,
but that was not the same: the tapes wore out over time. And the
elaborately designed cover of the LP was also missing.

Something happened gradually in the 80s and then in the 90s:
thanks to digital recording media (Sony minidisc, DAT recorder,
CD recorder) you could copy albums in super quality. And (almost)
everyone did that.

This meant that more people could enjoy the music - but the
uniqueness of an album and a song was reduced. Because almost
everyone had the song at home on DAT or CD.

Something else happened in the 2000s: almost anyone could
download any song from the internet as an mp3. The music industry
is still fighting against this, but ultimately without success. Today
(almost) everyone can have any song on their hard drive or
smartphone.

All these steps, digital copying capabilities and availability on the
Internet, have devalued music to this day. For young people today,
a song does not have the same meaning as it did for young people
in the 60s or 70s.

I just discovered Suno recently and I find it absolutely mind-blowing. That's some crazy tech sorcery. It can create any genre and the voices really sound like they have human emotions. The rhythm and emphasis of the vocals and song arrangement are so on point.

I think AI is drastically going to reshape the music scene. Anyone can create hits within a few seconds. And everyone will be like "Hey, listen to my music" "No, man, listen to MY music".

In this spirit: Listen to my AI music =)
https://on.soundcloud.com/Yf1m9

Cover images created with AI too.
It's got some happy pop, sad slow pop, synth-pop, flower power, hard rock and italo disco.

AI simply continues this devaluation evolution with the next step:
The fact that everyone can simply have "their" song composed by
AI devalues ​​music even further: Music and songs are simply a kind
of "good vibe" companion to everyday life that everyone can have
created for themselves while eating breakfast.

[WackyZoundz]

The devaluation is not even the worst aspect of it, it's the way people adjust to AI music. They lose (or never develop) the ability to understand music on a deeper level. No more feelings, no more message, no more thoughts, just convenient background noise based on statistics (because that's all you can get from AI). This inevitably leads to the devaluation of pre-AI music as well. Mozart, The Beatles and so on will be viewed as bad music. Too demanding, too much work to consume it, too unique, too unpredictable. It's simply the death of art or at least art as form of communication.

[enroe]

Maybe we shouldn't "make such a big deal out of it".
Because the first legal disputes show:

Suno and Udio don't compose or mix at all. No! They just copy finished
pieces of music and just swap the lyrics. Well, the "swapping of the
lyrics" is done pretty well, in AI style. But the music is just stolen -
sometimes in set pieces.

If you want to see for yourself, please click here .

So: The devaluation of all music by Suno and Udio is nothing more
than a fake. AI needs at least another 10 years to devalue it.

Until then, I would like to encourage you not to be discouraged from
composing your own music note by note on the DAW - or better yet,
on paper. Your own composing is unique, distinctive and one of the
highest forms of your own expression, your own art!

[igorius]

Maybe we shouldn't "make such a big deal out of it".
Because the first legal disputes show:

Suno and Udio don't compose or mix at all. No! They just copy finished
pieces of music and just swap the lyrics. Well, the "swapping of the
lyrics" is done pretty well, in AI style. But the music is just stolen -
sometimes in set pieces.

If you want to see for yourself, please click here .

So: The devaluation of all music by Suno and Udio is nothing more
than a fake. AI needs at least another 10 years to devalue it.

Until then, I would like to encourage you not to be discouraged from
composing your own music note by note on the DAW - or better yet,
on paper. Your own composing is unique, distinctive and one of the
highest forms of your own expression, your own art!

True. I noticed in one of the songs that Suno created for me that the beat had a striking resemblance to "Don't Look Any Further" by Dennis Edwards and Siedah Garrett. It was too similar to be a coincidence. The song itself wasn't just the lyrics swapped though, the singing had a different melody, so it was more like a remix or something.
But I know of this case where they fed Udio with the lyrics to Mariah Carey's song "All I want for Christmas is you" and the result was almost an exact copy. It can be seen and heard here: https://www.404media.co/listen-to-the-a ... suno-sued/

[osiris]

I never got that and I spent at least a couple of days feeding it famous song lyrics. Some of the results were quite hysterical.