@@blubbblubbblubbish This is why the major labels own Spotify and spend a fortune pushing the songs on radio. It's the same with books. Most blockbuster movies will have the same amount of money that the movie took to make, just to advertise it. “Hit songs” aren't a style, it's just that labels demand certain “safe” products for financing and pushing.
I've been a musician for 57 years and what was true back when I started is still true; you can have the "best" track out there, but if no one listens, it goes nowhere.
I am in a number of Facebook groups , I have not heard anyone write a song that even come close delusion is a big factor, knowing what is bad and good is an important starting point
You have to write music that non-musicians can understand. They will eat up the simple formulaic stuff. Forget the ribeye steak, the masses want McDonalds.
yep, correct, my mate wrote a number 1 hit in 1970, he still thinks he has got it, but has no idea, because he does not know how to write to the market
Only slightly is it understanding the market. It's more about creating the market. Most of these hits are probably signed to the same company. It's not ”Oh, you like this?”... it's ”You’re going to like this!"
Thanks for the video and all the work to put it together! There is a loose rule in advertising you need to get the product, thing etc. into someones ears or eyes 6 times before they decide to engage further like buy something. Same goes with music, which traditional radio was good at with a song playlist on rotation throughout the day. So marketing and curators make hits by feeding you the same thing over and over. All those studies are really mapping trends of what hit songs are made of, it's the diet you're fed but also there is musicality too. Case in point a hit from the 70's may not be a hit today given equal exposure and marketing etc.
Great video! Love taking the science-first approach strictly and rigidly. Now for anyone watching, the trick is to do it again, and again, and again. Creativity is like a muscle, and it helps to stretch it everyday. This can be as simple as a line, one melody, or even random ideas, but "show up to work" everyday and it will help. Using these ideas and guidelines are key, and then one day they will become ingrained naturally so that you don't have to think about it. This is when your natural style can begin to blend with the "formula" and create real gems. When the style and the science naturally happen, you can make hit songs that are hits because they are authentic, honest pieces of art that represent who you are as an artist. Wow, didn't mean to go on this long. Keep rocking!
Exactly... show up every day, wherever you do your work, show up, do the work, even if it's shit, even if it's only 20 mins a day minimum, go through the motions, develop your process, then trust the process. Once you've started to set up the right conditions, and have the skill set to actually handle and record/write/capture/etc. the magic when it comes, efficiently enough, before it vanishes back into the ether.... the inspiration will start to come for you more and more, and you'll be able to capture it more and more, until one day you'll be able to walk into the studio any time and just go off any random idea, melody, progression, lyric, etc. and put it through the process until it blossoms into something truly special, that is uniquely you, that nobody else in the world ever couldv'e made. It's a long journey, but it does get alot better, funner, more fulfilling, and the music eventually does start to become very special. Also the GENERATING of the ideas, other than the inspiration we can't really explain where it comes from.... use brainstorming, song maps, writing things out, teamwork, jam sessions, songwriting sessions, whatever you gotta do to initiate things so the inspiration and magic can come. Learning alone will NEVER cause this to happen, only DOING, like it's your job, purpose, and destiny... for years on end. 💯
The learning and practice is important too, don't get me wrong, but it only FACILITATES the special magical thing so it can come through you and into your creation. It's all very important, inspiration won't even come if you can't handle/interpret/capture it yet, and if it does it'll disappear just as quick as it came. Plus if ur also the artist recording it you gotta have some legit skills in order to perform it well enough to do it justice. So I'm not tryna downplay the learning part, and DEF not downplaying the practice/rehearsal/etc. part either... but I would lean heavily towards hands-on learning and practice, rather than passive learning and watching others. Just to clarify 😅
As always the best music comes from the heart and is made for the artist, and you always gotta keep making songs true to yourself to get good and perfect your unique style. Treat yourself right, and of course take inspiration. I've found I made my best music when listening to a song once then loosely basing it on my memory of that song, which ensures that my style is retained while also helping me to pick up new skills and really make the music I want to make (I love skrillex :D)
@@elijah2460 There are a lot of judgement calls in designing a good predictive statistical model. You have to decide which parameters to include and how to weight them. If you understand the environment you're in and know what you're doing well enough, you should be able to make it work pretty closely. It's possible, IMHO.
i think the max martin rules are way more helpful in predicting or crafting a hit. many of them are directly tied to limitations of the auditory system or evolutionary arguments. an example of the first: never introduce two new ideas at the same time. humans typically can only focus on one thing at a time, e.g. it is very difficult to talk to two people simultaneously about different subjects. an example for the second is hierarchy in music: not all elements in a piece of music are equally important. once a voice appears in a piece, most people will focus on the voice. we are wired evolutionary to listen to other people, that is why it is so hard to not listen in to the discussion on the neighboring table in the restaurant.
I havent looked at those papers in depth but I'd be skeptical that their algorithms would be that accurate to a brand new song. My thoughts are that their models would likely be making their predictions largely from the 'popularity' feature for each case, which would make the model very accurate to historic data and less-so for unreleased songs. Also depends what they are sampling. If it's taking a random sample of say 1,000,000 songs from Spotify, there wont be many (if any) Top 100 songs in there, and the most distinguishable feature between a hit and a regular song from an independent artist will be the popularity score. As you said there are so many external factors that turn songs into hits, and a lot of the time those are what actually make hits rather than the raw quality of the song
Popularity has got to be a huge factor in ranking for these algorithms. Since it's based off of the amount of plays and how recent they are, it's basically predicting hit songs off the songs that get the most plays. Well duh. That's what it seems like to me after reading the study a bit further in depth. But still, 88% still seems high to me even with that.
You overlooked a couple of things: You need a couple of short, memorable musical hooks that get repeated a lot. Start big, not low-key. Hits often start with the chorus or main refrain.
there is also the whole idea of marketing where even if its not the best song a fun dance or it being under a lable (who would work to have it been seen) could heavily sway the popularity of it
You should have studied the producer that has made the most hot 100 hits in the last 20 years. Max Martin. There are definitive patterns that can hone in on all the things you mentioned. There are a couple of BIG things that Max does that were nowhere in your research.
Yeah.... work for a dozen giant record labels who put millions behind your songs/artists every time, that's the BIG one 😂 Nah, I'm only half serious, I'm sure he's extremely talented and I'll bet ur correct that there are some valuable insights to be gained, not that I think AIMING FOR A "HIT" is the way to go, but for general use and to become a better producer myself.... I'll research him, thanks 💯
@@CyrusAugustus LOL, True, but wern' they all funded by Big Labels? Anyway the economist looked at Max Martin and some of the patterns here: ua-cam.com/video/F9pIZVFGa7M/v-deo.html
Excellent song, especially from the experimentation position. Really good. I learned a long time ago that the best and most powerful songs have simple melodies (singability, which cements it to memory), and are written in 3/4 or 6/8 (the whimsical meter I call it). I'm compsing orchestral/symphonic concert band music, after 25 yrs of homeschooling our 8 kids.😂
First of all I really learned something from this video so I subscribed. Me being an audio engineer/music producer for a living thats kind of rare these days. Second the song was catchy, your also a good singer and clearly know how to produce... However why did you decide to add that annoying "what" vocal sample that hits on every beat? This is just my opinion but that ruined the whole song for me 😅 its way too loud for a vocal chop, sounds robotic and repeats way too often. It does suite the song, but I would turn it down by at least 35-50% volume and maybe EQ the mid/high freq even more. But overall great video and song! 🙏🏼
Thank you for the conclusion! I was afraid you would say: aaaaand kids, this is how you suppose to write hits. Imho It is definitely the true challenge to aim for the 12/100. And I believe these are actually the songs that will last. Like Stings or Stevie Wonders hits. The 88/100 seem to obsolete very quickly. Because these have barely something to offer except the hit parameters.
Statistics: If Hip Hop is popular in the USA, and you want to break through, you either have to be _the best_ at Hip Hop, which is very hard, cuz that's where most of the contestants are, or you have to find a niche that is easier to dominate.
1st time viewer = New Sub!!! Great video... I have never heard the songs you referenced, well, at least that I know of, but I got a lot from the rest of the video. Thank you
This is what happens when you try to measure mostly subjective traits in music as if their objective stats. It doesn't work. For instance, the study tries to measure a songs emotion by the lyrical content, the use of major or minor chords, and the instruments used. None of those things actually matter that much in describing how happy or sad a song is. If there's any objective marker for a song's emotion it would be the rhythm and how "upbeat" it is. Even that isn't universal. There are tons of happy songs with depressing lyrics like Maroon 5's "Misery", and sad songs with happy lyrics like "Time in a Bottle" by Jim Croce (the lyrics' literal meanings are happy but when you know the origin of the song they're quite sad; not that the study would even know the difference) The mismatch of the tone and the words actually adds its own interest. Also there are tons of sad songs in major like Good Riddance by Green Day, and tons of happy songs in minor like the majority of Rock and Roll. You can actually make the argument that it happens more often THAT way around depending on what genres you listen to. Also no matter what instrument you play whether it be guitar, piano, ukulele, violin, or vocals/choir you can make a song of ANY emotion. This should be so obvious I don't even need to give examples. The only objective measurement in this study seems to be "popularity" which is perhaps the least helpful. "You wanna get on top of the charts? Get popular then." Why didn't I think of that?
Totally agree!! The vocal loop is annoying and boring. It’s too repetitive and it lacks originality. I would prefer a syncopated synth countermelody instead.
One issue with your predictive model is that the average alone is not enough to get an idea of the shape of the real world distribution of any characteristic measurement. You need to include additional statistics, like standard deviation, at least. The average may not be at the optimal point. There may be significant clustering at disparate ranges that you're overlooking with just an average as your guide. There are fast dance and slow dance numbers, somewhere in between is not the only sweet spot. Decide which one you're going for and commit to it.
The 80% of chart topping hits is the reason why I don't listen to commercial radio, it all sounds the same. Radio plays will always take the safe bet. So yeah the idea of genre is dead, bizarre sound design and simple repetition seems to be dominating the market. I'm perfectly fine with my nine subscribers - refuse to dumb it down, maybe one day I'll make something everyone likes.
I feel you man! I think its always the same question: - Is your goal to make a popular hit? Cuz then you need to follow rules and copy. - Is your goal to express yourself without any limits? Then you have much less chance to get popular. Its always about Popularity VS Authenticity
If people are so bad ar predicting "hits" ....then The Jam song Going Underground with its lyric "the public wants what the public gets" is truer than we thought. PS... it went in at #1 in the UK and was there for 3 weeks.
Billboard's top charts are very much about the songwriter's career. It's all about networking, luck, and being in the right place at the right time. Writing songs for famous artists with a large following automatically leads to high chart positions, which in turn leads to writing songs for other big artists, and so on. There are no recipes for success, just lucky circumstances and coincidences. 😎
It sounds like something that is forced on me while being in the supermarket and if I am gullable enough will create a earworm. - You cannot write a hit song for everyone, only for those easy to please, once you get their attention. Hit songs are written to complement the attention a medium creates. Combine Visual with Audio and you got a hit. Your sexy look won't make me listen, and your sexy voice won't make me watch, necessarily.
Very nice video! Can you please link in the description the various paper cited in it? (Also, next time if you can, try to filter that ground noise under your voice, it's very annoying)
I guarantee I could have gotten faster results going to Suno, prompting very specifically and then throwing the clips into FADR to generate the midi of the beats, chords and the melody
💯/💯... This song is actually a marketing GOLDMINE... Sonically it had my foot tapping and bobbing my head Subconsciously at times .. As an artist It had me already picturing visuals to box office films like spiderverse this song is a success.
It has great arrangement and structure but it feels boring. Maybe lower the background in the beginning with a side chain compressor and also do a pitch rise for the drop in the beginning vocals
The popular Hollywood musician Hans Zimmer said something in one of his interviews, I don't remember it word to word but it was something along the line of what makes music great is Human presence! As if you are in an unknown planet and you suddenly hear a sound, maybe of a voice, or a mouth organ or a string, you would immediately get drawn to it because of the way it is imperfectly produced. There are tons of AI music channels too and at first I couldn't tell the difference but I wasn't subconsciously liking it either (I unknowingly opened a song generated by AI and I groved to it but after a while it just stopped making me feel groovy, I think it wasn't because the music wasn't well produced but maybe because it wasn't produced with the pluck of a string struck by a finger of a agitated/peaceful/frustrated human being). The same is the case with these algorithm and the way they analyze music. We need human ears to truly analyze something produced by human. Because we make mistakes, and it shows/sounds in the music. which is attractive, again, because of human presence. One of the last things we wanna be is lonely and these human presence sounds will fill us up. AI can't replace that. Finger crossed.
Interesting, and thanks for putting this together. Unfortunately, the acquisition of all the knowledge expressed in this video is very much not sufficient to enable the vast majority of people to make a hit. And ironically, not having the knowledge in this video will not stop and has not prevented a significant number of people having a hit.
Hey Matti , Loved your videos! I’ve been enjoying your content , and i saw your hiring post on YT jobs . so, do you think there’s room for a collaborative effort in video editing?
And in our data driven society the tail also wags the dog. I tell some big wig execs that I have an algorhythm to predict hits. Instead of even listening to the songs, the big wig only backs songs picked by my algorhythm. They become hits only because songs that might have been better never even got a chance.
I think you went to the supermarket and bought the right ingredients you learned from the recipe. But do you have those mysterious skills that make a good cook?
yaeh bro, try using not 2011 but 2001 studies and find some weird average bpm like 78.154367 cuz its the middle, right? ngl ur final result sounds MUCH better with 1.25 speed gl!
I'ownknow bruh, this to me takes the fun out of just making good jams.. I could be wrong but like you said, unpredictability.. I believe a hit is something jokas put money behind.. There have been thousands of hits outside of bilboard.. They push the repetition into jokas subconscious on the radio to make it popular.. I feel.. I use to work in radio and you're like, who the flip picked this, then a month later, yo stankin A$$ singing that crap.. And it's all subconscious... I say do you unless, you're on that route.. Man, outstanding video for sure, you put some thought into this bruh.. Great work.. Good mix too dude.. I see the structure or hear it.. outstanding again young man..
I think it takes the fun of out music making as well! Sure, follow the 'formula' if that's the kind of music you like making. But for me, make music for YOU not for the faceless masses and the record label execs. Appreciate the kind words and the comment
I dont like the thinking of pleasing the market, rather please the soul! It will find the right people. It may takes hundreds of years but thats how it is. Always ahead of 99% of civilization. You can still create mainstream music tho, if you got nothing to eat. Thats totally fine
ok i'm 47 seconds to your video in but I now neck deep in the paper and the data has some errors. Some they acknowledge like "artist scores" being wrong for artists who had hits before 1986 but theres a bunch like Eminem's Ass Like That being counted as a breakout song (the epitome of, no it wasn't). ArtistScore Errors (64 total): Many established artists incorrectly marked as new (e.g., Janet Jackson, Metallica, Rihanna) Particularly bad with late-career songs Duplicate Entries (13 sets): Some songs appear 2-3 times (e.g., Eminem "The Way I Am") Affects hit probability calculations
There is a sample of somebody saying something playing all the time in the song. It does not sound good.The song sounds mediocre (at least to me). I also write songs and have them on my channel and they are even worse than this one you have just made. Maybe algorithms can predict whether a song might be a hit but I don't think they can make them (I might be wrong). I think you would have a better chance of success if you followed your heart and sang something that felt true to you. Anyway, really cool and interesting video.
@@Music-By-Mattie First time I've ever seen the creator reply yes to anything like this... hope you guys end up doing some cool work together, that's nice to see. Awesome! 💯
This is absolute nonsense. There is no "science" to songwriting. There are patterns and they generally vary from song to song. Sometimes to a less extent and at other times they vary greatly.
It’s one thing to create a “hit song”. It’s another to get people to listen to it :/
It only becomes a hit song if people listen to it, before that it's just a song.
@ that’s entirely my point.
Absolutely
@@blubbblubbblubbish This is why the major labels own Spotify and spend a fortune pushing the songs on radio. It's the same with books. Most blockbuster movies will have the same amount of money that the movie took to make, just to advertise it. “Hit songs” aren't a style, it's just that labels demand certain “safe” products for financing and pushing.
if it really is a hit song, people gonna listen. It has to be more than just pretty good in production
I've been a musician for 57 years and what was true back when I started is still true; you can have the "best" track out there, but if no one listens, it goes nowhere.
I am in a number of Facebook groups , I have not heard anyone write a song that even come close delusion is a big factor, knowing what is bad and good is an important starting point
I think understanding the market is what makes hit songs. Because you understand your people, and what they want.
You have to write music that non-musicians can understand. They will eat up the simple formulaic stuff. Forget the ribeye steak, the masses want McDonalds.
I think you're closer to the answer there
yep, correct, my mate wrote a number 1 hit in 1970, he still thinks he has got it, but has no idea, because he does not know how to write to the market
Only slightly is it understanding the market. It's more about creating the market. Most of these hits are probably signed to the same company.
It's not ”Oh, you like this?”... it's ”You’re going to like this!"
Thanks for the video and all the work to put it together! There is a loose rule in advertising you need to get the product, thing etc. into someones ears or eyes 6 times before they decide to engage further like buy something. Same goes with music, which traditional radio was good at with a song playlist on rotation throughout the day. So marketing and curators make hits by feeding you the same thing over and over. All those studies are really mapping trends of what hit songs are made of, it's the diet you're fed but also there is musicality too. Case in point a hit from the 70's may not be a hit today given equal exposure and marketing etc.
Great video! Love taking the science-first approach strictly and rigidly. Now for anyone watching, the trick is to do it again, and again, and again. Creativity is like a muscle, and it helps to stretch it everyday. This can be as simple as a line, one melody, or even random ideas, but "show up to work" everyday and it will help. Using these ideas and guidelines are key, and then one day they will become ingrained naturally so that you don't have to think about it. This is when your natural style can begin to blend with the "formula" and create real gems. When the style and the science naturally happen, you can make hit songs that are hits because they are authentic, honest pieces of art that represent who you are as an artist. Wow, didn't mean to go on this long. Keep rocking!
Very well said! Keep on making music and good things will happen!
Exactly... show up every day, wherever you do your work, show up, do the work, even if it's shit, even if it's only 20 mins a day minimum, go through the motions, develop your process, then trust the process. Once you've started to set up the right conditions, and have the skill set to actually handle and record/write/capture/etc. the magic when it comes, efficiently enough, before it vanishes back into the ether.... the inspiration will start to come for you more and more, and you'll be able to capture it more and more, until one day you'll be able to walk into the studio any time and just go off any random idea, melody, progression, lyric, etc. and put it through the process until it blossoms into something truly special, that is uniquely you, that nobody else in the world ever couldv'e made. It's a long journey, but it does get alot better, funner, more fulfilling, and the music eventually does start to become very special. Also the GENERATING of the ideas, other than the inspiration we can't really explain where it comes from.... use brainstorming, song maps, writing things out, teamwork, jam sessions, songwriting sessions, whatever you gotta do to initiate things so the inspiration and magic can come. Learning alone will NEVER cause this to happen, only DOING, like it's your job, purpose, and destiny... for years on end. 💯
The learning and practice is important too, don't get me wrong, but it only FACILITATES the special magical thing so it can come through you and into your creation. It's all very important, inspiration won't even come if you can't handle/interpret/capture it yet, and if it does it'll disappear just as quick as it came. Plus if ur also the artist recording it you gotta have some legit skills in order to perform it well enough to do it justice. So I'm not tryna downplay the learning part, and DEF not downplaying the practice/rehearsal/etc. part either... but I would lean heavily towards hands-on learning and practice, rather than passive learning and watching others. Just to clarify 😅
As always the best music comes from the heart and is made for the artist, and you always gotta keep making songs true to yourself to get good and perfect your unique style. Treat yourself right, and of course take inspiration. I've found I made my best music when listening to a song once then loosely basing it on my memory of that song, which ensures that my style is retained while also helping me to pick up new skills and really make the music I want to make (I love skrillex :D)
To be honest, you left out the most important thing, the artistic skill, this is definitely a hit, and needs a very experienced singer.
Nice... Let me know when you make a video covering all of what he addressed and missed 😅
its harder to write a pop song
youtubers trying to understand science. Gotta admit it didn't disappoint , good job!
i am very disappointed lol this is extremely unscientific
@@Mossbotmusic that's why I wasn't disappointed, I knew what to expect: a complete trainwreck and a good laugh
The statistics of getting the hit song right, is about the same as getting stock prices right. Let that sink in.
Elaborate 🙏
@@elijah2460 It's pure luck...
@@elijah2460 There are a lot of judgement calls in designing a good predictive statistical model. You have to decide which parameters to include and how to weight them. If you understand the environment you're in and know what you're doing well enough, you should be able to make it work pretty closely. It's possible, IMHO.
some hit songs sound kinda meh before the sound engineers get ahold of it .
Thanks for sharing this amazing study.Very hands on to get that magic cocktail.
Where is the sound cloud link, you cleared 1000 likes 🎈 🥳 🎉
Great study. Thank you!
has he written any hit songs
The section on lyrics reminded me of the fact that most of those hard and heavy 80s rock songs..... were love songs.
Thank you for your informative content. IJuno Roland can be used as a midi board with DAWs . I’m looking forward to learning more.
you've got a great voice mate
i think the max martin rules are way more helpful in predicting or crafting a hit. many of them are directly tied to limitations of the auditory system or evolutionary arguments. an example of the first: never introduce two new ideas at the same time. humans typically can only focus on one thing at a time, e.g. it is very difficult to talk to two people simultaneously about different subjects. an example for the second is hierarchy in music: not all elements in a piece of music are equally important. once a voice appears in a piece, most people will focus on the voice. we are wired evolutionary to listen to other people, that is why it is so hard to not listen in to the discussion on the neighboring table in the restaurant.
Women can't ignore a baby crying. That's why male falsetto always melts them.
I havent looked at those papers in depth but I'd be skeptical that their algorithms would be that accurate to a brand new song. My thoughts are that their models would likely be making their predictions largely from the 'popularity' feature for each case, which would make the model very accurate to historic data and less-so for unreleased songs.
Also depends what they are sampling. If it's taking a random sample of say 1,000,000 songs from Spotify, there wont be many (if any) Top 100 songs in there, and the most distinguishable feature between a hit and a regular song from an independent artist will be the popularity score.
As you said there are so many external factors that turn songs into hits, and a lot of the time those are what actually make hits rather than the raw quality of the song
Popularity has got to be a huge factor in ranking for these algorithms. Since it's based off of the amount of plays and how recent they are, it's basically predicting hit songs off the songs that get the most plays. Well duh. That's what it seems like to me after reading the study a bit further in depth. But still, 88% still seems high to me even with that.
75% accuracy in predicting hits is very impressive when you consider just how many thousands (millions?) of songs have been released.
@@ThreadBomb what a load of crap,
aye yall should be hitting like as soon as you finish hearing how much work he put into the vid. i dont know him. but love the work already.
You overlooked a couple of things:
You need a couple of short, memorable musical hooks that get repeated a lot.
Start big, not low-key. Hits often start with the chorus or main refrain.
the big thing is 99% just dont know how to tell what is good and bad delusion its the killer, hence only a handful are successful
Former teacher , current electrician part time engineer. Have to say, your editing skills and verbal presentations are A-1.
Great vid m8! I really enjoyed this content. When are you dropping that SoundCloud link? 👀
there is also the whole idea of marketing where even if its not the best song a fun dance or it being under a lable (who would work to have it been seen) could heavily sway the popularity of it
The hit song sounds like something that Suno would make.
Its a shure i think
@@mind_reloader670 Suno is an AI song generating service. Impressive, but formulaic and forgettable.
You should have studied the producer that has made the most hot 100 hits in the last 20 years. Max Martin. There are definitive patterns that can hone in on all the things you mentioned. There are a couple of BIG things that Max does that were nowhere in your research.
Yeah.... work for a dozen giant record labels who put millions behind your songs/artists every time, that's the BIG one 😂
Nah, I'm only half serious, I'm sure he's extremely talented and I'll bet ur correct that there are some valuable insights to be gained, not that I think AIMING FOR A "HIT" is the way to go, but for general use and to become a better producer myself.... I'll research him, thanks 💯
@@CyrusAugustus LOL, True, but wern' they all funded by Big Labels? Anyway the economist looked at Max Martin and some of the patterns here: ua-cam.com/video/F9pIZVFGa7M/v-deo.html
This guy is oblivious about music in general😂
Excellent song, especially from the experimentation position. Really good. I learned a long time ago that the best and most powerful songs have simple melodies (singability, which cements it to memory), and are written in 3/4 or 6/8 (the whimsical meter I call it). I'm compsing orchestral/symphonic concert band music, after 25 yrs of homeschooling our 8 kids.😂
First of all I really learned something from this video so I subscribed. Me being an audio engineer/music producer for a living thats kind of rare these days. Second the song was catchy, your also a good singer and clearly know how to produce... However why did you decide to add that annoying "what" vocal sample that hits on every beat? This is just my opinion but that ruined the whole song for me 😅 its way too loud for a vocal chop, sounds robotic and repeats way too often. It does suite the song, but I would turn it down by at least 35-50% volume and maybe EQ the mid/high freq even more. But overall great video and song! 🙏🏼
Well done, the song works from what I heard of it.
Thank you for the conclusion! I was afraid you would say: aaaaand kids, this is how you suppose to write hits. Imho It is definitely the true challenge to aim for the 12/100. And I believe these are actually the songs that will last. Like Stings or Stevie Wonders hits. The 88/100 seem to obsolete very quickly. Because these have barely something to offer except the hit parameters.
well done Mattie! very interesting content :)
Wow!!! These is the formular everyone needs.... That all it is
Statistics: If Hip Hop is popular in the USA, and you want to break through, you either have to be _the best_ at Hip Hop, which is very hard, cuz that's where most of the contestants are, or you have to find a niche that is easier to dominate.
Or you need a lot of money for publicity.
@@ThreadBomb Hot take: You don't really need a lot of money for publicity. You can do _a lot_ just with the free tier of Instagram and Facebook.
1st time viewer = New Sub!!!
Great video... I have never heard the songs you referenced, well, at least that I know of, but I got a lot from the rest of the video.
Thank you
This is what happens when you try to measure mostly subjective traits in music as if their objective stats. It doesn't work.
For instance, the study tries to measure a songs emotion by the lyrical content, the use of major or minor chords, and the instruments used. None of those things actually matter that much in describing how happy or sad a song is. If there's any objective marker for a song's emotion it would be the rhythm and how "upbeat" it is. Even that isn't universal.
There are tons of happy songs with depressing lyrics like Maroon 5's "Misery", and sad songs with happy lyrics like "Time in a Bottle" by Jim Croce (the lyrics' literal meanings are happy but when you know the origin of the song they're quite sad; not that the study would even know the difference) The mismatch of the tone and the words actually adds its own interest. Also there are tons of sad songs in major like Good Riddance by Green Day, and tons of happy songs in minor like the majority of Rock and Roll. You can actually make the argument that it happens more often THAT way around depending on what genres you listen to. Also no matter what instrument you play whether it be guitar, piano, ukulele, violin, or vocals/choir you can make a song of ANY emotion. This should be so obvious I don't even need to give examples.
The only objective measurement in this study seems to be "popularity" which is perhaps the least helpful. "You wanna get on top of the charts? Get popular then." Why didn't I think of that?
Very good points!
Juxtaposition my beloved
it was dope just take out the vocal loop with the drums
Totally agree!! The vocal loop is annoying and boring. It’s too repetitive and it lacks originality. I would prefer a syncopated synth countermelody instead.
One issue with your predictive model is that the average alone is not enough to get an idea of the shape of the real world distribution of any characteristic measurement. You need to include additional statistics, like standard deviation, at least. The average may not be at the optimal point. There may be significant clustering at disparate ranges that you're overlooking with just an average as your guide. There are fast dance and slow dance numbers, somewhere in between is not the only sweet spot. Decide which one you're going for and commit to it.
All the best music has been written. 60s, 70s, 80s and 90s. Here endeth the lesson.
The 80% of chart topping hits is the reason why I don't listen to commercial radio, it all sounds the same. Radio plays will always take the safe bet.
So yeah the idea of genre is dead, bizarre sound design and simple repetition seems to be dominating the market.
I'm perfectly fine with my nine subscribers - refuse to dumb it down, maybe one day I'll make something everyone likes.
I feel you man! I think its always the same question:
- Is your goal to make a popular hit? Cuz then you need to follow rules and copy.
- Is your goal to express yourself without any limits? Then you have much less chance to get popular.
Its always about Popularity VS Authenticity
I loved this video. ❤Had to subscribe! Gearing up to write one of the 12 the algorithm never saw coming. (What an epic way to end a video btw)😊
it's already over the 1k goal of likes, where's the soundclound link? :)
Love this song bro, great content
If people are so bad ar predicting "hits" ....then The Jam song Going Underground with its lyric "the public wants what the public gets" is truer than we thought. PS... it went in at #1 in the UK and was there for 3 weeks.
Hmmm it’s a vibe for sure
This sounds like AI made it. Well done, you;ve found the formula for the lowest common denominator.
How many hit songs have you made sir? Can you name them? I wanna add it to my playlist..
pretty much what like every song playing in the grocery store sounds like lol
Man this song is a bop!
Wow, nice work, thank you a lot.
watching you from Europe, so you are international dude
This was a banger!
Billboard's top charts are very much about the songwriter's career. It's all about networking, luck, and being in the right place at the right time. Writing songs for famous artists with a large following automatically leads to high chart positions, which in turn leads to writing songs for other big artists, and so on. There are no recipes for success, just lucky circumstances and coincidences. 😎
It sounds like something that is forced on me while being in the supermarket and if I am gullable enough will create a earworm. - You cannot write a hit song for everyone, only for those easy to please, once you get their attention.
Hit songs are written to complement the attention a medium creates.
Combine Visual with Audio and you got a hit.
Your sexy look won't make me listen, and your sexy voice won't make me watch, necessarily.
Very nice video! Can you please link in the description the various paper cited in it?
(Also, next time if you can, try to filter that ground noise under your voice, it's very annoying)
I'll link all of the papers I used in the description!
Also, not sure what you mean about the ground noise. I'm not hearing anything on my end...
I guarantee I could have gotten faster results going to Suno, prompting very specifically and then throwing the clips into FADR to generate the midi of the beats, chords and the melody
Can you link the article?
💯/💯...
This song is actually a marketing GOLDMINE... Sonically it had my foot tapping and bobbing my head Subconsciously at times ..
As an artist It had me already picturing visuals to box office films like spiderverse this song is a success.
dude, that's a banger.
What's the name of the song & where is the link?
Amazing content ✨
Great video utilizing data and studies to simplify the concepts behind hit songs, made it so simple to understand
The most popular song of all time was written in 3/4 time signature. Does that mean we should write more songs in 3/4 ?
Pretty good song. A bit more rhyme with the lyrics maybe. Bit nice vibe.
Thanks for the feedback!
It has great arrangement and structure but it feels boring. Maybe lower the background in the beginning with a side chain compressor and also do a pitch rise for the drop in the beginning vocals
This feels very general still kind of more like platitudes than specifics
there is a hum in your audio ,great video by the way
The popular Hollywood musician Hans Zimmer said something in one of his interviews, I don't remember it word to word but it was something along the line of what makes music great is Human presence! As if you are in an unknown planet and you suddenly hear a sound, maybe of a voice, or a mouth organ or a string, you would immediately get drawn to it because of the way it is imperfectly produced. There are tons of AI music channels too and at first I couldn't tell the difference but I wasn't subconsciously liking it either (I unknowingly opened a song generated by AI and I groved to it but after a while it just stopped making me feel groovy, I think it wasn't because the music wasn't well produced but maybe because it wasn't produced with the pluck of a string struck by a finger of a agitated/peaceful/frustrated human being). The same is the case with these algorithm and the way they analyze music. We need human ears to truly analyze something produced by human. Because we make mistakes, and it shows/sounds in the music. which is attractive, again, because of human presence. One of the last things we wanna be is lonely and these human presence sounds will fill us up. AI can't replace that. Finger crossed.
I suppose if you want to make cookie-cutter fast food McMusic for the masses, this is great.
15:34 sounds Classical ❤
listen back to your song sped up like 1.5
Idk i enjoy my music always. Its simple but i also don’t take longer than 4 hours on any song.
Interesting, and thanks for putting this together. Unfortunately, the acquisition of all the knowledge expressed in this video is very much not sufficient to enable the vast majority of people to make a hit. And ironically, not having the knowledge in this video will not stop and has not prevented a significant number of people having a hit.
awesome song, how did you learn how to sing? and how long did it take to get good?
What app do u use to write lyrics?
Naah this is actually too good. I would actually listen to it again and again
Hey Matti ,
Loved your videos! I’ve been enjoying your content ,
and i saw your hiring post on YT jobs . so, do you think there’s room for a collaborative effort in video editing?
And in our data driven society the tail also wags the dog. I tell some big wig execs that I have an algorhythm to predict hits. Instead of even listening to the songs, the big wig only backs songs picked by my algorhythm. They become hits only because songs that might have been better never even got a chance.
I think you went to the supermarket and bought the right ingredients you learned from the recipe. But do you have those mysterious skills that make a good cook?
And do you have a restaurant to attract customers?
tks u!!!
I’m pretty sure Eminem used the titles for that set (titles) for lyrics
Its actually good bro give it to an artist 👏
whats the mic u use called?
yaeh bro, try using not 2011 but 2001 studies and find some weird average bpm like 78.154367 cuz its the middle, right?
ngl ur final result sounds MUCH better with 1.25 speed gl!
Did all the algorithms take into consideration the DIDLER EFFECT 😂 I’m just saying it’s a real thing that’s been exposed??
What about of the level of cheesiness ???
What’s the secret for producing a song that is still listened to twenty years later?
great video minus the ground noise in your audio
>>> Saludos
Hii
I'ownknow bruh, this to me takes the fun out of just making good jams.. I could be wrong but like you said, unpredictability.. I believe a hit is something jokas put money behind.. There have been thousands of hits outside of bilboard.. They push the repetition into jokas subconscious on the radio to make it popular.. I feel.. I use to work in radio and you're like, who the flip picked this, then a month later, yo stankin A$$ singing that crap.. And it's all subconscious... I say do you unless, you're on that route.. Man, outstanding video for sure, you put some thought into this bruh.. Great work.. Good mix too dude.. I see the structure or hear it.. outstanding again young man..
I think it takes the fun of out music making as well! Sure, follow the 'formula' if that's the kind of music you like making. But for me, make music for YOU not for the faceless masses and the record label execs. Appreciate the kind words and the comment
The question should be asked: Can I make a hit song if I can't sing or play an instrument?
The song is almost a hit🙂
Appreciate it haha
Well okay but what is your hit song?
I dont like the thinking of pleasing the market, rather please the soul! It will find the right people. It may takes hundreds of years but thats how it is. Always ahead of 99% of civilization.
You can still create mainstream music tho, if you got nothing to eat. Thats totally fine
Nice vid
ok i'm 47 seconds to your video in but I now neck deep in the paper and the data has some errors. Some they acknowledge like "artist scores" being wrong for artists who had hits before 1986 but theres a bunch like Eminem's Ass Like That being counted as a breakout song (the epitome of, no it wasn't).
ArtistScore Errors (64 total):
Many established artists incorrectly marked as new (e.g., Janet Jackson, Metallica, Rihanna)
Particularly bad with late-career songs
Duplicate Entries (13 sets):
Some songs appear 2-3 times (e.g., Eminem "The Way I Am")
Affects hit probability calculations
oh well I guess that paper is out of date anyway
What is that app you write on plz?
The app is excalidraw
There is a sample of somebody saying something playing all the time in the song. It does not sound good.The song sounds mediocre (at least to me). I also write songs and have them on my channel and they are even worse than this one you have just made.
Maybe algorithms can predict whether a song might be a hit but I don't think they can make them (I might be wrong).
I think you would have a better chance of success if you followed your heart and sang something that felt true to you.
Anyway, really cool and interesting video.
perfect guide to make soulless music
Idk if there is a 'formula' for a hit song.
hey you need a video editor ?
I do! haha message me at mattie@musicbymattie.com
@@Music-By-Mattie First time I've ever seen the creator reply yes to anything like this... hope you guys end up doing some cool work together, that's nice to see. Awesome! 💯
Sssoooo...This means that my symphonic composition should first be tested on the rats in my barn. If the cats missed any. 🤣🤣🤣
This is great. Pretty soon we'll all be writing hit songs.
This is absolute nonsense. There is no "science" to songwriting. There are patterns and they generally vary from song to song. Sometimes to a less extent and at other times they vary greatly.
How many likes would it take to get this on Spotify? :D