Transcript (auto-generated, may contain errors). Hosts are AI-generated. Stories curated by Synaptic Plunge. sound professionally finished like why you have to EQ before you level crucial step then we're getting into why pros are using studio style prompting describing the sound instead of naming stars especially as courts weigh in on artist names oh the legal side is wild right now right and finally we'll look at the legal fine print shaping the actual boundaries of AI and pro workflows there is a lot to cover there is our mission for this deep dive is to basically shortcut you to being completely well informed on the cutting edge of AI music production. We've got a whole stack of sources, Suno blogs, X posts, mastering guides, legal filings, and we are breaking it all down without the information overload. Because, I mean, nobody has time to read all those court dockets. Exactly. Okay, let's unpack this. Okay, so if we wanna answer whether AI fits a studio workflow, we really have to look at where these AI tools are physically sitting. Right, because historic was they've been sitting in a web browser. Exactly. But Suno just launched a beta for something called Bridge. It's a standalone app that runs directly alongside Ableton Life 12. Which is huge, but we should probably lay out the disclaimers first. Yeah, fair. So right now it is Mac only. You have to be on a Premiere plan to use it. It is not on Android. It is not on iPhone and no Windows support yet. And no dates given for when that expansion might happen. Right, no dates. But for the people who can use it, it's a massive shift. To understand why, you have to think about what a DAW, a digital audio workstation, actually is. Yeah, because the old way of doing this was terrible. Like download a track from the browser, drag it into your DAW, and then manually try to line up the tempo. It was just a lot of forensic audio surgery. But the new way, Bridge actually syncs with your Ableton session. It temporarily uploads your session files so the AI understands the musical context. Wait, so it knows what you're already working on? Exactly. Exactly. You can using Ableton's arrangement view and your loop regions. understands the bar, the key, and the groove of your track. So let's say I have a session going, it's, I don't know, 92 BPM in A minor, and I've got piano and drums. Okay, yeah. So you just highlight a loop region, say bars 17 through 24, and you ask for a tailored warm string section. Bridge generates it specifically for that spot at that tempo in that key. It's kind of like a kitchen workflow. Instead of ordering a completely pre-made meal, like generating a full song, it's like having a sous chef who just chops exactly the carrots you need for the soup you're already cooking. That is a perfect analogy. And the community is really responding to this. The official AtSuno announcement on X got a lot of traction. And you see producers like AtMasci AI praising how it keeps the song right inside the session. Yeah. AtGreggCosis was talking about how it completely avoids that whole clunky import workflow. Right. But there is a catch. At Studio LHC pointed out on X that every time you generate one of these parts, it uses your Suno credits. Oh wow. So if you're, iterating a lot that could get expensive? It could. Okay. But practically speaking, this is exactly how groups like the Synaptic Plunge team, who make full original YouTube albums are working. You don't regenerate the whole song, you just fix or add one specific part at a time. That makes total sense. So the big takeaway for Story 1, Suno Bridge brings Suno inside Ableton Live 12 on a Mac for Premiere users, writing new parts that fit your session. But it's a beta desktop Mac only, no Android or iPhone. Exactly. So that's the studio door opening, but once you have the parts, what would a real engineer do next? Well, they'd have to fix the audio because AI generations are frankly kind of a mess acoustically. And this applies to any computer audio editor, by the way. We aren't talking about phone apps here. Right. And the instinct for a lot of people is to just throw a limiter on it, right, to make it louder. Which is the worst thing you can do. For anyone who doesn't know, a limiter is a tool that aggressively turns up the volume while preventing the audio from clipping. And we measure that loudness in LUFS. But turning a track up just exposes all the AI flaws. It amplifies them. There was this great at Mastering One on X. They analyzed 20 raw sueno files against commercial masters. Oh, I saw that. The numbers were crazy. The AI highs were what, 4.4 dB hotter? Yeah, and the air frequencies were 5.1 dB hotter, while the lows were 1.5 dB thinner. Now, that's not a strict lab rule for every track, but it's a very useful clue about the sonic footprint. It's like making a track louder without EQing it first is like putting a magnifying glass over a blurry photo. The flaws just get bigger. Yes. Which is why the Lupelin Mastering Guide dictates a very specific five-step workflow. Step one, you have to find the fizz. The fizz. Yeah, that harsh digital artifacting. You use an EQ, an equalizer, and search around 2,000 to 4,000 Hertz. When you find it, you cut it by about 1.5 to 3 decibels. Okay, so you tame the harsh highs with step two. Step two is cutting the low end mud. The AI tends to smear bass frequencies around 200 to 300 Hertz. So you cut that and you put a high pass filter around 25 to 30 Hertz. Because if you don't, that sub-bass rumble just eats up all the headroom for your limiter, right? The kick drum gets swallowed. Exactly. Now, step three is crucial. You have to compare your changes at equal loudness, what we call A-B testing. Here's where it gets really interesting because loud always sounds better to the human ear. It's an acoustic illusion. The ear is completely fooled by louder volume. If you don't A-B test at equal volume, And then step five is just matching every track's loudness before stitching an album together. Right. And the Synaptic Plinge team uses this exact step to make sure quiet tracks don't just disappear when they upload those long, full-album YouTube videos. Got it. So the takeaway here. EQ first, level second. Cut the fizz and the low-end mud before you turn a sueno song up. Judge masters at equal loudness and match every track before you stitch an album. That is the way to do it, but a professional workflow isn't just about mixing. It's so about how you communicate your vision to the AI in the first place. Pros describe sounds, not stars. Ah, right. Prompting. Which brings us to the big lawsuit. Yeah, the August 31st class action filed in Boston by several musicians, including Jason Isbell. Now this isn't just about copyright, is it? No, it's fundamentally about the right of publicity. That's the legal right to control the commercial use of your name and identity. They're arguing that using artists' names as prompts violates that right. Soon-O pushed back on that pretty hard on November 3rd, right? They filed a motion to dismiss. They did. And their motion basically hinges on three arguments. First, they say artist names are just being used as shorthand for a genre. Second, they claim that shorthand is protected by the First Amendment. you, that musical style itself isn't copyrighted. The really fascinating part is how Soon-O claims they actually handle those prompts internally. They say they convert artist names into generic style tags. Right, so if you type in a band like Pantera, the system just swaps it out for aggressive metal. Okay, so wait. If Pseudo converts names to tags anyway, does relying on a famous name actually give you a worse, more generic result than just being specific with your ingredients? What's fascinating here is that yes, it probably does. Relying on a name means you are surrendering control to whatever generic tag the system assigns. So what should producers do instead? They use a five ingredient sound word prompt formula. Okay, break that down for me. It's genre and era, tempo and feel, you know, the BPM, named instruments, a specific vocal description, like saying baritone instead of just male. And finally the mix or mood. So instead of a generic prompt naming a country star, you would type something like, slow Americana ballad, 72 BPM, fingered picked acoustic guitar, pedal steel. from weathered male baritone. Dry, intimate mix, hopeful but tired. Exactly. You are describing actual acoustic properties. Now I want to be clear, we are staying completely neutral on the legal battle itself. Of course. Yeah, the courts are still figuring it out. No responses are due to the separate first. Yeah, and if you look at X, users like at club diaspora, at tech signal, and at the stack music are all debating this. It's a huge test of the Elvis act and it's very much an unresolved debate. But practically speaking, the user today, the sound word formula is just better. It's how the synaptic plunge team ensures their album tracks actually sound cohesive. It gives you control. So story three takeaway. Suno says artist names and prompts get turned into generic style tags anyway, and the courts are still arguing about them. So describe the sound, genre, tempo, instruments, voice, and mix. Nailed it. And here's the fine print the whole industry is reading. Because That is Bell lawsuit connects to an even larger legal cloud over this whole space. You're talking about the reservoir media lawsuit. Yes. Reservoir is a massive publisher, meaning they control the underlying compositions and lyrics. And they just filed an amended complaint against both su no and anthrop. The numbers in this filing are just staggering. The su no suit points to 1368 allegedly infringed works. And the anthrop suit covers 1143. that combined maximum claims are topping $445 million. Yeah, and that number comes from statutory damages. Under the law, they can claim up to $150,000 per infringed work, plus additional violations under the DMCA, the Digital Millennium Copyright Act. Okay, but I have to ask, so what does this all mean for the listener? Because those numbers are terrifying. Are these 4 or just mathematical legal legal ceilings in a complaint? If we connect this to the bigger picture, these are absolutely ceilings, not verdicts. At the Stack Music and at 1 a.m the average user opening the app today, nothing changes. The app still works exactly the same. Right. But it does mean you need to be smart. If you are a creator, do what the Synaptic Plunge team does. Keep records of your prompts, write your own original lyrics, log your dates, and deliberately avoid copying specific songs. That makes sense. So, our final takeaway? Reservoir's expanded suits push its maximum claims against suenoananthropic past $445 million. But these are per song ceilings and complaints, not verdicts. And nothing changes in the app today. That's the reality of it. So to quickly recap today's deep dive, we learned that Suno Bridge is now living inside Ableton Live for Mac users. We learned you absolutely must EQ the fizz and being out of AI tracks before turning them up. Right. We learned that pros describe sound with a five-ingredient formula instead of naming artists, especially with the lawsuit's pending. And we saw that while publishers are claiming hundreds of millions in statutory damages, it doesn't break the tools we use today. Exactly. Which brings us back to our title question. Can AI music fit a real studio workflow? And the answer is yes, more and more every day. The tools are literally meeting producers inside their own software now. And it turns out the exact same traditional studio habits, making small EQ fixes, doing fair AB comparisons using precise sound word prompts and keeping good records are what make AI songs hold up even while the courts decide the bigger rules. It's all about how you use it. It is. And it leaves you with this kind of provocative thought to mull over. If these AI generation tools eventually become completely invisible, just standard plug-ins sitting in the background of every single DAW, will we eventually stop calling it AI music and just go back to calling it music. That is a wild thing to think about. That's today's Sculpting Sound. Thanks for listening to Sculpting Sound from the Synaptic Plunge team. If you've got a few more minutes today, check out our other shows, The Daily Plunge and Silicon Side Gigs. See you in the next one.