Will RIAA’s lawsuit against Udio and Suno really be the win we’re hoping for?AI and music had their first David and Goliath moment this week when RIAA issued a dramatic and scathing lawsuit against generative AI music platforms Suno and Udio.
In the lawsuit, the RIAA claims Suno and Udio are perpetuating copyright infringement “at an almost unimaginable scale”, alleging that both platforms are trained unlawfully on the catalogues of Universal, Warner and Sony. So blatant was this infringement, according to the RIAA, that the lawsuit outlines multiple cases when the notes, melodies and structure are almost identical to existing works, including music from the likes of Green Day, ABBA and Mariah Carey.
READ MORE: Splice CEO’s message for AI sceptics? “Trust the artists”
The 34-page lawsuit has been widely celebrated within the music industry and the broader field of creative arts. Why? Because wholesale ‘scraping’ of data by AI companies without permission or remuneration is a genuine threat — not just to the majors but to working independent artists, all of whom could be undermined if their catalogues are used to fuel models that churn out derivatives at speed and scale.
Uncontrolled and unregulated generative AI poses an existential threat to the music industry as we know it; that might sound hyperbolic but it’s not. These lawsuits could set the tone for how music and AI co-exist going forward and Suno and Udio — whose investors curiously include artists such as 3LAU, Common and will.i.am — will either be sued into oblivion or forced to clean up their act.
A familiar situation
But will the strong-arming become a grudging handshake? It’s not a stretch to assume that licensing deals — and a bit of equity thrown in — are a possible outcome of this lawsuit. Elsewhere in AI, The New York Times’s case against Sam Altman’s OpenAI is likely to conclude the same way – in fact, OpenAI has already signed similar deals with the FT and with News Corp. Few want to air their dirty laundry in court, including the plaintiff, and the stern language used in RIAA’s lawsuit feels more about obtaining leverage in negotiations than expecting Suno and Udio to fork up the roughly $1.5 trillion and $1.36 trillion it would cost them, respectively, to pay the damages due.
So how might they work? A blanket license for carte blanche access to train models on the back catalogues of some of the greatest artists of all time? Another micro-penny payment system if your artist’s name is mentioned in a prompt? A huge annual fee pocketed by the label and thrown on the pile, never to trickle down to the artists whose work makes up their vaults? These aren’t just whimsical hypotheticals — decisions made now could radically alter the upcoming decades of music industry economics.
Streaming has gutted the middle class of artists, shifting rewards to those who own catalogues at scale — micro-pennies mean little to those with three albums, but can mean a lot to labels with 30,000. AI, too, relies on scale: the more quality data you hoard, the more valuable your asset for licensees. Once again, this leaves independent artists sidelined with little voice or influence over the emerging tech that can define their future. And if you’re not at the table, you’re on the menu.
The long-term effect of decisions made at the dawn of streaming isn’t just affecting independent musicians, though; it’s coming back round to the majors themselves. As Billboard’s Elias Leight pointed out last year: “It’s common to hear grumbles about young acts who have hundreds of millions of plays of a single but can’t fill a small room for a live performance.” By opening the floodgates, streaming removed the bridge between artist and listener, and music’s value (as a commodity, at least) has plummeted to next to zero. Those decisions were made against the backdrop of piracy’s genuine threat — and here we are again.
Getting money upfront and a slice of the pie will likely be appealing, but the mistakes of the streaming era must be at the forefront if and when these licensing negotiations begin.
Move fast, break things, get paid
The tech industry has a long and tedious reputation as a disruptor, a badge it wears with honour. In an interview with MusicTech, Grammy-nominated producer, composer and developer BT said: “Our large label music partners told us a story about a CEO that came to see them [and] they [were] clearly in violation of training on IP-protected works just to speed-run a product to market. They asked how they trained and he said, ‘We would rather ask for forgiveness than permission’… This kind of thinking and irresponsibility could destroy music.”
The ‘move fast and break things’ narrative might feel irresponsible, yet it’s exactly the approach that led Suno to raise $125m to date. In fact, one investor audaciously claims that if deals with the labels were in place, they “probably wouldn’t have invested in it. I think that they needed to make this product without the constraints.” Viewing fundamental IP rights as constraints gives you a sneak peek into a worryingly common mindset in tech, but the music industry keeps falling for the Shiny New Thing.
If companies who steal, blitzscale and ask for forgiveness later are constantly rewarded, where’s the incentive to do things the right way from the start? ‘Ethical AI’ has become its own buzzword, with innovative companies rightly pushing for transparency in training data, traditional rights attribution and new technical solutions to allow AI and music to scale together fairly. But as Tatiana Cirisano wrote in MIDIA last month, if “one music-tech startup seeks permission, it risks losing the race to another startup that asks for forgiveness.”
UMG, Sony and Warner have all issued their AI guidelines and creeds, from the Human Artistry Campaign, to government lobbying, opt-out letters and AI for Music initiatives. If labels enter negotiations with AI companies who infringe on an “almost unimaginable scale”, they could risk undermining their own necessary, crucial guidelines as AI finds its feet, and trust, among creatives and artists.
Maybe it won’t happen. Maybe the major labels will seek to make an example of Suno and Udio, the case will make it to court and a novelty-sized cheque will be handed over to the RIAA. If not, and a settlement is reached with licensing terms attached, getting it wrong risks causing significant damage to an already faltering industry. The mistakes of the streaming era are still in the rearview mirror — let’s hope the majors look back before they look forward.
Read more about AI in music technology. 
The post Will RIAA’s lawsuit against Udio and Suno really be the win we’re hoping for? appeared first on MusicTech.

#ICQ

Correction #Plugins for creativity in #musicproduction and #SoundDesign

12 best stem separation software for vocals, rankedStem separation has long been a fascination in the audio world, the end goal being able to split a mixed track into separate parts without any noticeable glitches or audio artefacts. A few years ago, it felt like the stuff of science fiction, but more powerful computers and access to machine learning datasets have turned it into a reality that’s only getting better.

READ MORE: The best DAWs for music producers in all genres, styles and workflows

Which is the best stem separation software?
To find out which stem separation software is best on vocals, we’re using the track Restless Mind by Steven Beddall.
In all cases, the highest-quality configurations were chosen on the software. These examples should give you a rough idea of the stemming quality, but the results can vary from track to track. It’s possible that one algorithm may perform slightly better than another when fed a certain audio, but the results could be reversed with a different track.
Here’s what to listen for when comparing the results:

Do any other instruments leak into the vocal stem at any point?
Are there any timbral changes? Does the algorithm get confused with other instruments and take some information away?
Are natural reverb tails included?
How does it handle backing vocals?
Are there any additional artefacts like metallic whistles, glitches, hiss or phasing issues?

After carefully analysing the acapellas, and cross-referencing with a few other examples, we found the results from Gaudio Studio to be the cleanest overall — a surprise, given that it’s one of the cheapest.
Gaudio Studio provided an effective reduction of noise and instrument spill, a clean capture of backing vocals and reverbs, and the most consistent high-end frequency content.
If you’re willing to put the time into experimenting with the various models, then you can probably get similar results with the free Ultimate Vocal Remover 5. After that, Moises, Lalal.ai, AudioStrip and Logic Pro were all very impressive, with each offering variations on the amount of reverb and backing vocals captured.

Best stem separation software, ranked

Gaudio Studio
Ultimate Vocal Remover 5
Moises
AudioStrip
Logic Pro
Lalal.ai
Hit’n’Mix RipX
Akai MPC Stems
FL Studio
Serato Sample
iZotope RX11
Acon Digital REMIX

What do you think? Have a listen to the files and see which you think sounds best. Each option has different pros and cons including processing speed, price and workflow, so the final decision isn’t necessarily black and white. Different tracks could result in a different leaderboard. Whichever one you go for, we can all agree at least that the technology is mind-blowing — and it’s only going to get better.
Note that you should always respect the copyright of any original material you use these tools on. Also, some of these tools will process the audio using their own servers, so an internet connection is required and the waiting times can vary.
Stem separation software available in 2024
iZotope – RX11
iZotope RX11. Image: MusicTech
Music Rebalance is a feature inside of RX11 that uses a tweaked version of Spleeter to split stereo audio into Vocals, Percussion, Bass and Other stems. You can then change the volume of each to rebalance a track, or solo an element and bounce it to a new audio file. You also get controls for each stem to edit the strength of the separation and three different quality settings.
It comes as part of RX11 Advanced, but you can get it cheaper inside of RX11 Standard, and both go on sale regularly. Music Rebalance has had an update for RX11 and features upgraded machine learning based on modern neural networks.
Price: £389 RX10 Standard, £1,149 RX10 Advanced
VST3, AU, AAX, AAX Audiosuite Plug-in
Separation options: Vocals, Percussion, Bass, Other
Library/Algorithm: Spleeter algorithms implemented in iZotope’s own code
[products ids=”4PfP5kzaOHCiAHGOnI5IKf”]
Acon Digital – Remix
Acon Digital REMIX. Image: MusicTech
Remix is the only option here that processes audio in real-time in your DAW. It runs as a plug-in (or inside of Acon Digital’s Acoustica Suite) and features a low-latency algorithm from HANCE.
You can split your audio into Vocals, Piano, Bass, Drums, and Other stems and then balance the volume or solo each. You also get sensitivity controls to fine-tune each stem, plus the ability to route the output of each channel to its own track.
It may not have the competitive quality as some of the offline options, but it might prove useful for quickly rebalancing stem volumes within your DAW.
Price: £39.90 (also comes bundled with Acoustica Standard Edition for £49.90)
VST, VST3, AU, AAX plug-in
Separation options: Vocals, Piano, Bass, Drums, Other
Library/Algorithm: HANCE Realtime Stem Separation
[products ids=”35L1YdV0zj4SNhR7iQYvjT”]
Hit’n’ Mix – RipX DAW
Hit’n’Mix RipX DAW. Image: MusicTech
RipX DAW and RipX DAW Pro are based on the unique Rip Audio format, which deconstructs audio down to its individual elements and treats audio and MIDI as one. You can split audio into stems for voice, drums + percussion, bass, other sounds + instruments, and guitar + piano. The beauty of RipX is that you can further refine the detection to get the best results, and even edit the harmonic and unpitched elements separately.
Other features include Melodyne-style note editing, the ability to change and randomise the key or generate new harmonies, adding effects to individual notes, replacing any note with a different sampled instrument, advanced repair and cleanup, and more.
Price: £99 RipX DAW, £198 RipX DAW Pro
Standalone
Separation options: Voice, drums & percussion, bass, other sounds & instruments, and guitar and piano
Library/Algorithm: Ripper Engine + Modified Demucs
[products ids=”5rZcis5Lt4B0USFvT3xhDY”]
Omnisale – Lalal.ai
Lalal.ai. Image: MusicTech
Lalal comes as both a web-based and desktop/mobile app stem separator. You can try several files for free, or choose from a range of one-off fees that get you a set number of processing minutes. This is a great alternative to monthly payment plans if you think you’ll only occasionally want to use the service.
The latest version is based on a state-of-the-art, in-house developed neural network called Orion, which uses direct synthesis to recreate stems with exceptional quality. Currently, you can extract Vocals, Backing Tracks, Drums, Piano, Synthesizer, Acoustic Guitar, Electric Guitar, Strings, and Wind, making Lalal one of the most versatile options in the list.
Price: £18 for 90 minutes, £70 for 500 minutes (other options also available)
Web-based, plus desktop iOS and Android apps
Stems: Vocal and instrumental, Voice and Noise, Drums, Bass, Piano, Synthesizer, Acoustic Guitar, Electric Guitar, Strings, Wind
Library/Algorithm: Orion/Phoenix
www.lalal.ai
Serato – Serato Sample
Serato Sample. Image: MusicTech
Serato Sample is a refined sampler with a lightning-fast workflow. It features Serato’s renowned Pitch ’n Time algorithm for stretching audio, and the recent V2 update added built-in stem separation.
You simply load an audio file, and Serato Sample will automatically work out the bpm and key, and will then sync it to your host tempo. You then get access to four buttons for vocal, bass, drums, and music/other parts. These can then be switched on and off at will whilst you trigger samples in the normal manner.
The separation isn’t the best, but being able to import audio into a plug-in, and then instantly get to work chopping and composing is a bonus for workflow.
Price: $149 (or $9.99 pm subscription
VST, VST3, AU plug-in
Stems: Vocal, Drums, Bass, Other
Library/Algorithm: Serato’s proprietary machine-learning algorithm
[products ids=”7IRpldrXCwCZhZBVO8kj9b”]
Image-Line – FL Studio 21.2
Image-Line FL Studio 21.2. Image: MusicTech
Image-Line added built-in stem separation in a recent update to FL Studio. Alongside the extensive production features of the program, you can now right-click on a piece of audio and select Extract Stems from Sample. You then get a dialogue box where you can select from Vocals, Drums, Bass and Other Instruments, and it gives an estimate of how long the processing will take.
FL Studio was the first DAW to include this feature and the fact it’s all built-in is a massive workflow boon. Logic Pro has recently followed suit, and hopefully, it won’t be long until Ableton Live and other DAWs jump onboard with their versions.
Price: £164 Producer Edition, £245 Signature Bundle, £409 All Plugins Edition
Standalone DAW
Stems: Vocals, Drums, Bass, Instruments
Library/Algorithm: Customised Demucs
[products ids=”2wQk1OAYJfqgpMuc5G5BXR”]
Gaudio Studio
Gaudio Studio. Image: MusicTech
Up until recently, Gaudio Studio was a completely free online splitter that used the Gaudio Source SEParation (GSEP) model to separate audio into Vocal, Drums, Bass, Electric Guitar, Piano, and Other Instruments. The website and services have recently refreshed, so you can now purchase bundles of minutes for a very reasonable price.
You can also try out 20 minutes of audio for free with an MP3 format output, or if you upgrade then you get the option of a fast-track service and WAV files.
As mentioned above, this is the most convincing option for separating vocal stems.
Price: $7 for 50 minutes, $16 for 200 Minutes, $50 for 1000 minutes
Online
Stems: Vocal, Drums, Bass, Electric Guitar, Piano, and Other Instruments
Library/Algorithm: GSEP (Gaudio source SEParation)
https://studio.gaudiolab.io
Moises.AI
Moises. Image: MusicTech
Moises comes as a slick web application, plus a desktop and mobile app. The pro version lets you separate audio into vocals (with separate backing vocals), drums (including individual elements), bass, guitar, background vocals, piano, and strings, making this a versatile option. It also has the ability to separate multimedia tracks into dialogue, soundtrack and effects.
If you want to enhance your workflow, then there’s also a Stems plug-in available as part of the Pro Plan, with Voice Studio and Mastering plug-ins coming soon.
The free Starter plan will get you 5 audio separations per month, with the other plans offering unlimited separations, higher quality and a host of bonus features. Moises also updates its algorithms fairly regularly.
Price: Free, £4.99pm Musician Plan, £24.99pm Producer plan (discounts for paying annually)
Online, Desktop app, mobile app
Stems: Vocals, backing vocals, drums (with separates), bass, guitar, background vocals, piano, and strings
Library/Algorithm: Moises AI Proprietary Stem Separation Model
www.moises.ai
AudioStrip
AudioStrip. Image: MusicTech
AudioStrip is a high-quality online option that uses the results of both the MDX-Net and Demucs libraries. The company also notes on its website that the algorithms will be updated as and when new methods come out.
When you upload a file, you can choose from three different algorithms, and then separate out to vocals, instrumental, bass, drums, other, piano and guitar.
You can choose from a basic free option that includes 3 isolations, 3 masters, and 3 transcripts per month, with a slower isolation speed, limited file size and MP3 output format. Or alternatively, the Premium option gives you unlimited uploads, faster isolation, larger files, WAV, FLAC and MP3 formats and batch uploads.
Price: Free / £7.99pm Premium
Online
Stems: Vocals, Instrumental, Bass, Drums, Other, Piano and Guitar
Library/Algorithm: DemucsV4, VB-Splitz V3, VB-Splitz V4
https://audiostrip.co.uk
Apple Logic Pro 11
Apple Logic Pro 11. Image: MusicTech
One of the front-running new features in Logic’s latest update is its built-in stem separation. You can right-click on any audio region and then choose whether to separate it into Vocals, Drums, Bass and Other instruments. It’s a feature that uses the grunt of the Apple Silicon chips, so it’s only available for more recent computers. However, the offshoot is that it’s incredibly fast to process.
As Apple is always incredibly tight-lipped about what goes on behind the scenes, it’s unlikely we’ll ever find out exactly what models and libraries they’ve been using. Suffice to say that the quality isn’t the best in show, but it is still impressive.
Price: $199 (Stem separation requires Apple Silicon chips)
Standalone DAW
Stems: Vocals, Drums, Bass, Instruments
Library/Algorithm: Unknown
[products ids=”4cOxjY9hweLJpjNiHjUXjk”]
Akai MPC Stems
Akai MPC Stems. Image: MusicTech
Akai MPC Stems can be purchased for £9.99 as an add on for the MPC 2 software. This is a feature-packed DAW that runs on your desktop, but it can also be used in Controller Mode to seamlessly sync to the full MPC hardware lineup. The ability to use MPC Stems without the need for a computer will be coming to certain hardware units at some point in the future.
Whether you’re using the interface on the software or the screen on the hardware, you simply select the audio, hit the Stems button and then select which of the vocals, bass, drums and other stems you’d like to create.
Price: £9.99 for MPC Stems, plus £199.99 for MPC2 software (comes free with hardware)
Standalone
Stems: Vocals, Bass, Drums, Other
Library/Algorithm: zplane STEMS PRO
www.akaipro.com
Ultimate Vocal Remover 5
Ultimate Vocal Remover 5. Image: MusicTech
If you want to delve deep into using a range of different models, then you might want to check out the free Ultimate Vocal Remover 5. With this standalone program, you can choose from a wide range of free algorithms, with more downloadable as they become available. You can even use an Ensemble mode to run your audio through two models to get improved results.
The only downside, is that it’s not very user friendly, and the processing can take quite a long time with certain settings. It’s also a little overwhelming as it’s not obvious which settings will yield the best results, although you can check mvsep.com for separation quality leaderboards of results from other users.
Price: Free
Standalone
Stems: Vocals, Bass, Drums, Other
Library/Algorithm: Multiple Demucs, MDX-Net and VR Architecture models
www.ultimatevocalremover.com
The post 12 best stem separation software for vocals, ranked appeared first on MusicTech.

New multiband saturation #plugin from the creators of Gullfoss. Need to test, it can be really good. #Musicians #Producers #musicproduction

Google’s AI can now create an “unlimited number” of soundtracks for videoGoogle’s AI research lab, DeepMind, has shared an update on the development of its video-to-audio (V2A) technology, a process that makes synchronised audiovisual generation possible.
According to Google’s AI page, V2A “combines video pixels with natural language text prompts to generate rich soundscapes for the on-screen action.” DeepMind says the technology can “understand raw pixels”, allowing for the real-time generation of video soundtracks.
READ MORE: EU AI Act explained: What does it mean for music producers and artists?
Additionally, the feature uses text prompts to generate rich audio soundtracks. Together, Google’s V2A technology supports the creation of dramatic scores, “realistic sound effects or dialogue that matches the characters and tone of a video.”
Google’s new AI feature means “enhanced creative control” meaning V2A technology can create an “unlimited number” of soundtracks.
While the tech company’s update on AI sounds creatively promising, there is still some progress to be made until the tool is rolled out to the general public.
The development of V2A is “improving”, however, it will be some time before the update will be widely available: “Still, there are a number of other limitations we’re trying to address and further research is underway,” the blog post reads.
The statement continues: “Since the quality of the audio output is dependent on the quality of the video input, artefacts or distortions in the video, which are outside the model’s training distribution, can lead to a noticeable drop in audio quality.”
Watch example clips of Google’s new V2A tool in action below.

In related news, Google recently came under fire from Sony Music, after the label raised suspicions that Google — along with OpenAI, Microsoft and over 700 other tech firms — have made “unauthorised uses” of Sony’s music catalogue to train AI. The public letter was titled ‘Declaration of AI Training Opt-Out’, and reads:
“We support artists and songwriters taking the lead in embracing new technologies in support of their art. Evolutions in technology have frequently shifted the course of creative industries. AI will likely continue that long-standing trend.
“However, that innovation must ensure that songwriters’ and recording artists’ rights, including copyrights, are respected. For that reason, SMG’s affiliates, Sony Music Publishing (SMP) and Sony Music Entertainment (SME), on behalf of themselves and their wholly owned or controlled affiliates, are making this affirmative, public declaration confirming that, except as specifically and explicitly authorized by either SME or SMP, as the case may be, each of them expressly prohibits and opts out of any text or data mining, web scraping or similar reproductions, extractions or uses.”
Learn more music technology news.
The post Google’s AI can now create an “unlimited number” of soundtracks for video appeared first on MusicTech.

5 creative ways to share your music using QR codes. #Music #artists #Producers #Promotion #Musicians

Universal Music signs deal to train AI to make “ultra-high fidelity vocal models” of its artistsUniversal Music has inked a deal with artificial intelligence start-up SoundLabs to create “official ultra-high fidelity vocal models for artists using their own voice data for training while retaining control over ownership and giving them full artistic approval and control of the output”.

READ MORE: From Kraftwerk to AI – Berlin’s love affair with music technology

The deal means that singers on Universal’s roster can train AI models using their own voices and retain ownership of the results. They will also have full control over how they’re used. In theory, this could also mean they could record a new album without actually singing a single word themselves.
In a press release, Universal Music say: “It empowers artists and producers to explore bleeding-edge vocal transformations, including voice-to-voice, voice-to-instrument, speech-to-singing, language transposition, and a myriad of previously impossible vocal transformations. Together, UMG and SoundLabs are collaborating to allow UMG artists to create custom vocal models that will be available for their exclusive creative use cases, and not available to the general public.”
“It’s a tremendous honour to be working with the forward-thinking and creatively aligned Universal Music Group,” adds SoundLabs founder BT, who famously worked on the 1986 hit Blue Skies, featuring Tori Amos. “We believe the future of music creation is decidedly human.
“Artificial intelligence, when used ethically and trained consensually, has the Promethean ability to unlock unimaginable new creative insights, diminish friction in the creative process and democratise creativity for artists, fans, and creators of all stripes.
“We are designing tools not to replace human artists, but to amplify human creativity.”
Although the use of AI in music remains controversial, its use is accelerating fast; almost two-thirds of young creatives are embracing AI in their music making, according to a recent study.
The post Universal Music signs deal to train AI to make “ultra-high fidelity vocal models” of its artists appeared first on MusicTech.

UMG partners with SoundLabs to launch AI vocal plug-in MicDropUniversal Music Group and SoundLabs are launching MicDrop, a groundbreaking AI vocal plug-in. Find out how this new tool will change the way artists create music. via Celebrity Access SoundLabs,. Continue reading
The post UMG partners with SoundLabs to launch AI vocal plug-in MicDrop appeared first on Hypebot.

REWIND: New Music Industry’s Week in ReviewIt was a busy week by any definition, and the music industry was no exception. There were big accusations regarding Merchbar, music publishers’ fight with Spotify escalates, action on group. Continue reading
The post REWIND: New Music Industry’s Week in Review appeared first on Hypebot.

Niche is mainstream #MusicIndustry and mainstream is niche #Music #Producers #artists #Midia

What does Suno AI mean for music producers and the music industry?Sharooz is an electronic music producer, studio owner and entrepreneur. He’s also known as Principleasure and is the founder of Wavetick.
Whether a producer or songwriter, it’s impossible not to feel some emotion around the hyped generative music startup Suno. Especially in light of its recent $125 million funding — the biggest music tech equity investment in over three years.
How we create music and the potential to earn revenue from our skills may be about to change forever.
If you haven’t already played with it, Suno is fun and powerful. Like a ChatGPT for music, it creates unique songs based on a simple text prompt – and does so with impressive, albeit generic, accuracy. Vocals sound realistic, even guitar solos and string sections are spliced together with a nuance rarely seen before in generative AI music. In just a few years, AI has advanced from dodgy, artifact-riddled soundalikes to a personalised jukebox capable of spitting out songs that could probably sit unnoticed in the Billboard Top 100.
READ MORE: Learn how to create custom voice and instrument audio stems with AI
The potential to damage virtually every aspect of the music industry is obvious. While virtuosic composers and experimental curators of their craft may have little to worry about, Suno could conceivably chisel away at the stock music industry, sound designers, foley creators, lyricists and the work of songwriters in virtually every genre. This could be particularly true for those who practice more traditional arrangements and chord structures, like those commonly seen in charting pop songs.
I’d like to think that organic human emotion and the poetry of heartfelt lyricism will transcend anything a machine can offer. But it’s not inconceivable that, in the space of a few years, AI output may be indistinguishable from human endeavour, especially to the untrained ear. After all, Suno is a mere glimpse at what may be possible in the near future.

Suno’s public message offers utopian promises of “moving the bar” of music creation. It’s clear the Massachusetts-based company has plans to disrupt, with the online discourse opining that the wider mission is to fully remove the barrier between music creation and casual listening — imagine personalised playlists made up of fully unique AI-generated songs, fuelled by user prompts.
If these services are creating a future where the music creator and listener become one, this gives real potential to disrupt DSPs, labels, aggregators and everything in between. At the time of writing, Suno recently announced it plans to pay the platform’s most popular “creators” $1 million in “prize money” during June 2024.
To grasp Suno’s impact, one needs to understand how their output has become so much more polished than anything else that’s come before. AI is traditionally fed on real recorded music — human-created intellectual property (IP) with complex copyright restrictions. In theory, the more ‘data’ the network can train on, the more realistic the resulting output can sound.
Nobody is quite sure of the data Suno is trained on, but keen listeners have already identified scrambled elements of distinguishable works in their creations.
Public details on training data are scant, with many suggesting there could be lawsuits from major publishers and labels in the offing. Sony Music recently sent 700 letters to leading generative AI firms warning them not to infringe their copyrights. But if the current landscape of the music industry has taught us anything, it’s that there’s no guarantee disruptive technologies will favour human creator rights or livelihoods.
The dominance of digital streaming platforms (DSPs) has only diluted existing songwriter and performer revenue further. There’s an ongoing conversation on the unfair economics of streaming, with commercial law slow to catch up on AI’s impact on existing copyrights and publishing rights.
It’s not inconceivable labels may soon license our recordings and songs into Suno by the truckload. When those deals are done, they may net you less than any DSP currently does: fractions of cents. Will the majority of subscription revenue Suno generates line the pockets of its investors and the major labels that could one day own a share in it, if or when it goes public? After all, the investors in it will be keenly expecting a return, such is the nature of venture capitalism.
To create this technology is an awesome feat. Suno sounds remarkable. It’s fun, powerful and easy to use. To embark on this journey by going a step further, disclosing training sources and collecting metrics, directly compensating dataset contributors would be a welcome play…But I don’t suppose that pays investors well.
We all want to make technology more accessible, but we won’t get there by powering our product on the work of writers who have yet to even be acknowledged, let alone compensated. Creators may be in real danger of being squeezed out of operating altogether.
Brand trust begins with responsible practice. High-profile artists will likely boycott and vilify Suno — we saw something similar with the SAG/AFTRA strikes in 2023. Suno’s millions could aid its legal challenges but could set an uneasy precedent — steal now, seek permission later [producer BT told MusicTech a similar anecdote].
Will future legislation render AI music services a gimmick unfit for public broadcast or distribution?
User terms for most generative AI services are also unacceptably vague. We see “Use at your own risk” through to “your creation is uniquely your copyright”, with little comfort for pro/broadcast use or publishing to a DSP. Licensing a bona fide, human-created sample or track may be far more beneficial than hours spent prompting an AI-generated output that’s legally unfit for purpose.
Is there an exaggeration of AI’s impact on the music industry? If Suno are to be believed, we may evolve into an entirely new class of creators, further democratising music making. After all, why settle for spoon-fed major label playlists when you can just roll your own bespoke experience? But the real threat to professional creator livelihoods seems a long way off.
The winners in this space will likely be the well-researched, steady adopters who subtly integrate AI-assisted features to aid their existing creative processes. They’ll meaningfully democratise access for tomorrow’s creators without stealing from humans, whose work has been an endeavour of learned skill, political upheaval, emotional intelligence and the very meaning of what it is to be human. They’ll translate their authentic experiences through music.
Read more music technology features. 
The post What does Suno AI mean for music producers and the music industry? appeared first on MusicTech.

Droplets is an innovative sequencer application/plug-in equipped with a built-in synth engine and an unusual interface.
#musicproduction #PluginsNews #Synthesis

ICQ era is over. I remember the days when I used dial-up modem: public chat boards, ICQ messenger, mp3 file download whole night, etc. Internet was different those days. I have met really great people

Tencent Music’s AI-powered tech can ‘predict the next hit song,’ and 5 other things we learned from its latest annual reportTME is also facing 695 copyright infringement lawsuits, and it's negotiating a 10% stake in an 'overseas entertainment company'
Source