ReMidi is a tool that can help with the split and recombine.
What are some ‘intelligent’ orchestration tools to help break me out of my box?
- Thread starter DennyB
- Start date
Tatiana Gordeeva
Long time member
Never heard of it. I'm intrigued by the website description... Thanks for sharing!
Ha! The company is my "neighbour" !
Last edited:
youngpokie
Senior Member
Despite the derision of AI, this view nails it in my opinion.
Everything in the process of making any art is a series of logical operations. Perfect voice leading is a great example of this: it's just a set of logical rules that are easy to codify and program. Another - orchestration techniques (e.g. chord voicings). Another one - chord progressions. Modulations. And so on.
The first genuinely "convincing" AI music is probably going sound extremely generic, because it will follow all these rules to a T. But then at some point, a "humanization" feature will be introduced, followed by a question - can AI actually make ART?
Last edited:
Remidi targets a specific task. The included public domain midi help you learn how to use it. It becomes more useful when applied to your own midi. You can accomplish the same results with lots of cut/paste. But, just like an arpeggiator, you can quickly try out many permutations.
Tatiana Gordeeva
Long time member
Thank you for your kind words and thoughtful opinion. I agree 100%. I invite you to take a look at this other thread where I posted something that you might find pertinent and even maybe interesting:
Music Making Tools, or Tools Making Music?
Dear composer friends, I'm hoping I've posted this in the right section! I have recently completed a Masters course in 'Sound for the Moving Image' at Glasgow School of Art (UK). My final dissertation/thesis was a research project on media composers, music technology and artificial intelligence...
vi-control.net
antret
Active Member
In-house music prototyping solution: splitting a bunch of Orb midi files into 1 bar chunks and recombining based on scale, key and harmonic content of adjacent bars. Crud but seems to work
Cool idea! I may have to give that a go some day.
Crossroads
Senior Member
But I for example am not using an orchestra. I am using orchestral SAMPLES inside said sequencer. The tools and tech are part of what inspires me. The tools can definitely be leading if used right.
InLight-Tone
Senior Member
Willowtree
I make music
Tatiana Gordeeva
Long time member
Thank you @InLight-Tone. Interesting video. A few years ago I toyed around with OpusModus during my H.A.C.K. phase.
I finally opted for cellular automata as the basis for this piecewhich is now in dire need of re-working the piano samples used (Ivory II if I remember well) and re-mastering. Score is available. I might post an extract here later if anyone is interested.
I noticed that Julio mentioned working from the DNA level of the music, the genotype. I also came to this conclusion but moved to the phenotype level instead as shown here
upon which I expanded in this recent post
Horizon 2025: What does it have in store for us, musicians?
Consider the following statement: Consider music, already 99.9% of all music is experienced virtually; less than 0.1% of music is experienced by physically listening to physical musicians physically performing in real time. Is the world worse or better because of this fact? Music has been...
vi-control.net
Last edited:
ProfoundSilence
Senior Member
Edit : have fun, but that won't be orchestral music, that will be sequencing music using orchestral samples in a way that has nothing to do with an orchestra.
Also not important to the OP, because it doesn't sound like that's what they are doing either
Last edited:
chocobitz825
Senior Member
There are a few tools for kontakt that auto-generate some ideas. Sonuscore "the orchestra", and "Elysion"....as well sonokinetic has a bunch of fast orchestration tools mixed with loops and drag & drop midi.
orb composer is a bit more flexible, but might not get much support moving forward. I still have yet to try Orb Producer for orchestral purposes.
In my opinion, rapid composer and synfire are nightmares to use.
Captain Chords and Scaler might not do any full-on composition work but can be good tools for setting the groundwork before you start going deeper into the composition.
Other midi tools would include harvest, cream, Melody Sauce, Nora, Obelisk, Phrasebox (pretty good actually), Cthulhu, Kameleono, Riffer, ReMIDI, and EZKeys. Some are just arps, some are reharmonizers, some are riff generators. I actually find EZkeys to be a bit helpful in that if you feed it various user-generated midi, and match it to your chord structure you can find some interesting alternatives. Only problem is it can't voice lead at all and tends to create some whacky midi that needs to be transposed and edited a bit to sound more natural.
Use whatever tool inspires you. I don't particularly care what the purists say. All the masters they love might do things the same way, but almost all the people here referencing those masters ARE NOT those masters, regardless of how much they mimic the same method, and they will likely never be remembered as much as their idols either. All of those skills are valuable, and it's good to work on them, but if you just want some outside input, go for it. No one on the listening end will care how you got there.
orb composer is a bit more flexible, but might not get much support moving forward. I still have yet to try Orb Producer for orchestral purposes.
In my opinion, rapid composer and synfire are nightmares to use.
Captain Chords and Scaler might not do any full-on composition work but can be good tools for setting the groundwork before you start going deeper into the composition.
Other midi tools would include harvest, cream, Melody Sauce, Nora, Obelisk, Phrasebox (pretty good actually), Cthulhu, Kameleono, Riffer, ReMIDI, and EZKeys. Some are just arps, some are reharmonizers, some are riff generators. I actually find EZkeys to be a bit helpful in that if you feed it various user-generated midi, and match it to your chord structure you can find some interesting alternatives. Only problem is it can't voice lead at all and tends to create some whacky midi that needs to be transposed and edited a bit to sound more natural.
Use whatever tool inspires you. I don't particularly care what the purists say. All the masters they love might do things the same way, but almost all the people here referencing those masters ARE NOT those masters, regardless of how much they mimic the same method, and they will likely never be remembered as much as their idols either. All of those skills are valuable, and it's good to work on them, but if you just want some outside input, go for it. No one on the listening end will care how you got there.
chocobitz825
Senior Member
ok, but realistically speaking....in the grand scheme of things, most of us are nobodies worth anything.....
ProfoundSilence
Senior Member
So who do you think they should listen to? Other nobodies? Or the same advice that's been common practice for hundreds of years with consistently good results?
Willowtree
I make music
I respect you a lot chocobitz, but I don't think it's most of us are "nobodies worth anything". Lots of talented folks on here, whether they're professionally / commercially "successful" or not.
And even then, while I don't think there's anything bad with using outside tools to help learning orchestration, I think orchestration's primary purpose is in making whatever you're writing playable by real players, or (more relevant to samples) help you form a coherent consistent whole (where conscious human decisions about themes are essential, eg. oboe is the duck, clarinet is the cat, see Peter and the Wolf).
Ultimately, a good composition orchestrates itself. Good voice leading and harmonic voicing is where it is at.
This is where I'm coming from. A lot of people on here seem to confuse orchestration with arranging or composition.
chocobitz825
Senior Member
they can listen to whoever they want. they're responsible for themselves and the results they get. People throw this advice out with no consideration of what the goals of the person are, or what "realistic advice" actually is. If this person is creating for the sake of making their own art and music, then anything is fine. If this person is trying to be a consistent professional that can push out orders on time, AI tools likely won't help them as much as traditional methods. If they're trying to sell their music to the public, then none of it really matters. Most listeners don't know or care, and what works works. Tons of trap beatmakers and EDM producers are making money "cheating", and no one cares.
My point was not about the talent of this group, its a response to the comment
"But hey, what does everyone else know /shrug if you gave common sense advice it's elitist, so you need to give advice that nobody worth anything follows".
Few of us are John Williams, or Hans Zimmer (take those two examples for what you will). we're a vast range of creatives running from hobbyist to professional, and the idea that we need to chain ourselves to the "common sense" practices of composition just because famous composers do/did it that way is pointless because just doing it their way does not make us those composers. There is value in learning it, but mandating it must be done that way is wrong. There are lots of simple but enjoyable songs...there are also plenty of pieces done by the book that are forgettable. In that reality, what does it mean to tell people that writing without tools or a DAW is "better"? These AI tools won't ruin art, and writing score to paper won't automatically make us musical geniuses or legends in the field. So we should just make music, enjoy music, and calm down about all the stuff in between.
Willowtree
I make music
Until the AI replaces us as composers and we are forced to bend the knee to them and accept their supremacy.
(I joke, but also ... #AIVA)
ProfoundSilence
Senior Member
I give up.
What a pointless conversation to have.
We aren't world strongest man's, but if they all give the same advice for getting stronger - there's a 99% chance it applies to you, and is more worth listening to than some bullshit your coworker regurgitates. We have actual college professors that visit the forum, should their students stop listening to them because they won't be john williams?
It's not defeatist, it's just stupid. There are plenty of orchestrators that will tell you about all the mistakes they've made and you think AI will do better?
I can't wait until orb composer and it's kin write 16th note figurations that last eight measures long for the trombone
chocobitz825
Senior Member
I give up.
I’m not saying AI is better, or will produce a better orchestration by any traditional measure. What im saying is, art doesn’t exist for the sake of tradition. It’s not validated by the method in which it was made if it was made for the sake of artistic expression. Whatever tool gets the OPs ideas out, then fine.
I don’t know about you, but I work in a field where the most talented people are not always the most successful, they’re not the ones working the most, and not the ones with the most accolades and quite often they’re not that unique either. Oddly enough many of them are also not particularly satisfied with the results of all their effort.
im not saying throw out the traditional knowledge and I even said with some of my suggestions you need to know good composition to correct the mistakes AI and algorithms makes. If OP wants to try some idea generators, it matters to no one else but them.
No but this is the classic trend of academia that pushes the purity of tradition, condemns the new, until the new becomes standard and gets added to the curriculum without any sense of humility.
I’m not saying AI is better, or will produce a better orchestration by any traditional measure. What im saying is, art doesn’t exist for the sake of tradition. It’s not validated by the method in which it was made if it was made for the sake of artistic expression. Whatever tool gets the OPs ideas out, then fine.
I don’t know about you, but I work in a field where the most talented people are not always the most successful, they’re not the ones working the most, and not the ones with the most accolades and quite often they’re not that unique either. Oddly enough many of them are also not particularly satisfied with the results of all their effort.
im not saying throw out the traditional knowledge and I even said with some of my suggestions you need to know good composition to correct the mistakes AI and algorithms makes. If OP wants to try some idea generators, it matters to no one else but them.
chocobitz825
Senior Member
It’s funny because it takes a lot of skill to make a piece from AIVA sound like music
