FEATURE:
New Waves
IMAGE CREDIT: Irenebination
The Use of AI Compositional Tools That Reduce Studio Costs and Foster Creativity
_________
I have published features…
PHOTO CREDIT: ThisIsEngineering/Pexels
in the past that talk about various styles of music we do not hear anymore. Certain sounds that are rare for various reasons. One of the most recent features discussed Steely Dan and Donald Fagen. How there are hardly any artists today where you can hear their influence. Not such an obscure artist(s), why do we not hear anything Dan/Fagen-esque in modern music?! Some suggested the sheer studio costs put people off. Maybe the creative and intellectual levels needed to write such music. I would argue people could pen something similar to a Donald Fagen song that was really great without having to be exactly as good. In terms of composition, I feel that people could have ideas and starting blocks that could form into these fully-formed songs. Maybe time in the studio experimenting would take a lot of money and time. Is the sheer cost of recording at professional studios with musicians limiting creativity?! Not that there is anything wrong with small studios or D.I.Y. recording. Indeed, some of the very best albums can be made for very little. In terms of doing something bigger and more involved, I guess you need quite a few musicians. If you had to do fewer takes and had the compositions fully realised before they get to the studios, then that would help creativity and future inspiration. Rather than using AI to completely replace artists, one of its benefits is that there are algorithms and options where you could put together notes, lines and full compositions into a demo and then present to musicians. This is nothing new you’ll say, yet AI offers something more detailed and comprehensive. Rather than it being mere building blocks, you could work together these compositions that just need to be repeated in the studio – still costing a little but not as much as doing endless takes.
IN THIS PHOTO: Steely Dan’s Walter Becker and Donald Fagen in 1977/PHOTO CREDIT: Chris Walter/WireImage/Getty Images
There are programmes and software like AIVA, that allow anyone of any level to create songs in a variety of styles. There has been a lot of discussion about AI in music over the past year or so. We have seen entire songs and albums generated by AI. There are definite disadvantages of AI in music around authenticity and quality. Songs through AI that do not sound that good or lack the human depth and nuance. I think we are a very long way away from AI being able to replicate musicians or replace them. Artists fear they can have their voices used without permission. That we will have a scene of AI music. This real nervousness around technology in music. There are benefits to AI. Think about The Beatles’ final song, Now and Then, which was released last year. That was a John Lennon demo that featured the other three Beatles. AI was used and made it sound like a full and proper band collaboration. AI can give you tools to produce like a professional. Rather than replace those essential people, it allows options for artists who want to self-produce or do not have the budget for a producer. Thinking about the fact that some bands and sounds should have modern-day torch-carriers but do not. I wonder whether studio fees and a daunting hill to climb puts people off. AI does make it more likely that artists today can create quite layered songs with incredible musicianship. They do not need to spend weeks in the studio. This LinkedIn feature discusses AI-composed music and its benefits:
“Introduction
The fact that artificial intelligence can produce music is exciting and scary at the right time for many people. With technological advancement, AI has left no stone unturned and is now doing wonders in the music industry. AI has significantly grown in music composition! Thanks to the development of sophisticated algorithms and the rising availability of large datasets. The algorithms use machine learning techniques to analyze unique music patterns and generate new compositions.
Through this newsletter, let us explore how these algorithms work in a symphony to hit the right notes of the music.
PHOTO CREDIT: Blaz Erzetic/Pexels
AI Algorithms Composing Tunes
Data Collection: The process starts by collecting massive amounts of music data, including audio recordings and user-generated compositions. This data gives the algorithm a diverse range of musical styles and genres to learn from.
Machine Learning Models: Deep learning is the most common machine learning approach in music composition. These are designed to process sequential data and learn temporal dependencies, making them suitable for capturing musical patterns over time.
Model Training: The algorithm is trained on the collected music data to analyze the patterns and structures in the music. The model learns the relationship between notes, chords, rhythms, melodies, and other musical elements.
Creativity and Generation: The trained model can generate new musical compositions, take input in melodies or notes, and use this knowledge to develop music further. The generation process is often guided by specific parameters or constraints set by the composer or the user, such as genre, mood, tempo, or musical motifs.
Evaluation and refinement: Not all generated compositions will be musically appealing. Therefore, AI composers often use evaluation metrics or feedback mechanisms to check the quality of the generated music. They may compare the AI-generated music against a large corpus of existing human-created music to ensure that the generated output aligns with established musical norms.
Repetitive Improvement: The AI algorithms meticulously improve by analyzing feedback and user interactions and refining their output over time. This process of continuous learning enables the algorithm to produce increasingly better compositions.
Collaborative tools: Some AI music composition tools are designed to assist human composers rather than replace them entirely. These tools can provide suggestions, harmonizations, or variations based on the composer's input, helping to speed up the creative process and inspire new ideas.
Conclusion
While technology is continuously evolving, AI is becoming an essential tool in the creative process for musicians and composers worldwide. However, it's important to note that while AI in music composition has demonstrated impressive capabilities, it still faces challenges in fully understanding and replicating the complexities of human musical creativity and emotion”.
PHOTO CREDIT: Wallace Chuck/Pexels
It is important to make a distinction. I do think that AI creating entire albums and tracks seems rather pointless. I would not want to buy any albums where all the music was made by AI. I know that there are plenty of albums that are made digitally. Electronic albums that have been created by machines. I mean AI replacing instruments and voices. This attempt to replace humans. With some saying they cannot write and record certain types of music because of the incredible studio costs and time to get the sound write. AI tools and algorithms can allow composers and artists to exactly hear what is in their heads. In terms of incurring studio costs, when musicians know that what a composer wants, they can record their part remotely. I favour musicians in the studio though, rather than having AI replace their parts, a remotely-recorded take would be fine. It takes me back to my concern that there are certain groups and artists (like Steely Dan or even The Beatles) where you cannot really hear their influence today. Not the more experimental side of them anyway. This valuable and hugely rich sound of music that is stuck in the past and is begging for modern-day revival. I can understand that there are limited budgets. I don’t feel a lack of talent is the reason for an absence of certain sounds and updates on older artists who should be adopted by new artists. AI is a way of inspiring melodies and harmonies. A way of building compositions and cutting costs because of the reduced studio hours. Then, once a composition or song is worked up by AI, it can then by played by real musicians. Rather than AI being used to replace existing artists or revive deceased ones, the technology is so useful as a way of working up demos and almost-realised full features before working them in a studio. Even for non-musicians like me, is makes compositions and songwriting a lot easier and more intuitive. Let’s hope more and more artists do use AI this way and we will see a revival of older and wonderful sounds…
PHOTO CREDIT: Thapelo Boateng/Pexels
MISSING in the present day.