I dont think you actually read my comment before you "removed" it. It was supporting your video and condemning crybabies. Devil is in the details lmao That is of course you actually removed it. Doesnt bother me, peace
I have just used Synthia with this prompt : "Hans Zimmermann's Dune style emotional music , 30 seconds long, strong female voice, strong drums and pipes". Synthia made a miracle for me :) Thanks a lot!! :)
I just started to use Synthia too, it's very promising still need improvement but I really love the vision and I will definitely look after it. thank for the video EDM tips
I've been going through a Yacht Rock phase recently, so I started using Udio to create songs for a fictional band called "Keep on Truckin'" I take the generated track into Logic Pro, split the stems and then rebuild the track using Logic's tools such as the Session Drummer, Bassist and then use Logic's effects and Mastering algo to strip out the low bitrate mp3 sound and make it sound professional. I also use it to create ambient and rave techno tracks.
it totally does sound bad, and if you notice the stems there was BASS, DRUMS and OTHER. Well, the "other" is just fluff that the Ai uses to 'fill out' the spectrum. Mostly garbage, but it's going to improve in the coming months. Suno and Udio actually do better.
Suno is amazing but it often has that shimmering static residual sound leftover from the diffusion process bleeding through various parts, especially the voices and sweeps
Another, maybe less direct AI tool is just to use an LLM like chat GPT, an dprompt it to recommend keys for specific genres, then chord progressions. So I could say what are some good keys for a dark UK Garage track. PIck one and then ask for UK Garage chord progressions in the selected key. Take that over to your DAW and pug it in. Put in a quick bassline and chords in that progression and see if you like it.
Suno added vocal stems recently, but not the other bits yet. I'm not worried about a digital entity dominating a physical space. People can feel however they want about it, it's an open source tech backed by multiple nations and companies at this point...it's not goin anywhere, regardless of your feelings and people can make like the Amish at any point that they want and choose when and where they step off the tech/culture train. There's people that refuse to use CDs/DVDs, cellphones, social media, etc., if ya don't like it, don't use it, simple as, but don't tell other people what to do. Makin the quantity over quality argument is like arguing between Junk Food and High Class Cuisine...which of those do you eat more of on a day to day basis? If anything people's interests are gettin so diverse and specific to the point that we have sub-sub-sub genres of shit now, there's plenty of room to make content and art for, and people will always be into another person's ability to create and express themselves and connecting with the human experience through their vision/art. Arguing that it's theft when you take from thieves is funny too. I'm a freelance artist, it's how I make a living, and I and many others learned by tracing and copying shit we thought was cool until we could form our own styles. We all stand on the shoulders of giants, actin like you're gonna be the last giant and no one is gonna stand on your shoulders is dumb. Plus, no offence, but who the fuck that is 45 years or younger didn't grow up recording songs off the radio, usin double tape decks, recordin tv shows, recordin vhs, pirating games/music, etc? You're fuckin thieves man, stop whinin about it. Are we freedom of info types or not? :D
I intentionally don't use AI. The music it produces is so ordinary and mainstream that it tends to hinder one's creativity. I want to write the melodies myself and determine how the instruments harmonize.
lately been listening to a lot of the 'turn of millenium' NYC/global prog vocal & trancey house 'classics' that got me into this... danny tenaglia, junior vasquez, johnny vicious, hex hextor, plasmic honey, brainbug, oakenfold, sasha n digweed... it all had *soul* and *edge* that like heavy metal or hip hop 'classics' could never be have been created by machine intelligence. in my humble opinion.
Agreed. This whole thing sort of feels like Paul Bunyan vs. the steam machine. Accept we won't be losing by a hair, more like a million miles. I have used some A.I. to make visualizers to go with my songs. Once I upload the rest I paid for, I do not plan to use it again. We all have a choice. I would rather give real humans real work. Doesn't mean I can always afford it. I don't make anything for music production yet. I just don't think that automation in work is the same as art. I feel it's important to use our hearts/emotions to make art and not roll a dice.
@thomashambrecht6535, Don't use AI to make music for you, use it as a tool to overcome writer's block. Trust me, there's value in these tools when used wisely. It’s not about relying on AI completely, but about using it to push your creativity forward when you're stuck. Don’t dismiss the idea, my friend. Human creativity will always find a way, and having something that helps you break through creative blocks can be a great asset.
Good stuff as always...thanks! Loving DJ Studio for a number of reasons. As someone who has lived through this before, your comments on the future of AI music are spot on. The luddite comments echo those of the peanut gallery when the 1st mainstream computer music was being created. I was in the thick of things back then, I'm still in it now, and it just keeps getting better. Appreciate all your contributions.
@EDM Tips, thanks for this wonderful tutorial! I noticed something, though, that I wouldn't recommend people do: sampling audio from a RU-vid video into Ableton. The reason is that RU-vid encodes audio at a maximum of 192 kbps in AAC format, which isn't great for high-quality sampling. When you play that on a high-end audio system, all the imperfections from RU-vid's compression will be noticeable, and the sound quality will suffer. You're a pro, and I’m sure you’re aware of this, but it’s good to make others aware so they don’t start sampling from RU-vid and then wonder why their track sounds poor on loudspeakers. Always use lossless formats like FLAC or WAV for the best results.
Technically, yes, but realistically - in my experience, at least - it depends on how much of the sample you use, how much you manipulate it and whether it’s recognisable. However, I’m not a music lawyer, so don’t take my word for it!
@@EDMTips I thought you said you didn't encourage this WIll? @Producer-student in the UK EU and US ANY amount of sampling is copyright infringement the length is irrelevant.
Merci Will, it is a very nice overview with concrete examples. Personally I still prefer using my old school tools such as Live and Cubase.But never say never !
Really interesting dude, a real eye opener! Testing some of these for sure... 🙂 As for the question over whether AI will in some way make us redundant- I work in IT where it's obviously having a direct impact on my role already. I share your viewpoint- I think we'll definitely need to adapt to keep on top of AI in music and in all other areas of life, but I think if we do adapt, we'll use the tools to enhance our inspiration, but still keep that human touch. I know I will. If it get's to the stage where it's so advanced, fully mastered and release ready tracks can be 'generated' with AI, then that will be strange, and open up the debate further; but until then- let's get inspired!
Great tips! How the copyright works if we get into a really nice idea, we polish it, and when published these platforms find out that the original melody/lyric were created using their tools?
Those online stem separation tools seem impressive - as long as you haven't seen Steinberg Spectralayers. It works locally (same like RipX, which is worth mentioning as well), but goes far beyond any other unmixing tool I've seen to date in both, quality and capabilities, like separation of background vocals, separation of individual voices, drums separation, crowd noise separation, de-bleeding (reducing artifacts), de-reverb, etc. - it's widely unknown for those capabilities (bad marketing?). Check it out! I'd be interested whether anybody knows a tool which is even better?
Someone needs to come up with a tool that you give it a sound you'd like to use for I.e a lead, then you give it the style and tempo and it goes through your samples and suggests other sounds that would best fit for other elements. Sound selection has to be one of the hardest things to get right, I mean it's the difference between a great sounding and well known track and one of these midi download versions of the same track. It certainly wouldn't put sample pack creators out of business; if anything, it would encourage them to make the best quality packs possible.
izotope plugins are generally highly regarded in terms of quality, but I'm afraid to say that the Ozone 11 AI "Master Assistant" is a joke. It always pulls in all of the modules, some of which are highly invasive and make things rather worse than better. It doesn't make any choice about what's needed and what not - just try with an audio file from which you know that it's perfectly mastered. The procedure is this: it listens to a few seconds of audio from the "loudest part" and based on that, it sets up the individual modules. For example, it adds a dynamic EQ with parameters based on those analyzed 3 seconds, which obviously doesn't make sense. Similar applies to the other things is does. It's maybe nice for playing around, but what it does is kind of the opposite to "professional".
That is a very interesting and well produced video, some great advice many thanks. I already use LANDR and I have been looking at Phil Speiser and purchased a couple of the products.
I guess since I learned most of these producers are loopers and drag and drop producers I guess this makes sense , one thing I be the first to hold true is when things break down , what tech are you going to cap ai , ok what ever it takes I guess
@EDM Tips Good Video! Keep sharing this stuff. Been seen a lot of comments talking about stopping this, stopping that. Of course this is a controversial thing yet these AI Tools are for improve the music. Sure, there’ll be guys complaining as this goes on but, sooner or later they’ll fall into this and it’ll be good cause it’s a big aid to us producers and singers. Obviously, the industry will put an end when it be distributed but for fun purposes, let the peeps do whatever they do; violate copyrights, plagiarise, covers without mechanical license. They do not understand the business. The same happened years since with Autotune and such. Turns out everybody can now “Sing”. In the end, life without music would be horrible. This’s the best art of all and those who have it naturally, the AI’ll make it better and it MUST BE used to collect royalties. The AI doesn’t do everything by itself, someone has to set it up.
I agree. I guess the main concern is that it will get to the point where it doesn't even need detailed prompts (basing it on your likes, dislikes, gender, age, country, etc.) all fed by social media, so it know what you want to make better than you! Now THAT will be freaky
@@EDMTips May be, true. Let’s not forget that the government, laws etc. could regulate AI to ensure it doesn't get out of control; that its use remains ethical and supportive, otherwise, anyone could take advantage of it. That’d be freaky. Fake artist and producers, copyright infringements, plagiarisms, people getting sued and so forth. Also, distributors will have a clear understanding of this “Futuristic Tool” and have already begun regulating distribution when content is primarily created by AI. Not to worry about. Only true professionals know how to use it ethically. So, what you shared in this video is great, assistance tools as long as you know ‘bout the business unless as I myself went: “For fun purposes”.
This is the end of creativity. After pretty much every producer nowadays he’s using just software instruments, plugins, advanced DAW techniques, pre set loops, midi packs, and now AI. I’m an old nostalgic guy? Yes, I’m working mainly with hardware synths, hardware sequencer, drums module or drum machine.
Thank you, it was interesting. I use Captian Plugins occasionally. Ozone isn't always good, so I tend to do more mastering job on my own. Suno really does things. On my channel, the Suno tracks is much more popular than the tracks created by me. Sad but true.
I have a bunch of songs that I wrote while jamming. They are demo quality but great finished ideas. I have Cubase 13 but will never find time to re-record them. IS there an AI tool I can plug a demo into and have it reproduce a high quality version. And maybe even change vocals, bass sounds, lyrics, and make other musical tweaks?
AI is brilliant for established paradigms, and if you give it enough specificity by virtue of a prompt, ‘eventually’ it will be able to deliver cutting edge results. What I feel saves the top 1% of producers, is they’ll be able to conceive of cutting edge ideas and deliver them exactly as they would wish, whereas ai does an prompt based approximation (or at least will do on the cutting edge sometime in the future).
Love your tips and channel - among the best. For this, though, I just have no need or desire for anything AI. With all the tools and resources we already have, there's no shortage of non-AI ways to get the inspiration or music going. And for mixing and mastering, I want to learn and perform these skills myself, so AI assistants just aren't needed. It's just that AI really does nothing for it. Love Ozone and tried their AI assistant just to see what it did. Pretty bad results, and I'd rather use their real tools applying human skill.
i liked your video with Dj studio.. i was very close to buy it , but then i saw that Traktor Pro 4 was out, and it could extract stems to.. maybe you can do a "Vs" review on these two when you got time for it.. ? best regards Ananda.
Are there any AI tools that can edit specific parts of music? Such as taking two pieces of music that are not very different and joining them together. I want AI to make music transitions smoother.
They all use spectra layers for splitting the stems so it would be more economical to get the software than use online services. Get your mastering in Lando and a well-known software, and you guest it.... they just provide a service of running it through it.
A.I will never replace the whole concept of music or anything. It is a great source to help you get things done faster, more efficient, and gives you different ideas that you might have never come up with otherwise. It's a "helper" just like a teachers aid. If you FEAR that your teachers aid will take your job then that just means you weren't that good to begin with unfortunately
I think its great for coming up with more musical tracks. I'm a song writer but I'm not that great at producing beats so I'll often write my music and have ai come up with something that goes with it. I'm a bit scared to release anything in the public though because of this AI backlash.
So, on SUNO, the vocal that was applied, where does that come from? Could it be a legal minefield? Could Taylor swift say it sounds to much like her (ai vocals are already very ubiquitous)? I don't use ai, but, i do think it will replace us sooner rather than later! Think about it, Dan Ek is already saying Musicians are not paying anything now to make music or very little!
Really interesting Will. For better or worse it's here and musicians are just going to have to deal with it. Agree with your comments. Guy Michelmore pretty much said the same thing in his latest vlog.
During you last couple of videos I saw you switched back from the Hd490 to your old 650. Why? I‘m close to buy the HD490 and don’t wanna make a mistake.
Why are people acting as if A.I is this super bad thing that should never be allowed in music? Everything evolves and the best thing you can do, is evolve with it. Creativity isn't about what you use, it's about doing the most with what you've got access to. People acted the same way about DAWs and now look at us. If your favourite producer came out and said the snare in his latest song was generated by A.I, would you stop listening to the artist? It is simply just another tool guys. Just like your parametric EQ that didnt exist not so long ago, or your VST that emulates a piece of analog gear. This argument is exactly the same as "People who use samples" vs. "People who make everything from scratch". Do what you feel is right and just because another person chooses to drive instead of walk, don't assume that person is lazy or any less capable.
At 14.00 your ai generated Madonna Like a Prayer chord progression, and with your sound selection it pretty much sounds like it. Question. what happens if AI generates -(randomly) something that exists exactely, and you use it?
Another great video Will! 😀 What are your views on Neutron 4 for semi AI mixing? Also, I’m gutted I bought RipX before seeing what you did with DJ Studio!! 🤦🏼♂️
I haven't tried Neutron 4 yet but I will give it a go as soon as I get a chance :) DJ Srudio was a fun experience for me and I am looking forward to finishing the track!
I understand you're just playing the youtube game and keeping the editing snappy but do you not feel responsible to give a legal disclaimer on sampling? or are you fine encouraging the sample first get sued later approach?
I play the “think of things my audience might find interesting or useful” game - not the RU-vid game. I like to show people how I like to make music, and neither advise on the legalities or encourage it. So, to answer your questions: no, and no.
@@EDMTips don't be ridiculous Will as you are a tutor and influencer, to an aspiring audience of mostly beginners who are copying you, and following your advice via these tutorials, telling those people to "Go try this" is absolutely the definition of encouraging.
I use Splice for their sound library. And Suno for making song. I let Suno make the song and then I use stems in Serato Studio to isolate the vocals. Then I make my music in Ableton. 😊
Are there any AI tools that can edit specific parts of music? Such as taking two pieces of music that are not very different and joining them together. I want AI to make music transitions smoother.
Ooh, Suno won't accept artists' names eh? Could it be because they trained their AI on copyrighted tracks and they're worried it shows in their output?
yeah.. lets disconnect beginners brains fully and get their money... the worst music we had have is comming out of mainstream speakers😊ASK WOMAN HOW THEY LIKE IT... when woman are not crazy for a song anymore.. they will not go to dance anymore.. You guys have effectively killed the music... the vibe.. and tha passion😊😊
Couldn't get DJ Studio to export correctly. I asked them for help and despite them saying they'd get back to me in an average of 30 minutes later so one replied for over 24 hours. At that point I was frustrated, so just asked for a refund .... And the guy was rude. So get DJ Studio at your own risk!
I lost my interest in music because of AI. My music on spotify is now on playlists between AI crap that their creators didn't even pay attention to what they upload. So, in some way my melodies which I spent months to create are now considered as AI rubbish. Why would I be interested to continue making making music the way I did a few months ago?
@@EDMTips I wonder what will be left for humans in the near future. Maybe soon comes a new platform (or maybe Spotify itself) which writes it`s own prompts and distributes automatically on all platforms. We are close to a point that AI will not need us anymore.
Hi you said .that ai is not still there. I think it is . I have used ai to make stunning music with my special promt.Not because im good but because ai is
I tried Synthia and I can't understand what's going on. Whatever words I fill in, it gives me the following error, "Request failed with status code 500"
I had the same experience. I tried multiple accounts and ended up having to delete cookies because there’s no option to sign out. While I do like the simplicity of the website, there isn’t much to explore. To top it off, the verification email says, “If you are having any issues with your account, please don't hesitate to contact us by replying to this mail,” but it comes from a no-reply address.
@EDMTips hey, love the vid. Duly liked and subscribed. I’m looking for a tool that will help me remix a really great classic track. I want to input the original, have it analyze and then between me and the agent, we craft the remix. Anything out there that fits the bill? Thanks, gonna trawl through your vids now 👍🙏🫶
it won't be AI that factors artists out of the creation equation, it'll be greedy execs who decide they can pad their profit margins by having their secretaries work unpaid overtime feeding prompts to music generators.
I've been using RipX Daw to get stems that I couldn't find anywhere else. I've recently done remixes of Cocteau Twins (Blood Bitch) and Led Zeppelin (Immigrant Song) lol
You can make money with these, but realize if you go head-to-head with a record label or publishing company, you'll lose far more than you ever made. They also work with the IRS CID who can legally freeze your bank account and file tax evasion charges. This will crash your credit-rating as well. Just say'n.
I always thought AI was going to do things we don’t like doing so we’d have more time to be creative in making music, art, films, writing etc. Instead it seems to have the op and made people lazy, anyone can do it now with no practice or talent, probably why everything sounds and looks so homogenous and predictable now 😢
To me, this is just gross. If a melody is generic enough to be generated by one of these programs, it’s also generic enough for you to have come up with yourself in five minutes. The sample it tool is useful, but it’s not AI. I’d rather never come up with a song good enough to release than have a machine generate the idea and pretend it’s my own.