Loopy Pro: Create music, your way.
What is Loopy Pro? — Loopy Pro is a powerful, flexible, and intuitive live looper, sampler, clip launcher and DAW for iPhone and iPad. At its core, it allows you to record and layer sounds in real-time to create complex musical arrangements. But it doesn’t stop there—Loopy Pro offers advanced tools to customize your workflow, build dynamic performance setups, and create a seamless connection between instruments, effects, and external gear.
Use it for live looping, sequencing, arranging, mixing, and much more. Whether you're a live performer, a producer, or just experimenting with sound, Loopy Pro helps you take control of your creative process.
Download on the App StoreLoopy Pro is your all-in-one musical toolkit. Try it for free today.
You’ve never heard A.I music like this
This is a very interesting take on A.I music.
I can’t believe how far it’s come already.
Comments
just to point out the obvious, its samples from a real female voice slapped into an AI algo.
Yah there is the ML part for sure. To get these results though you cant just train on one voice. It is many many voices as a basemodel and then a specific voice is used as a finetune on that mass of data.
But yah I will be impressed when someone can do this entirely with code and not use machine learning hehe.
yes…
and not only voices
e.g. sound of a guitar (Clapton, Hendrix…?), synth…
or Randy Travis used it…
Maybe many people will shun all this and live music will become much more popular again - then you’ll know who you can trust.
No samples at all. All audio generated via A.I. That’s been trained on audio data.
Either way it's someone else's stolen talent and uniqueness.....
Most people tend to appreciate music made by people they would like to have sex with… so AI has a long way to go to achieve that level of audience that don’t just love the music… they love the artist and want to get close to them.
I think most of us here think this aspect of pop music is fucked up. So, we may be more open to appreciating music created by good AI programs just as we’re have learned the sampled or modeled instruments are good enough at these prices. I’m fixing to sell most of hardware because I just don’t need it even though I love most of that gear. Hopefully others will use it again.
I’ve yet to hear any AI music that grabbed me as being the slightest bit interesting or unique. On the surface it sounds ‘musical’ and I guess that’s impressive in it’s own way, but this demonstration is an almost cynically generic take on a standard pop/folk ballad and nothing about the melody, inflection, rhythmic phrasing, harmonic language, instrumentation or timbral qualities stand out in any way (no offense to the artist or the genre, it sounds like they write nice lyrics and plenty of human music in this style does have creative and original qualities). The most interesting part is probably the slightly awkward stress patterns and strange warbling vibrato of the voice, and I’m sure that would be considered an artifact of the algorithm to inevitably be ironed out.
It seems like this process is only good at generating a pallid facsimile of the exact material it has already been trained on. I’ve spent some time trying to prompt it to do anything outside of surface reflections of well established western convention and it usually outright ignores any part of the input that goes against that, or at best serves a simple mashup of the concepts with no understanding of the underlying idiosyncrasies and no attempt to meld them cohesively. All that is to say that if you feel you have something to express that is personal to you or slightly original in any way at all you should absolutely keep doing what you’re doing and not worry about current AI techniques taking that away. Music for advertisements and other commercially oriented areas will likely suffer however and it’s a shame to see one of the few ways for composers to make a living being thwarted. Also seems like it would be great for memes.
I never want to hear AI music like this.
Imagine how I felt when the Linn LM-1 drum machine came out in 1980 with real drum samples…. Being a 29 year old working drummer I went back to college to study computer engineering.
Of course, many drummers survived to make a living but they also were asked to write parts for these new devices too.
When Walter Becker of Steely Dan fame made his solo album he used a drum machine and you can’t tell.
It’s coming… software is eating the world. AI is just the new programmable paradigm.
Humans can make better music but the financial benefits of AI for production will make the results good enough. But when can huddle here and cheer each other on as the @Linearlineman packages his 1,000 album using Neuralink from his hospital bed. Latency will be something we talk of with great humor by then… “remember when Apple thought we’d use Bluetooth Headphones?” Crazy times before Ultramarine(tooth) Neural implants and brainwave syncing using Quantum encoders.”
The best quantum computer already exists and you can use it to decipher the meaning of love.
this is why i added the comment, AI uses an original human voice sampled through different methods. all involve an original human voice. Its then analysed and regenerated.
in the example you show its clearly a regenerate human voice. if we could be bothered, we can probably find out the name of the human used for the source voice material.
that why you can use AI to generate your own voice and do text to speech or singing.
Ditto art.
Maybe I’m weird, but I don’t actually like some of the musical output from some of my favourite musicians/bands - I like them, and find their ways of doing things interests and inspires me, but a chunk of their output is rarely listened to. I still buy it to support them though.
Conversely some of the most played tracks on my playlists are from artists I actively dislike. I’d never buy their albums or go to their gigs, but the occasional track might hook me in.
It’s a package deal. And I’m never going to bond with a script, or give them my money.
it may very well be a composite blend of several and not one particular person.
Likely they are not training on copyrighted material. People doing music models are way more cautious than the image generator crowd.
If it’s trained on one person, then it’s stealing; if it’s trained on large data sets (likely), then it’s not. Are we going to pretend that humans aren’t stealing from their idols when they create art?
It's a legit concern asking about the training data used for this. You could argue that existing large language models have already been trained on the entire internet and that it's required and a legitimate use of knowledge to train these models.
One things for sure is that A.I generated music and A.I music tools is going to become a lot more prevalent. It does make you wonder about the creation process.
Depends what you mean by stealing. Some people actually license their material / give permission to these companies. But if you mean in the camera takes your soul way that is a whole other discussion.
A.I. should be fantastic it’s built on the knowledge and creativity of humanity from over many centuries, so when we ask ‘is this stealing’, we must determine who actually owns this.
Music ML companies have made it pretty clear they are not doing broad 'everything on the internet' scrapes like their image cousins have because the music industry is far more likely to and capable of defending itself. I am inclined to be believe this given how inflexible and seemingly overtrained this first wave of almost kinda good enough music generators are. If they were indeed trained on a dataset even a quarter of the size of Laion5B (the datbase oft cited for image ML gens) then you could specify exact key, time sigs, arrangement, voice type etc which is definitely not happening with these currently. The more fringe I do prompt weighted blends the more things turn to mush with these. So I just dont think the datasets are that large and they are highly fitted to a pretty small amount of source material, relatively speaking.
Yah looking forward to actually music tools like how we have Control Net for images. The music side is really a couple years behind in that department.
The stealing bit is a good discussion in itself. No one creates music in isolation. We've all learnt to play using other peoples work. Lot's of artists have been called out because their tracks sound similar to something else. Was it deliberate or subconscious. Is A.I just doing the same?
It seems like some people feel you dont really (creatively?) own what you have made because it draws so much from others. But then how can it be stealing when they also did not own it? Just because they were your idol or a genius does not mean they can lay claim of ownership anymore than you.
In the end I think the term stealing is really being hyper extended these days.
That’s not weird, that’s being an actual human appreciator of actual human music, rather than just a passive consumer of fabricated crap.
I would agree. It's an important part of the creative process. Standing on the shoulders of giants etc...
Excellent point! And I really think the average person doesn't realize that AI is not sentient. We have this imagination that AI is alive and thinking - and that's the scary part. But in actuality, AI is just a mirror of patterns across humanity.
But our lifestyle and creative patterns are THAT predictable. That's the shocking part, for me -- and a whole other conversation.
I think it's interesting that society is having AI making its debut now, in the middle of our capitalistic downtrend. The music industry model (of monetizing our musical creativity) has long been surrendered to massive corporatization. Now without a way to survive solely on our creativity, we have machines mimicking our creations very accurately.
So how do we process this new situation? Many of us are starting to re-analyze why we create in the first place. But I know it's hard to walk away from the 'rags-to-riches' narrative the industry has been trying to sell us. It's been a lie for many decades.
AI/ML is very powerful in some areas and super weak in others. Likely the ones who will advance and survive creatively/financially are the ones who can leverage the strengths to fill the weaknesses or just avoid the digital realm all together and focus on winning in the analog real world.
It certainly has not affected my music impulse much because that is just self expression and making things mainly for myself but my career path is completely altered and much more complicated now because of it,
I'm not going to get into how much every single person on this planet is going to be massively impacted by A.I but do think that self expression for the sake of self expression is going to become a very important part of what it is to be a human. At the moment we alone have that agency so enjoy that while it lasts.
gulp
Great discussion.
Collective intelligence: It’s a term I recently heard that made me think.
I think most can agree that using an AI model that is trained by data from a single entity, like the style of one guitar player, or the voice of a particular person, without their knowledge and consent, without acknowledgement is stealing.
However, if these models are trained by many many data sources from many many artists then artificial collective intelligence is what we are really talking about.
Collective intelligence is the basis of education and experience. That is how we learn and get inspired to create new things. That is why we are able to write and respond to each other here.
And no one can ever claim that everything that they’ve ever learned came from material that they somehow paid for.
The example of a star working with many other artists to produce a song, which was mentioned in this thread earlier, is also a great point. Not all of us have access to a professional production team.
Now we kinda do.
I once heard a quote, “There is no compression algorithm for experience.”
Now there is.
And I think that’s not only ok but also exciting because it’s a whole new ballgame.
Curious to hear your thoughts on this point of view.