Video

Reaper is the best and most lightweight DAW for someone with experience of DAWs. It's barebones UI makes it quite hard for a beginner to get their head around it. I suggest Cakewalk for a free DAW on Windows that can basically do almost everything a fully paid DAW can... And after getting comfortable, switch to Reaper. I primarily use Studio One for all the macros, folder busses and basically how intuitive and how quick and easy it makes my workflow.

Started with PlayerPro in the 90s, went on with Logic and later switched to Cubase. Due to other circumstances I wasn't able to record for the past 15 years. Last month I downloaded Reaper and bought the license a week later. To me it provides the best experience - I wasn't able to create music at the same level of comfort with any other DAW I used. Reaper allows me to streamline my workflow in a way I've never thought to be possible. The most convincing fact for me was that Reaper doesn't come with tons of instruments and other stuff, but provides probably the most powerful foundation for you to make it whatever you want it to be: It lets me make my own choices of how to modify it with the exact upgrades I needed without having gigabytes of modules preinstalled that I wouldn't use.

So you're really using the "machine" to create the idea, melody, and eventually any lyrics for the song. Right? In other words, you did not write the melody. Correct?

Do we have a way of bringing in a chord detector in Reaper?

As a non-musician and just a common simple user, AIVA is complicated. I think what most people like me would appreciate is a music creator that's as simple as an AI image creator.

I am currently a S1 user and a small time music producer. The biggest issue I have is that the playback quality of the output through S1 never comes in the output that is rensered with the highest possible quality. I tried to contact Presonus support team and they seem to be unbothered. I think I will soon buy the Reaper license.

I love Studio One, but it’s handling of a very powerful Ryzen 9 7950X CPU, is deplorable, like most DAWS out there. Reaper is the oNLy one that allows my CPU to work with its full compliment of juice. That is the sole reason that I stayed with Reaper after trying Studio One. S1’s workflow is incomparable. I’ve used Reaper, S1, Logic, Ableton, and others. I don’t need all of the plugins and instruments that come with certain DAWS because I simply don’t use them. I’ve got a really fun and fairly vast collection of third party VSTs and VSTi plugins, so I always go there when I’m writing and producing.

I feel there is a difference between hitting generate with no human input and spending time writing a song, getting the structure down, and then generating the end product with AI. No difference between using AI to generate the song and hiring someone on Fiver to play the song I personally wrote. I think the argument is that people with no talent can just pump out 100 bangers a month and put artists out of business. I view it as another tool that can be used by artists to complete work. I didn't see drummers out of a job because I can make drum tracks. It is lame that people can pump out songs but in the end it will just be clutter and garbage. It will take actual human talent to input the structure to make a good song. Plus AI cant play live shows. It will just be another tool used by all musicians, not an end to music. Just my opinion, I am not the authority on how you feel about it. Stay blessed everyone

I go back and forth between Reaper and Studio One as my linear workhorse. I love how crazy customizable every little thing in reaper is, but I also find myself getting lost in the sauce of it all pretty often. I forget hotkeys I setup to do things and then get bogged down. I like how studio one does all the important stuff very well and is much more straightforward GUI wise, but then sometimes I miss the customization. I fear i may juggle them both for awhile. Bitwig is my other main. Nothing quite touches what bitwig can do from a modulation/sound design workflow, I just dont love it for arranging and it lacks proper tools like video and ara support for media composition

I have nothing to say about this video, because I definitely don't know much about mixing, masking, and what not... But I do have this to say: I was there, I read it all! 😉😁😎👊🏻

And now i see the song which you have been using in your informational videos. This is awesome!

I saw a test on how different DAWs handle fades on audio items. What I remember is that Reaper was one of the best, Pro Tools too, and StudioOne was one of the worst. Who thinks about such things working in the DAW? Close to no one, I suppose. But such differences do exist. Of course, I'm not sure if StudioOne has changed anything in new versions. Also, Reaper has one of the best resampling algorithms, which comes into play when working with samples at different sample rates or rendering to a different sample rate while processing in a project sample rate. There're also tests on the web. Izotope RX is also amongs best alrorithms, if I remember it correctly.

The only robots i actually wanna listen to are Daft Punk man how did we let ourselves get here 😭

Some day I went to listen to Release Radar on my Spotify. I noticed some releases which list artists I follow as collaborators of a noname artist. I instantly became suspicious, played a track and it was apparent that first of all, my followed artists are of course not related to the tracks and they were added by a shady actor to boost their play counts. Then as I was listening, it also was obvious that the audio was AI-generated as it had some distinct sonic characteristics. AI-generated "music" sounds muffled, something like it's processed through a vocoder with white noise a little bit. I don't remember if I sent a report to Spotify regarding using unrelated artists in songs metadata. I remember the album was called Invasion of Galatto (quite telling). I can't find it on Spotify now, so maybe I did send a report and the album got removed.

could you show how you work with pitchbends and other cc lanes in reaper? it's very hard to copy parts of it or i just don't know how. in ableton you just select the part of it and duplicate, but how here?

I make my own music without any ai. I did release one song per week this year and have already sent songs to my distributor to release every Friday till the end of this year. I agree with everything you said but i can't agree to the fact that the only way possible to release so much music is by using ai.

Started with Reaper knew nothing about DAWs not a good move settled on Cubase because it automatically did stuff I couldn't figure out being new to DAWs. After a while I found i didn't like the way Cubase was dealing with electronic music production i was thinking about Ableton but hated the mixer, and went back to Reaper ( which I knew I would do at some point). Yeah it creates like Abelton for electronic music. Very easy sound design stuff. It sounds amazing. I use the old school white tie theme. It looks like how I want my interface to look and the mixer is perfect. Reaper really has no limits. But you have to know what a DAW is what you need and don't need. I never really used the stock programs in Cubase that much. Which was another factor. These DAWs you're paying for the extras, Cubase without the bloat would be cheaper and more useful. To me. Plus Cubase is more for live recording with added samples. Reaper,doesn't matter you can set it up to be what you want. I do house so it is geared for that. But Reaper is my DAW not changing to anything else. No need. I'll keep my Cubase for some instruments but otherwise I don't use it at all, thought I would switch between the 2 but....nah

While AI music is definitely gonna replace a huge amount of music but it has some stuff it just can't replace, 1. Languages, for English it of course has the most data for singing and grammar but for less spoken languages like Korean? It might be able to make kpop but it won't be able to make some Korean shoegaze (eg. Parannoul) and make it speak coherent Korean, and Korea has a decently sized music scene so what for languages like Croatian or Hungarian? 2. It can't exceed the quality of the best music just due to the way it's made, like yeah, the average quality of an AI song might be better than even most but sometimes quantity doesn't beat quality. 3. I'd rather say that the most popular music as of now relies on image of the artist as well (eg. two artists who I listen to barely: Charli xcx brat which sounded more unique than Crash but is performing way better or Taylor Swift, whose appeal is pushed to be the most popular artist currently alive due to her image as a story-teller and her songs being her diary entries (I'm not saying that the quality of the music does not matter in fact I think it's the most important reason but her image is what pushes her to be the most famous artist right now)) Though these are just the thoughts of a miniscule 17-year old artist from Croatia so correct me if you think I am wrong.

The biggest challenge is that AI music generators have only recently begun evolving. This is just the beginning! Their only limitation is computing power. Software development has virtually no limits, so AI music generation will rapidly evolve, and in a few years, generated music will surpass our current levels of imagination. Eventually, there may only be room for live performances. Composing music could become just a pastime with no financial reward. I confess that I've completely fallen for this "composing," or rather "prompting." I can't stop it. The output from the generators is still far from perfect, but it's captivating. It's not difficult to "prompt" an entire album in a day, which can be listened to in the car as background music. You can compose the lyrics yourself or with Claude/ChatGPT. It's starting to be faster to "compose" everything on your own than to laboriously search through the flood of "music" on Spotify.

I am a real musician using AI to create music that I LOVE and that sounds 100% genuine to my ears. A good song is a good song! I continue to be blown away by the excellent melodies and authenticity of the vibe generated by AI: youtube.com/@miragemusic83

As someone who uses AI for music generation, I completely disagree with the video. I feel like you simplify Ai music to just prompting, but in truth it is much more than that. I would compare Ai generative music to a wild horse that an artist/designer is trying to tame to do what they want. First you have to write your lyrics, then you have to arrange them in a way that Ai will get you more or less what your initial creative thought was. Then you keep prompting for hours or even days and every time improving your instructions for the Ai to get what you wanted. Ai music that people are listening to is not made in seconds, and it also requires some skills and lots of creativity. I agree that streaming platform should clearly separate Ai music from 100% man made music, but it should not be not allowed to post or distribute it as it has its own value and use cases

$H1T-MUSIC is compatible with our "metadata guidelines"

i joined an fb group out of curiosity and a lot of these guys think of themselves as legitimate musicians. 🙄

video looks heavily edited. Maybe ai commentary?

People eat sh*t. Most people have terrible taste and simply consume bad stuff. That's how we see stuff like Temu, Ali Express or Amazon... Selling a lot of bad quality products that no one should ever buy. That's how we see stuff like Walmart's "Best Value" being sold and people eating that. Generally people don't care, they just consume bad things. Think of all the horribly bad designed products on t-shirt websites and stickers, etc. It's 99.9% pure AI generated (the bad kind, bad prompts, not-curated from hundreds of results), or unprofessionally done sh*t not at all designed by professional designers or illustrators with decades of experience and degrees... yet people still buy it. That's also how social media content creation and consumption works. The same goes for music. Most people will listen to anything. They don't know any better. They CAN'T RECOGNIZE any better. They are used to bad things and completely disassociated from any sense of quality. Now, if you're reading this comment, that means you are watching this video and you care, so you are the 0.01%. Know that you're not the target audience, nor the most effective persona that will create content for the 99.9% of people that simply don't care. Just create for your lane, the 0.01%. The root of the issue you are describing in this video is simply that the world population is too big, and we have a problem of quantity VS quality. There's not enough quality taste. And because the tasteless population consuming things is huge, that's where the money is at. Any company which the purpose is to make money will produce for the tasteless population and will endorse any methods of quickly producing more for that population. Asking Spotify to remove AI music is the same as asking Amazon to stop selling Chinese crap and start curating companies and products they sell. Big LOL. edit: thanks for linking my comment, OP! People might actually read this so I removed some of the "sh*t"s, as there were to many (this was a hastily comment as an emotional response to the video). Also made some revisions to spelling, grammar, and weirdly enough also math... It would be a bit of hypocrisy to have a bad-quality comment complaining about quality 😅

I make a.i. music where I put my own lyrics in it

You do not know what you do not know. I have used both Studio One and Reaper, but has ended up in Cubase/Nuendo. Many of the fuctions you describe is already part of Cubase. Depending of what you do - reaper lack a lot of musical functions like: No good pitch correction, no easy way to find a tempo of a track, no chord track etc. Seems like you write MIDI a lot, here Cubase is a lot more advanced. And it even more CPU effective for large projects then Reaper.

Suno and Udio are now being sued by Universal, Sony and Warner. Depending on the outcome, it could lead to some important changes... I mean, if the majors win, Udio and Suno are financially dead, and it will be the end of A.I. generated music companies. Now, the use of personal software will be far more complicated to prosecute, as it was with samples back in the days...

I agree 100%. And yes, this is what is going to happen and is already happening. World Is full of AI people, AI art, AI music. World is fake.

Rick Beato said about a year ago that in the near future, 9 out of the top 10 songs in the Spotify Top 10 will be AI created and nobody will even care. Casual music listeners don't care who makes their music, they just want to hear music. Whether or not a human creates it is of no concern to them. As a musician it hurts me to hear that, but I know it's the truth. This is what the future holds, unfortunately.

I guess this won´t be a popular Opinion here but you seem to at least try to think about this topic and don´t condem it all right away. It seems pretty strange to me that you are judging music by how it is made. When you found that "Artist" you talked about - why did you need to find out that it is AI to say it is bad? Bad music can be made with or with out AI - AI just made it easier so more bad music without any artistic vision will be made. But at the same time, good music WITH an artistic Vison can be made with AI too. Its pretty strange to me that ppl accept Hip Hop / Rap Artists that buy Beats from someone else and then "Talk" their Lyrics to that Beat as being Artists, while they won´t even listen to the Lyrics some Artist wrote and made Music for it with AI. To me AI is a bit like what a DJ does - you select what parts will fit together and represent an artistic vison of your own. Oh and btw: If someone gets AI to write Lyrics that are good and not obviously AI at first glance, they really are an Artist for coming up with such a prompt to make that happen. So yes, there are ppl making bad music with AI trying to make a quick buck - but there were also ppl making bad music for a quick buck before AI - blame it on bad music without a vision and not on AI.

where were you when writers started being replaced by AI? what about digital artists started being replaced by AI artists? and now follows the music you should have spoken for all artists, not only musicians now that you are affected In the end art is art and if it pleases someone's eyes, someone's ears, they wont care who made it or what made it

You seem to be conflating craftsman with artist. I have a whole video on why I love Ai art.

It’s hilarious to watch people standing AGAINST PROGRESS genuinely believing that they’re on the right side of any history ever.

​ @synapticschism Ultimately these are all tools to a the same end goal, prompting is just a different form of input notation, currently we have all the software and tools that you can write chord sequences, click squares on many a PC APP to define what notes are played and how long for, I started out writing sheet music on sibelius back in school, then later could feed the midi files into a better rendering software. As a software engineer (who has also worked on NLP/NLG and someone (somewhat) musically trained I know that I don't have the time or resources (or knowledge of all the tools and steps!), at my age to make music (even to the level I could 20 years ago!), and some of this new stuff has been really good for my enjoyment as it reopens those doors to be creative and share with other humans and make them happy. At the end of the day, even something as 'dumbed down' as Suno -> you still need 1000 hours of raw experience or a decent grounding in tokenization and how and why an AI tool creates what it creates (considering most training data will be 4/4 for example) to have half a chance of consistently making anything sound decent. Otherwise it's just Loot Box Music for the masses... a few good ideas will come out of 1000s of bad songs even if you are clueless, this low bar probably upsets some musicians. (But from a listener perspective, this feel the same to me as listening to 1000 teenage bands to find one that actually sounded decent... and hearing what sounds good for the masses, is technically also a skill) At the end of the day AI music as 'powerful as it is', once you listen to DECENT human music there is a chasm of quality difference, and current prompting is in its infancy as a form for describing music, (I suspect as the training data needs many more layers of tags to get to a richer more powerful prompt based input working at the 'next' level) Which is why outside of musing ideas, for music - this pseudo-instruction format is NOT USED traditionally over sheet notation etc. - What we have in the AI space, leads to a lot of limitations, proper roundels / counterpoint - forget it... Something like suno doesn't have a structure close to representing 2 people singing over the top of each other with different melodies / words concurrently... and yet you 'might' achieve it via the lootbox random luck. For me the real issue is any artist (however they create the music) barely gets anything from a streaming service, and the amount of people / skills needed to make a decent song is quite high, even the BEST AI todate barely lowers that requirement if you want to make anything better than a 'recorded in your garage' vibe. My understanding / suspicion (could be totally off-base conspiracy this!) is the RIAA crew who are filing lawsuits, just see money and want personal control the music gen scene, they would rather licence and launch their own tools so they can track the 'influences' of artists / tracks and line they up what they do own with anything you make (with algorithmic proof) - and then if you did Launch your music claim a large cut as licence holder, they'll sell the idea by giving 2% back to the orig. artists, or something tiny.... meanwhile a 'skilled' human who takes influence from multiple bands (as happens today) does not face this issue, although even the best artists, certainly they are getting a very raw deal. Isn't it roughly like $4? for 1000 streams? - this is absolutely shocking. Music is probably too cheap to access, lets face it a multi-user monthly sub for less than a CD? - it's economical nonsense, unless you listen to max an hour or so of music a month. Apologies for rambling. Certainly many topics for debate here.

As a visual artist I agree with everything, however there's one thing prompters can't do. Live shows. Maybe I'm in a bubble, but I've seen a lot of people ditching social media and heavily dropping their internet usage. I myself changed my smartphone to a dumbphone for contact, Walkman for music, PSP for entertainment. Surprisingly, lots of people around me did the same in the past year or so. It's still not very common, but I believe a bigger "back to real life" movement begins. EDIT: also there are two points to remember. 1. People who don't care about if song is AI generated don't buy music. CD era is over, now it's all in streams, and it's better to have 1000 devoted fans that buy every album you put out, than 10 000 listeners who just stream your song from time to time and then forget you exist. 2. No copyright. The tracks you mentioned can be downloaded and uploaded on Spotify as "yours" and nobody can do anything about that. AI and prompters having no ability copyright strike you is something that will be weaponized in the future, and I'm here for that :)

can I use them as one shot in say ableton drum rack or sampler?

Thanks for bringing up the issues :) . I'm a software engineer of many years but shifted to making my music and playing instruments and teaching arrangements to people with disabilities. I am figuring out a new platform and a way, publish 10% to spotify/apple music, the rest from own site I can help with, no analytics or ads or promotions over each other, more localized and very little ability to skip music. Will try it out and test out my theories with artists and hopefully get some support and some financial support as I am running dry on savings.

About a year ago, I saw that someone called himself "Schnittharty", which is very similar to my artist name. He created songs and one of his songs has the exact same name as one of mine. I saw the name of the "label" in his Spotify account details. Unfortunately, I can't remember the name of the homepage, but probably don't exist anymore, don't know. On this site you can create AI songs for different genres and upload them to Spotify. The annoying thing was not only that my name was stolen, but that this garbage also produced even more clicks, although it is clear that the higher numbers could not have been human. I didn't want to be associated with that so I contacted them and threatened to take legal action and they took everything down.

thanks for this video, well said brother🙏

Well said. I have spent a lot of money on equipment and software over the years not to mention the time investment but when I tried Udio earlier this year I knew the writing was on the wall and I haven't spent anything since then as it just seems futile now unless you already have a big fan base.

7:39 stop please complaining about things you can´t change. I am a musician myself. The whining is useless. Make the best of it and make your own money with the new possibilities that AI provides.

One correction; Studio One Hybrid gives you a perpetual license after 12 months. So they at least finally figured out that people don't want to pay so much for their product.

A few years ago I almost bought back into ProTools after a hiatus from making music. It was all I knew. But by chance I was watching a review video for Eleven Racks and the guy just casually mentions Reaper. I'd never heard of it. And now that it's what I use, I can't fathom why I'd never heard of it. I'd heard of everything else. Reaper seems impossible to me. It's so powerful, it's so efficient, and it's so affordable! It's literally free unless you have a conscious! The creators are absolute legends of the industry!

If you write lyrics then AI is a good solution to express yourself. I hate it when it's put down as easy. Writing a Nice lyric is work to...

I don’t think you’ll find a musician on the planet who doesn’t whole heartedly agree with you. It is heartbreaking. 8000+ listens! Like many I only have a handful of monthly listeners on my Spotify channel as it is, now I’m competing with this! I guess most people who make music have to be resigned to just making music for themselves. Soon there will be no point putting anything out there for others at all 😢

I fundamentally disagree with the extent of the problem. As you saw, their music was all over the place, they will not gain a large following that way. You have nothing to be afraid of. Also, if you are worried about uploading too infrequently, you could embrace the change and find a way to use AI to speed up your workflow. I agree, writing a simple prompt to generate anything is not artistry, but I believe real artists can remain artists, even if they use AI in their workflow. I'm not too familiar with AI music generators and how easily you can make it actual art, but in image generation at least, there is definitely a possibility to be an artist who uses those tools. Some people use ready-made loops and synthesizer presets, which imo is just as bad (not bad) as using AI to generate vocals or drum tracks for their music. Creativity is about combining things, it's literally impossible to come up with something completely new.

Finding more and more of this generated "content" right here on youtube. The ai slop-merchants don't care about about your final point, it's all about easy, selfish, short term gains.

Why would Spotify have an issue with AI music? They themselves use AI to create endless playlists of AI-generated music to suit every passive consumer-listener's mood and taste (hence no need to pay real artists any potential royalties).

I have to respectfully disagree. Your point reminds me a bit of the times when synthesizers and loopers were first introduced, and the "traditional" musicians were appalled by the music created with the help of computers, saying this is no longer art, that the creators are engineers rather than musicians, that it's artificial and fake, etc. Same when Fatboy Slim emerged and the wider public realized one can make new tracks entirely by sampling existing records. Did it make musicians die out? No. Firstly, it's easy to be impressed by AI output when you first hear it, because it's such an improvement compared to a few years ago. But if you give it a closer listen... the tracks AI spits out have all kinds of problems you'd never encounter when working with humans, that still make the song useless (random noise, fake-sounding vocals, nonsensical lyrics, AI hallucinations, inconsistent melody, very basic solos... to name just a few!). It takes actual skill to make it spit something presentable. Forgive me if that's not what you intended, but I get the impression you don't recognize that effort. Secondly, what you get is usually the most generic version of the style you asked for. Listen to it and any music created by a decent artist (emphasis on 'decent' here) and you'll realize just how low creativity the AI output is. The users of these platforms constantly demand them to allow for more control, more direct editing, and more granular options to actually inject their own creativity into the output. And finally - the account you showed in this video has a total of... 4 followers. To me, this means their weekly releases are not good enough to attract an audience. They'd probably have to put a lot more work into the tracks to gain popularity, especially with next to zero chance to have a recognizable style. But speaking of work, let's not forget that artists get paid for the effect of their work, not the time spent on it. So while I don't think AI will be able to spit out anything better than humans overall anytime soon, if there are some individual AI creators who already make better music than some individual traditional musicians... I don't have a problem with the latter having to change professions, sorry. My point is, I think the real artists who make actual use of human creativity and original thought, will prevail. Some by just doing what they're doing and being great at it. Others, by learning to incorporate AI as a new production tool. Thanks for the interesting video!