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.
Comments
MrStochastic:
no. its not only AI voices!
all is AI! done in Udio.
3 pieces generated there and mixed/blended/mastered after.
I'm not really a fan of AI, but I was curious to see what Udio can do.
)
I started with one of my favorite music genres: 70's Heavy Blues Prog (yes, I'm old
I didn't expect too much, but was blown away from the very first try. Cool singer and an awesome guitar solo.
If the band actually existed, I'd definitely want to hear more from them
For those who are interested:
https://www.udio.com/songs/weK39oP94Nmjr9NRFjL1w1
Sorry, neither hip hop nor techno or ambient
...> @ZankFrappa said:
I agree. The level of sophistication possible with this system is just unreal. It's comes down to the "taste" the Udio system exhibits. Just six months ago, all these systems would create was nearly unlistenable garbage. Can you imagine what they will be capable of in another six months?
By the way, how do you change the 'album cover art' in Udio, @ZankFrappa? Do you change the cover art after the final assembled song? And how is that done exactly? (I find both Udio and Suno really lacking in detail on their sites)
Outstanding. What was your prompt?
New one:
https://www.udio.com/songs/j8FayVo8HTzD4xRZnxn9nG
Prompt: slide guitar rock song called "Countrified"
Sounds like the system referred to Aerosmith to create this one.
in the end, i find it all a bit creepy. i hear and like a song that nobody played? the singer didn't sing the song,
the guitarist doesn't exist either, but still delivered a great performance....?
i think it's time for me to find out how it all works. Just samples? Probably not, right?
to change the cover, i proceed as follows: click on a song title in "my creations".
on the song page, select the "edit" button to the right of the "extend" button. the cover now says "edit image".
click on it and you can enter your prompts. The cover can be changed again later.
I have just noticed that the song title can also be changed using this edit button
Amazing….. I’m looking forward to ‘stem separation’ thingy.. like 16-24 tracks.. I think this is where the magic will shine. One must wonder about the sample creators… BlocWaves, etc…
Great idea, why didn't I think of it myself...😀
Yeah, I think basically all sample-based sellers will need to find a new revenue stream. Especially when Suno and Udio features start getting integrated into DAWs (and as these music generation systems gain in features, they'll certainly end up replacing a lot of DAWs).
Ah. OK, thanks. So it just generates a new cover at the site instead of allowing you to upload your own art... I guess I can see the reasoning there, so they can avoid any problematic imagery showing up.
@AudioGus
Works much faster for me now, a 33s piece takes about 1m or so.
I don't think that it will affect the sample pack creators because they'll just offer more and sell the curation aspect of it. It will hit studio musicians and those that got hired to produce those pack the most, or artist who live off royalities from sampling.
That said the music industry probably won't sit still. Neither Udio nor Suno reveal their training sources and it's kinda obvious that it's trained on popular artists in each genre and the patterns of their hit songs. If you extend tracks for example it can be quite hard to avoid certain patterns or switch ups from showing up in one section of the song.
I had my first try with Udio and was totally blown away with this track. I asked for an epic 70s style Progressive rock track about the Knights Templar - and that is what I got (although there are a few glitches with the lyrics and the Prog style is maybe more recent than the 70s). I particularly like when the guitars kick in around the 2 minute mark. I would buy this if it was realised by a real band!
As a musician, I still want to play and produce my own music but such tools as this are fantastic idea/inspiration generators - although, sadly I am sure some will start to pass off some of the AI generations as their own work and it is now getting so clever, it can be difficult to tell what is A.I. and what is human made.
Sounds amazing. The vocals it produces are really stunning.
I tried to create 70s styles as well, one funk/soul track I've created reminded me a bit of a more modern version of classic Steely Dan stuff:
Yes, I can hear that too! But amazed at what can be produced with this!
I think it will definitely hit sample pack creators in that it will be flooded. It is happening now in the stock image world.
There don't seem to be any laws about training on copyrighted data powerful enough to inhibit it from happening.
Definetely. I'm now mostly focused on generating material I like that's normally very hard to clear and sample, like certain Jazz labels from the 70s (CTI, ECM etc.). Can't count the Nautilus knock offs it already spit out for me haha.
The AI isn't precise enough yet to really serve the needs of sample pack buyers though. It's more raw material that you can't just drop into a track and it works, you have to generate enough, curate and cut it to shape in order to satisfy the needs of sample pack buyers.
I mean it must be still similar with stock photos, not everything you generate is a winner and the time to sift through the stuff is still time many just don't want to put in, meaning there's a market for those that want to save that time.
But if they come they'll apply retrospectively like the sampling laws. With instrumental music it will be hard or impossible, but voices and styles related to that are low hanging fruit for the music industry.
Yah sound is where Midjourney was two years ago.
For sure. I am not talking decimation, just hyper saturation.
Yes, if by music industry we are talking lawyers making money, they will find a way for sure. But this wont be like current copyright histograms making auto take downs of YT vids. They would be hand picked specific charges against AI/ML service providers that may at some point be legally obliged to list all copyrighted data in their data sets (something that is gaining traction as bills). Meanwhile people will just be making massive amazing mixes of whatever they want using open source tools that can't be autodetected which floods the market and competes against traditional music industry for listeners ears. But it is not like recorded music / streaming route isn't a dead end on the current trajectory anyway. It seems the only 'industry' left will be live shows, which is where it has been limping towards anyway.
You can actually buy the guitar/bass/drums!
Listen to this, first the KI song, then Rush's Xanadu ;-)
It's still not good for very detailed pixel-perfect work like logos or icons. Many designers still buy PSD templates and stock photography instead of prompting an AI and waiting for the right output. You could generate a cardboard texture with an AI, but you're likely better off taking a photo yourself or buying stock textures.
Imo it's the very complex and epic works where AI really shines. Background images or performances where you layer more detailed stuff on top.
If you zoom in on the drums for example, or details in images it becomes obvious that it will take quite some time for AI to reach a level that satisfies the most demanding designers and artists. And if it becomes capable of detailed edits it's likely not going to be easy, kinda like messing with nodes in comfyui. Not everybody wants to spend time doing that.
The thing is that if you're a pack designer or producer and embrace AI you can now use your knowledge to create complex sample material besides the higher detailed stuff like drum samples for example.
From an artist perpective, Instead of licensing the right to a mediocre horror movie soundtrack on Tracklib they may buy a pack of 20 better fitting soundtracks for that money.
But is this a problem for folks who already seek their sources themselves and dig for quality? It's unlikely that those places will flood their stores with low quality stuff. They're rather the ones who will curate the best AI output for their existing audience.
Of course if you're an impulse buyer buying randomly, let's say a coloring book for your kid on Amazon and other ultra competitive stores it's likely that you're seeing 100 AI generated KDP results before one from an actual artist shows up, but that's because those stores don't really curate their offerings. They also sell the visibility in their store to their business audience in form of ads and benefit from leaving it somewhat unregulated.
Yeah, music is already a promotional tool for those that can't live off a large back catalog from the pre streaming days. Many labels I like maybe get a couple of hundred clicks on their videos on YT and Soundcloud and aren't even available on the usual streaming services. These artists make their money as touring DJs and maybe a few bucks on bandcamp from other DJs buying their tracks.
Yes, I can hear the resemblance. It is quite similar but it is not close enough for Shazam to identify that section of the AI song as Rush
When I first heard it, it did remind me of another Prog band but I can't place it - maybe Landmarq or similar?
Right, and it will be 100 times easier for producers to make and provide them, which was my point. I am not talking about how anyone with a need will simply fulfill it directly themselves with AI from scratch in the short term but there will be sooo much more mailable content to pick from.
Right, and there will be tons more of it, even at the highest levels. 3D is also evolving at a rapid pace and the clean up time is going way down. I am currently making finished in game character models in 3 days that took 3 weeks with three different specialists and lots of feedback loops. Now I can generate 3D sketches within a few hours where once 2D images were needed. Generating images with high coherence using 3D tools only adds to this. I think stellar 2D images will be as valuable as a drum loop, a buck on a site sort of thing and with the addition of image Gan like sliders for the purchaser to manipulate them to taste so it is not static with other people having the same loop (image etc).
To be honest I am not sure if we have a specific through line on our discussion heh, but things are being massively disrupted in the graphics world right now and it will only continue. Where things are at in a year, from a market perspective, is entirely unknowable at this point.
As far as I can tell you're more into 3D, game graphics and photorealism. I'm more into 2D, typography and the boring stuff. The graffiti thing is separate, but having that background made things a lot easier.
Logo generators (shape spinners and font selectors) and spec design services lke 99 designs exist for a long time, but they only cover the low end of the design market. They're the ones getting replaced by AI, not actual designers.
Incidentally, did you hear about the integration of generative video features by various competing services into Adobe's Premiere? It's a smart move on their part.
https://blog.adobe.com/en/publish/2024/04/15/bringing-gen-ai-to-video-editing-workflows-adobe-premiere-pro
I see this as a possible next move for both GarageBand and Logic Pro to integrate generative elements. It would be especially helpful if Apple gave musicians and producers an option to generate virtual backup players which would remain responsive to changes made to music as it develops during a recording session, just like real people. This would also solve the recurring problem of trying to find a last-minute musician, sample or loop to fit a missing part during a recording session.
From Harper's Index in the current issue:
"Minimum number of layoffs in the U.S. media sector since January 2023: 26,102. "
"Portion of U.S. human-resources officers who expect AI to replace jobs at their company within the next 3 years: 7/10."
Frankly, projections like this are completely unknowable. New jobs will be created as a result of these services becoming available and some jobs will go away forever. Who could've imagined "YouTuber" could be a career choice ten years ago? The same will apply here.
Currently, maybe for most, but for these specific use cases (logos, icons, textures) there will be simple refined, curated procedural tools that leverage 3D with PBR materials and diffusion rendering etc. to provide any of those. It is likely within a year or two for these niche areas to be covered procedurally for 90% of people's needs. But then yes, what is the larger product it serves? A game? A website? etc. People who adapt to provide those larger products are the ones who make it. People used to write sheet music for music scores and there were conductors, musicians etc involved in making a music score. Now one person at a home studio can satisfy most of this even on feature length films.
I'm more into stylized stuff than photo realism. But stylization is typically extrapolated from a real world base when it comes to automation.
Yah I imagine that will definitely be common place. The new Adobe/Sora stuff is certainly blowing some minds of people that I work with, have yet to try it myself.