AI-powered stem splitter’s Andromeda model interprets audio tracks with near-human precision.
ZUG, SWITZERLAND, December 8, 2025 (Newswire.com)
– LALAL.AI has officially launched Andromeda, its most advanced audio separation model to date, that not only performs faster, processing audio up to 40% quicker than its predecessor, but delivers more accurate stem separation, reducing distortion and improving clarity.
Andromeda Reduces the Need for Manual DAW Cleanup
Rather than simply processing sound as a single waveform, Andromeda takes a more nuanced approach, analyzing audio in terms of time, frequency, and tone. This allows it to separate complex layers of sound with greater precision, making manual cleanup in a DAW largely unnecessary.
“In practice, this means that even the most delicate elements, like soft backing vocals or subtle instrumental details, come through with newfound clarity,” says Nik Pogorski, LALAL.AI Product Lead. “What once required meticulous post-processing can now be achieved directly with Andromeda’s enhanced capabilities, saving users both time and effort.”
In the past, LALAL.AI users faced a trade-off: choosing Clear Cut for cleaner stems but less detail, or Deep Extraction for finer detail but more bleed between tracks. Andromeda removes this compromise entirely. Now, users can extract rich, detailed stems without worrying about cross-bleeding, getting clean and precise results in a single pass.
The launch of Andromeda extends to several LALAL.AI web services, ensuring API clients too can harness its power:
Stem Splitter (for Vocal and Instrumental + Voice and Noise stems)
Lead & Back Vocal Splitter
Echo & Reverb Remover
Voice Cleaner
Soon, Andromeda will be available in the LALAL.AI’s first VST plugin, allowing producers and audio engineers to perform seamless vocal and instrumental isolation directly within their DAWs.
Speed Meets Precision: LALAL.AI’s New Model Enhanced Performance
Andromeda processes audio tracks up to 40% faster than its predecessor, which allows for quicker workflows and less waiting time, benefiting professionals who need efficient processing for large volumes of content.
The new model shows a 10% increase in SDR compared to its predecessor.
The frequency range remains consistent at up to 22 kHz with stereo processing, providing high-fidelity separation the industry expects. Andromeda handles challenging audio scenarios and complex instrument mixes with greater precision, reducing artifacts and increasing consistency.
Thanks to advanced DSP (Digital Signal Processing) techniques, Andromeda delivers consistent separation across quiet or loud tracks, letting creators and audio engineers work across different types of audio sources without worrying about volume inconsistencies.
With Andromeda, LALAL.AI has set a new standard for audio separation, providing both professionals and enthusiasts with a tool that offers unparalleled precision, speed, and clarity. Whether you’re a content creator, producer, or audio engineer, Andromeda is the next generation of stem separation that will take your projects to the next level.
For more information or to try Andromeda yourself, visit LALAL.AI.
From Daft Punk to Fred Again, Avicii to Peggy Gou: a two-decade journey through festival anthems, stitched together with the help of AI.
ZUG, Switzerland, June 3, 2025 (Newswire.com)
– In a tribute to two decades of dancefloor anthems, the masked Italian DJ duo, DJs From Mars, have released a 13-minute mega-mashup featuring 100 of the biggest festival tracks from 2005 to 2024. From The Chemical Brothers’ Galvanize to Peggy Gou’s Nanana, the duo slices through time with surgical precision, and one tool helped stitch it all together: LALAL.AI.
Known for turning pop and EDM into interstellar sound collages, DJs From Mars say the key to this project was clean, high-quality acapella and instrumental extraction, a long-time struggle for mashup artists. AI changed that.
“We always struggled to find good quality acapellas,” they said. “Now we can extract clean vocals from basically every existing track. That’s a game-changer for our mashups, our live sets, and our studio production. We tried LALAL.AI, and BOOM! It was love at first sight.”
The mashup, which features tracks from Daft Punk, Swedish House Mafia, Fatboy Slim, Calvin Harris, Avicii, and over 90 more artists, has already caught fire on social media, tapping into collective nostalgia while showing how AI is transforming music creation in real time.
The mashup is a perfect showcase for how AI tools like LALAL.AI are opening new creative workflows for DJs, remixers, and producers, making it possible to rework and reimagine music in ways that were previously unthinkable without access to official stems.
“Seeing artists like DJs From Mars use our tech to reinterpret the sound of the last 20 years is exactly what excites us: not just innovation for its own sake, but tools that unlock new forms of creative expression,” said a spokesperson for LALAL.AI. “This isn’t just a fun mix; it’s a case study in how DJing and remix culture are evolving. Once limited by what you could buy, borrow, or rip, artists now have a powerful toolkit in AI audio separation. For mashup artists especially, it’s a revolution.”
Some of the featured tracks include:
The Chemical Brothers – Galvanize (2005)
Justice vs Simian – We Are Your Friends (2006)
Swedish House Mafia – One (2010)
Skrillex – Make It Bun Dem (2012)
Avicii – Levels (2011)
Fisher – Losing It (2018)
Fred Again & Swedish House Mafia – Turn On The Lights Again (2022)
Peggy Gou – (It Goes Like) Nanana (2024)
Shouse – Love Tonight (2017)
The full list spans 100 tracks across EDM, house, techno, D&B, and pop crossovers – all seamlessly blended together into one high-energy festival flashback.
Known for their viral mashups and bootlegs (and signature cardboard box helmets), DJs From Mars have built a global following with their genre-hopping, high-energy performances and creative edits.
About LALAL.AI LALAL.AI is a next-generation stem separation service that uses neural networks to extract vocals, instrumentals, drums, and more from audio and video. Trusted by producers, DJs, dubbing professionals, and creators worldwide. EDM.
LALAL.AI introduces its highly anticipated Lead & Back Vocal Splitter, offering music producers, remixers, and creators the ability to separate lead and backing vocals with unparalleled precision.
ZUG, Switzerland, October 9, 2024 (Newswire.com)
– After the success of their initial product in 2020, which focused on splitting songs into vocal and instrumental stems, LALAL.AI has responded to user feedback by expanding its capabilities. The Lead & Back Vocal Splitter is now available, providing users with the ability to isolate lead and backing vocals as two separate tracks from any audio or video file.
“We’ve been listening closely to our users, and many of them asked for a way to separate lead and backing vocals. After perfecting our AI technology, we’re thrilled to introduce this dedicated product,” says the LALAL.AI team. “This tool is easy to use, offers uncompromising quality, and opens up new creative possibilities for anyone working with vocal stems.”
A Tool for Every Creator
LALAL.AI’s Lead & Back Vocal Splitter utilizes advanced AI algorithms to accurately extract vocal layers from audio files. Whether you’re remixing, producing, or teaching music, the tool allows for independent manipulation of lead and backing vocals, making it easier than ever to create remixes, backing tracks, or training materials.
The tool works best on tracks with clearly defined vocal roles and can handle songs featuring multiple lead vocals. However, users may experience variable results with tracks where vocal separation is less distinct, such as those with synthetic or heavily processed vocals.
LALAL.AI Lead & Back Vocal Splitter is now live and available for use, set to transform how artists and producers approach vocal isolation, giving them greater control and flexibility over their creative projects.
LALAL.AI is a leading AI-powered audio technology company that specializes in vocal and instrumental separation tools. Launched in 2020, the platform has quickly gained a reputation for its accuracy and ease of use, empowering musicians, producers, and audio professionals worldwide to explore new creative possibilities.
LALAL.AI’s new Perseus AI uses technology similar to OpenAI’s ChatGPT model for vocal removal.
ZUG, Switzerland, September 26, 2024 (Newswire.com)
– LALAL.AI is to announce the launch of Perseus AI, an advancement in vocal removal that leverages the power of the technology akin to OpenAI’s ChatGPT model. As one of the first transformer models in audio processing, Perseus AI is set to redefine how users create instrumental versions of songs and acapellas.
Originally introduced by Google and widely utilized in applications like large language models, transformers are now revolutionizing audio processing too, with LALAL.AI being one of the pioneering solutions that leverages it for vocal isolation.
The core of Perseus AI features an entirely new architecture, distinct from its predecessors. Although the training procedures mirror those used for the Orion network (previous LALAL.AI model), the development of Perseus demanded an average of 3.5 times more resources, which resulted in a 15% enhancement in vocal extraction quality.
What does it mean for musicians, video creators, producers, and businesses that leverage vocal stem separation in their workflows? The vocals users isolate from a song or to clear background noise have become drastically clearer, containing fewer artifacts and infiltration of other stems. Such quality improvement will unequivocally result in a quality boost of their business or creative projects, be it a remix, mashup, karaoke track, podcast, dubbing of a video, or sync licensing a track.
Special Launch Offer
To celebrate the release of Perseus AI, LALAL.AI is launching a time-limited special offer, which is an excellent opportunity to stock up on minutes and save in the long run by purchasing selected LALAL.AI packs with extra minutes added. The packs with extra minutes are available for purchase from September 25 to October 7.
Users can experience the next generation of vocal removal technology with Perseus AI by visiting the LALAL.AI site. Perseus AI is specifically designed for vocal sound extraction, currently available for the “Vocal and Instrumental” stem. The new neural network is enabled by default with no additional steps required.
About LALAL.AI
LALAL.AI is a leading innovator in AI-powered audio processing, dedicated to providing state-of-the-art solutions for music enthusiasts and professionals. The company’s mission is to make high-quality audio editing accessible and effortless for everyone.
LALAL.AI, a leading provider of AI-powered audio manipulation solutions, is releasing Voice Changer, a new tool that makes any voice sound like a star.
ZUG, Switzerland, April 23, 2024 (Newswire.com)
– LALAL.AI, the AI-powered stem splitter renowned for its capability to separate a track into ten stems, announces the launch of its latest innovation: the LALAL.AI Voice Changer. This cutting-edge AI-driven solution allows users to transform any voice into the voices from a select roster of renowned artists.
With LALAL.AI Voice Changer, currently available for free in beta, users can harness the power of artificial intelligence to sing like their favorite artists. By simply selecting a voice pack and applying it to their recordings, users can effortlessly emulate the distinctive vocal styles of their musical idols and check how a star’s voice would fit their mix.
Key features of LALAL.AI Voice Changer:
Access to 16 curated voice packs, featuring world-renowned artists.
MP3, OGG, WAV, FLAC, AVI, MP4, MKV, AIFF, and AAC formats are supported.
Currently available in beta, offering users the opportunity to experience the Voice Changer completely free of charge.
“Our AI meticulously modifies pitch, tone, and timbre, ensuring a natural and high-quality transformation of your singing voice,” said Nikolay Pogorskiy, Lead Engineer at LALAL.AI. “Catering to a global audience, we’re working with a multitude of languages, making it a versatile tool for creators around the world. No matter the language of your track, Voice Changer can handle it.”
In addition to its pre-packaged voice options, LALAL.AI Voice Changer also welcomes collaboration on custom voice packs, allowing users to tailor voices to their unique preferences and creative visions.
To learn more about LALAL.AI Voice Changer and try it for free, visit the LALAL.AI Voice Changer page.
About LALAL.AI
LALAL.AI is a leading provider of AI-powered audio manipulation solutions, specializing in source separation, noise cancellation and voice transformation technologies. Leveraging cutting-edge artificial intelligence algorithms, such as Orion, LALAL.AI empowers users to unlock new creative possibilities in music production and audio editing.
There is perhaps no better person to incorporate into a song called “Nothing Lasts Forever” than Madonna. Not just because she has made a career out of proving that trends come and go, and that reinvention is the only way to survive the fallout of a certain “fad’s” death (even if fads always end up swinging back around—like voguing), but because to stay stagnant is its own form death. Of course, there’s also the more obvious way the song applies to Madonna in that she happened to have a near-death experience over the summer that reminded her (as if she needed to be) of just how fleeting existence can be.
With the concept for the video written and directed by Willem Kantine, the visuals for it accentuate just how much the music video art form has evolved since Madonna’s “heyday,” when far more fanfare and linear conceptualization was put into such endeavors. Now, all an artist needs is a “concept” without much else behind it (save for trippy special effects), least of all a narrative on par with the video for “Express Yourself” or “Bad Girl” (both of which were directed by David Fincher). Even so, the video is actually inspired by 90s-era gabber music, which often featured such displays of underground machismo. So naturally, Madonna is game to contribute her visage (#givegoodface) to the project intended to subvert ideas of what “masculinity” is. And, in her own way, maybe it’s a troll on all the blowback she got a couple years ago for putting her face on another woman’s body. Although the altered image Madonna had posted was from 2015, it took six years for the culture to become outraged about it in our newly evolved state of perceiving everything as a violation (which, truth be told, it kind of is). Madonna’s social media choice brought up a larger conversation about “reality” in the internet age, well before the AI panic of 2023. Not to mention the issue of stripping female bodily autonomy when it comes to AI, deepfakes, Photoshopping, etc. Thus, for a feminist like Madonna, such behavior was a big dichotomy.
In any case, it wouldn’t be unlike M to display a sardonic sense of humor toward that moment by, once again, having her head placed atop another person’s body. This time, a bodybuilder type. Indeed, everyone in the video—Sevdaliza, Grimes, Madonna, Julia Fox and A$AP Ferg—is down to have their physical person warped by this odd “cut and paste” of their head. Better known as: deepfaking. For the most part, it’s Sevdaliza and Grimes who seem to be loosely training for a competition (if all that slow walking side-by-side on a treadmill is any indication). After all, it’s their song. One that combines their shared love of AI in particular and technological manipulation in general. For example, Sevdaliza offered herself up to become the first femenoid robot (named Dahlia) and Grimes has been very open about inviting fans to deepfake her voice via the website she launched to do just that, Elf.Tech. In both instances, each artist has taken a contrasting approach to the new world order compared to other musicians, who view AI and deepfaking as massive threats to art. Rather than fighting it, however, Sevdaliza and Grimes have gone whole hog on embracing it, perhaps adhering to the old adage, “If you can’t beat them, join them.” Mind you, that seemed to be the adage that many a Nazi employed as well.
Madonna, too, has employed this method with technological advances throughout her career, embracing every shift as it comes—from the dawning of the internet (see: her concert at the Brixton Academy being among the first of its kind to get livestreamed) to her creation of a TikTok account. Whatever it might be, Madonna never shies away from the new tech that bombards us because she, like Sevdaliza and Grimes, prefers to use it as a tool rather than view it as a threat.
Granted, it was Grimes herself who ominously foretold that “we’re in the end of art, human art.” And she’s been the first to practice what she preaches by surrendering herself over fully to the matrix (as she basically did when she started dating Elon Musk and then had his children). Madonna, too, has made that surrender, in case you forgot about the NFT she made of herself (in conjunction with Beeple) that shocked the nation. For Madonna, that kind of immortality is precisely what she’s always yearned for. With technological manipulation a person can not only live forever, but they can look young forever, to boot (see also: Madonna’s face as Andy Warhol’s philosophy). Which brings us to a key lyric from “Nothing Lasts Forever”: “I don’t wanna waste my youth (nothing lasts forever)/Love me then let me go.” If that hasn’t been the credo that Madonna has lived by, then nothing is.
To be sure, the themes of “Nothing Lasts Forever” apply not just to Madonna’s own approach to life and the pursuit of fame (which took a toll on many of her personal relationships during her climb to the top), but to humanity itself as we start to reconcile with the notion that maybe our jig is about to be up. Just like it was for the dinosaurs. Whether that refers to our extinction by way of AI or climate change (or both) remains to be seen. But when Sevdaliza sings, “We are machines made for dreamin’/Dreams are done, are dreams dead?/Something inside still believes it,” it only further proves that we’ll die living in the delusion that there’s still a shred of hope left.
Unfortunately, there’s no such hope for music videos as we once knew them in their postmodern prime. Going back to how the medium of the music video has devolved irrevocably since the decades when Madonna was going all in on film-like efforts such as “Like A Prayer” and “Bedtime Story” (and no, Taylor Swift’s “All Too Well” isn’t that), Kantine had this to say of “Nothing Lasts Forever”: “With the rise of TikTok and other trends, the actual time consumers watch videos like ours has been steadily declining. Our project is a response to this phenomenon and challenges the notion that art needs to cater to a specific audience to be successful.”
In other words, we’re living in the era of Just Toss Something Out and See What Sticks/Goes Viral. Madonna, ever the ready adapter, is only too prepared to absorb and wield that trend. One that, hopefully, won’t last forever… then again, maybe it’s better if attention spans, short as they already are, stay the same for as long as possible. Because it can only go further downhill from here.
As talk of AI being the biggest threat to humanity (apart from climate change) since the invention of the atom bomb (also still a threat by the way), one very human aspect of life that’s been made more vulnerable than anything of late is music. More specifically, the wielding of AI to “make” artists sing any song a person wants them to. Hence, AI versions of Britney Spears singing Madonna or Lana Del Rey singing Nancy Sinatra or Billie Eilish singing Olivia Rodrigo, and so on and so forth. And yet, amongst all the negativity about the detrimental effects of this type of technology, a lone positive story to emerge is a resuscitated demo that John Lennon wrote in the late 70s called “Now and Then.” In the mid-90s, the other three living Beatles decided to turn Lennon’s demos into Beatles “reunion” songs for a project called The Beatles Anthology. Unfortunately, at the time, the technology wasn’t available to bring “Now and Then” up to par with the other previously unreleased singles that were included on the album, namely “Free As A Bird” and “Real Love.” And yes, funnily enough, the movie Now and Then came out in 1995 just like The Beatles Anthology.
Luckily, in the wake of Peter Jackson making the documentary for Get Back, he and his team had developed a software system for separating/parsing out audio that they used throughout production. One that, at last, enabled the separation of John’s vocals from the piano on his demo, which was plagued with the cursed 60-Hz mains hum (one far louder than what the remaining trio found on “Real Love,” which had a similar, but more salvageable 60-Hz problem). And, since George Harrison was the one who had written “Now and Then” off as “fucking rubbish” during the first go-around of trying to make it into “something,” there wasn’t much effort put forth in trying to find a method, however fallible, to better the single. As Paul McCartney would then tell Q Magazine (RIP) in 1997, “George didn’t like it. The Beatles being a democracy, we didn’t do it.” But, clearly, that was for the best, as more time had to pass so that technology could catch up with the needs of “Now and Then” and its rough-hewn state. Plus, now that George has been out of the picture since 2001 (having died of lung cancer after surviving a brutal knife attack in 1999), a democracy of two is much easier to work with, and Ringo Starr has never been one to turn down a few extra bob. All of that said, the final product of “Now and Then” is nothing short of gut-wrenching. Particularly when paired with the accompanying music video (also directed by Jackson), awash with equal parts archival footage and what some would call a “nefarious” use of technology in that it revives John and George as, let’s call them, holograms. Younger versions of themselves that perform alongside Paul and Ringo for an effect that’s both eerie and poetic. And an effect that, of course, highlights the “now and then” theme through a contrast of Beatles at different ages.
Alas, Lennon will never be known beyond the age of forty (perhaps something he would call a blessing, likely poking fun at how Paul looks as an “elder” from on high). He is frozen in time just before that tipping point between “middle age” and outright “agedness.” Something about that lends an additional melancholy to the timbre of the song, imagining him writing it in the Dakota in 1977, when he would have been thirty-seven years old…and still relatively fresh from his “Lost Weekend” (from 1973-1974) with May Pang in Los Angeles. Hence, “Now and Then,” framed within its “in real time” context is yet another clear mea culpa directed at Yoko Ono. He couldn’t have known how the wisdom and lament of his words (even then at still such a tender age) would transmogrify in the future, one in which, had he lived, he would have been eighty-three years old.
Although the lyrics were once aimed at being grateful for the salvation Lennon attributed to Yoko’s love, when taken into context as a project that was revived by the last living Beatles, it becomes a song about being appreciative/eternally tied to his bandmates. Thus, lyrics like, “I know it’s true/It’s all because of you/And if I make it through/It’s all because of you” transcend into Lennon’s grand thank you to the band for not only the success they shared together, but its continued ability to reanimate in new and unexpected ways. With Lennon now “making it through” once more because McCartney and Starr have willed it to be so. Indeed, in the official statement regarding the single, it is mentioned, “This remarkable story of musical archaeology reflects The Beatles’ endless creative curiosity and shared fascination with technology.” At least when it came to music and its manipulation. After all, The Beatles were always willing to tinker with their sound, usually courtesy of George Martin—which is how albums like Sgt. Pepper’s Lonely Hearts Club Band, Magical Mystery Tour, The White Album and Yellow Submarine came to fruition as the band was more prone to experimentation after their “teen heartthrob years” of the early to mid-60s.
In the present context, The Beatles’ openness to experimentation has extended into AI technology, perhaps with more willingness than many of the younger musicians (apart from Grimes) that have expressed an aversion to it and what it might mean for the “purity” of one’s artistry. And with The Beatles still being a foremost “tastemaker” and “standard-setter” in the business, it means the floodgate has further opened in terms of embracing rather than bothering to rebuff the use of “cheating” with technology in music. What’s more, in a world that has already surrendered entirely to the ersatz, perhaps The Beatles are aware that “Now and Then” is actually more authentic than most of what gets released in the current landscape. By the same token, it’s easy to dismiss the dangerous effects of technology’s takeover in music when one has come to the end of their life, therefore the end of their musicianship. It’s sort of tantamount to boomers throwing a peace sign up to caring about climate change because they won’t be here for its most severe consequences anyway.
Despite this, there’s no denying that “Now and Then”—billed, definitively, as “the last Beatles song”—will be a comfort across generations beyond the band’s own birth cohort. If time goes on even for another century, it will be as James (John Hannah) in Sliding Doors said: “Everybody’s born knowing all The Beatles’ lyrics instinctively. They’re passed into the fetus subconsciously along with all the amniotic stuff. Fact, they should be called The Fetals.” “Now and Then” has jolted listeners into remembering why, exactly, that is.
Later in the song, still sparse with lyrics beyond the chorus in spite of its “clean-up” (ergo, the intense layering on of additional instrumentation), Lennon and McCartney sing, “And now and then/If we must start again/Well, we will know for sure/That I will love you.” That utterance “if we must start again” coming across as part of Lennon’s acerbic wit, which, in this instance, pertains to being dug up from the grave anew to “be a Beatle.” Yet, since the Fab Four did share such a unique experience together, their forever bond is still apparent even though half of the quartet is no longer with us. So it is that John is able to tell his brethren, from beyond the grave, “Now and then I miss you/Oh, now and then I want you to be there for me/Always to return to me/I know it’s true/It’s all because of you/And if you go away/I know you’ll never stay.” That last line being a peak Britishism/Lennonism in terms of wordplay and the exhibition of a cocksure ego.
With The Beatles bringing back 1995 all over again with this reminder that everything old can be dusted off to be made new, The Beatles Anthology-style, it raises the question of whether or not John and George truly would “consent” to the use of this song. On the one hand, both were extremely “pro-fan,” yet, on the other, each was a meticulous artist who wanted their work to be a certain way. Nonetheless, one would like to believe that this “certain way” would have been sufficient to bring a smile to Lennon and Harrison’s face. If for no other reason than because AI has given them both a last gasp in the music biz.
ZUG, Switzerland, October 17, 2023 (Newswire.com)
– LALAL.AI, a cross-platform AI-powered stem splitting tool, presents Orion, the algorithm that delivers a leap forward in audio separation technology, boasting capabilities in extracting vocals from a mix in a much cleaner and faster way than all existing solutions.
Unlike other stem splitting AI tools, Orion hallucinates the vocals and directly synthesizes them, not relying on a spectral mask that can’t deliver anything more than in the original track, which is a breakthrough approach to audio separation. It means the new algorithm doesn’t just extract vocals from a source, it assumes what a vocal stem is before mixing, and, if it’s not entirely suitable for extraction, recreates the vocal stem.
The peculiarities of the Orion algorithm are as follows:
Orion significantly reduces the phasing effect and distortion in the extracted stems compared to the older generations of the LALAL.AI algorithms.
The results split by Orion are 70% cleaner from various artifacts.
Orion preserves all tonal characteristics of a vocal stem.
Orion can accurately extract unison singing and echoes.
“We spent 20 times more training data on Orion than we did on Phoenix, our previous solution. Besides, in addition to all the differences from our former algorithms quality-wise, Orion is twice as fast as Phoenix, meaning that users will get the split results much quicker, especially in batch processing,” says Lead Engineer Nikolay Pogorskiy.
To evaluate the separation quality, the LALAL.AI engineers use signal-to-distortion ratio (SDR), indicating how many artifacts the separation process introduces to the vocal and instrumental parts. The difference between Orion and Phoenix is 2.5 dB, which is a huge leap in quality in comparison to the SDR differences between the previous algorithms.
Sound producers, DJs, musicians, and other creatives can now test Orion, as well as compare it with the previous algorithm on the LALAL.AI website. The new solution can currently be used for the extraction of vocals, instrumental backtracks, and noise cancellation. More instruments, such as acoustic and electric guitars, drums, bass, piano, synthesizer, strings, and winds can be processed by Orion in the upcoming releases.
To learn more about Orion and listen to the samples split by it, check the LALAL.AI blog article.