Audiobus: Use your music apps together.
What is Audiobus? — Audiobus is an award-winning music app for iPhone and iPad which lets you use your other music apps together. Chain effects on your favourite synth, run the output of apps or Audio Units into an app like GarageBand or Loopy, or select a different audio interface output for each app. Route MIDI between apps — drive a synth from a MIDI sequencer, or add an arpeggiator to your MIDI keyboard — or sync with your external MIDI gear. And control your entire setup from a MIDI controller.
Download on the App StoreAudiobus is the app that makes the rest of your setup better.
Making songs memorable
A month ago I met an 18 year old guy in Texas who has written thousands of songs during the past few years. He said that he usually writes between 5 and 10 new songs per day. On a good day he can do 20 or 30.
I only listened to roughly a dozen of the songs from start to finish. They were all good. Then I tried bits and pieces of many more songs, just by randomly clicking my way through folders. I liked what I heard, but I can't really judge his overall output.
The guy had a system. He explained some of the details. He listed various apps and software that he uses to write songs. The songs are entirely apps and software. He doesn’t play any physical instruments.
If I understood him halfway accurately then he uses a combination of music theory and software to quickly knock out the foundation and framework of songs. Then he gets into the custom construction that makes each song unique.
A lot of what he explained was beyond my level of understanding. He’s clearly onto something, and it’s unlikely that he’s a singular case. I think that thousands of people like him are going to change music.
Right now humans are still better at songwriting than algorithms are at songwriting, but I don’t expect that to last. I think that software will soon be writing beautiful and transformative music.
Songwriters will be able to pick from hundreds of great voices that can sing whatever lyrics are fed into the software. It will sound like humans singing because actual human voices are built into the software.
Virtual instruments are making progress. I don’t see any reason to believe that the progress will stop. It’s more likely that virtual instruments will gradually become better than humans at playing most things.
Any great guitar part played by a human will be analyzed and replicated in hundreds of variations that any songwriter can use. Everything will be available, with every existing instrument.
Humanization technology will improve to the point where there’s no way to know if something is played by a person or by software. I don’t know how long that will take, but it seems to be inevitable.
People aren’t likely to be completely replaceable. I’m not sure that recorded music will have much value in the future, but live music played by live people will probably remain viable. I can hope anyway.
For now my songwriting attitude is to add in lots of raw ingredients, with as much dirt and sex and love and anger and mystery as I can conjure. A hundred years from now I want people to still know that I was once here.
Comments
Cool Post. Interesting thoughts.
However there will always be a demand for 'human-mind' compositions.
No software will ever write a 'Bohemian Rhapsody'.
>
It it possible to create such bombastic variance within a song using software. But the imagination required to use that software and assemble the blocks in a pleasing, innovative fashion requires a human.
Similarly, while software can suggest words or themes, coming up with a concept such as Bohemian Rhapsody and turning that into a poetic lyric is beyond AI.
On the voice side of things, Google is getting scarily close to perfection. One they turn this tech toward song, then we will be able to buy John Lennon or Scott Walker, et, etc, as downloadable vocalist.
https://youtu.be/IFDnJeLgIZk
What level of polish are we talking about with these songs of his? Completely mixed electronic productions with vocals, without, more details please
20 to 30 more or less complex really finished songs per day (even a good one) is not some something possible. Finding 20 or 30 melodies, chord progressions and beats ... not almost sure it is. Even it the guy works 20h a day it makes a song every 45/60 minutes. Even if it was possible don’t know where can be the artistic fun here. But it’s sure that having some methodology can make a musician more productive. But quality, originality, artistic vision and real fun still needs to remain. Sometimes 18 years old people like to exaggerate numbers. Some of them have 20 or 30 girlfriends per night!
Does he write the lyrics?
If the robots take all our jobs, what will be left for us actual people?
Music is probably the most subjective art form there is. What can be seen as a "classic" to one may be seen as old cliched shit to another.
I like arpeggiators & a small degree of 'generative' composition (like Fugue Machine's capacity for turning one phrase into a completely different form) but the thought of 100% computer created "AI Music" sounds about as appealing as flaying of the scrotum.
I'm not one of those GenX boomer kids that believes music hasn't been good for 20+ years but I'm also a realist. The popular music of the day is the barometer most will use to judge the music of a certain time. Yes there's always underground, fringe musical movements that are the exception to conventional wisdom. But anyone who can say the music of 1969 to 1975 is not as good or better than the music of 1999 to 2005 is suspect at best.
One of the hallmarks of the period of the late '90's to now is the heavy reliance on heavy digitally creations over people playing in a room together with the red light on. Just my opinion. I love apps like DrumPerfect & Patterning but I would choose having a bigger space, nice drums & cymbals, some good mics set up right & a really good drummer playing over ANY digital drum solution any day.
@Janie hit on it by saying live music will always remain. But for many of us here recording & a one man band/producer approach is what we either have to or choose to do. In that context some of the great digital technology available is a real positive: let's us play cool Moog sounds without spending $3000 on the old hardware; allows us to record guitar in the middle of the night with amp sims, etc. But once the technology usurps every role a songwriter, musician, producer is responsible for then to me it isn't music anymore. It's digitally created, soulless muzak, and I don't want any part of it.
I think you're missing an essential element of music as a connection between a performer and a listener. These "songs" you mention are potential vehicles for a performer to interpret for an audience. People connect to the person associated with the delivery of that song. The act of writing songs as a craft can be computer generated but the audience will want context and without a human to connect to they will consider the song to be irrelevant to their lives.
This artist producing 20-30 songs a day... do they get performed or even listened to by an audience or listener? You heard them. Did you want to have copies to play again? Would you pay for them?
There are factories that pump out Elvis on Velvet paintings but do you know anyone that owns one (proudly) and thinks there's a gifted artist behind that. They just love Elvis. Even after all this time.
Don't be afraid of the computers or the technology. They are still just wonderful tools we can deploy to connect to each other. We create to communicate.
Don't worry about the second sentence now; it won't matter to you then. The first sentence? Brilliant. A reason to live. Keep on.
Thirty songs in one day makes me think of Van Morrison’s contractual obligation album (also known as the Bang Sessions).
Apparently he owed about thirty songs to the record label he was on at the time, and had a better offer from another label in the table.
Does he sit down and write thirty brilliant songs for his current label? No, he goes into the studio, armed with an acoustic guitar, and records, one after another, thirty absolute garbage songs (some less than a minute long, many of them virtually the same song with slightly altered lyrics). It’s one of my favorite music industry stories and a hell of a hilarious album to boot.
Im having trouble wrapping my brain around what this kid is up to. Even if I already had the ideas and tools ready I can't imagine working that fast and being satisfied with anything. Why give up on a good song after 30 minutes, for sure it could be better with more work. Id rather have 1 hit I slaved over than 50 filler tracks made quickly as possible.
If I really pushed myself I bet I could generate 20 complete songs in a day using apps like Odesi, Figure, ikaossilator and blocs wave but I probably wouldn't like any of them and would get no artistic satisfaction from it.
Without having heard these 5 to 30 songs a day, I remain skeptical.