When Siddharth Bhardwaj and Mansoor Rahimat Khan started building AI-based music creation tools, they always had the idea of allowing creators to create original background music without the hassle of licensing. Bhardwaj, who in 2021 launched his Bengaluru-based Beatoven.ai with Khan, said:
31-year-old Bhardwaj has always wanted to combine music and technology, so after graduating from IIIT-Allahabad, he completed his master’s degree at UPF’s Music Technology Group in Barcelona, Spain. Having previously worked with startups specializing in signal processing, deep learning and music technology, Bhardwaj understood artificial intelligence and its applicability to music generation. After Bhardwaj met his Khan, a seventh-generation sitar player in his Dharwad gharana, it was a natural progression for him to start his own AI-focused company.
Music licensing and the copyright issues that come with it can make any creator uneasy, especially those new to the field. Do I have the right to use this background music in my podcast? How long does the copyright last? What is royalty-free? This is where this AI-based music creation tool makes sense. Empower creators by simplifying how to create original, royalty-free background music for your videos and podcasts.
According to Bhardwaj, the music generation AI tool works like any other music director. Basically, you have to set the length of the song you want, pick the tempo, genre, etc. and let the AI analyze the database and give you five of his songs that sync with the theme of the video. Or a podcast. You can then publish and release your video or podcast wherever you like. You can customize the sound and tweak the instrument layers according to your needs.
The system works using deep learning networks, a type of AI that analyzes large amounts of data. “The AI arranges all the loops and samples we get from the artists … we train an AI-based model to generate a structure that combines all these layers vertically and horizontally to create a coherent whole track. create, ”he explains Bhardwaj. “For a 1 minute track, it takes him 10 seconds for the AI to generate 5 options for him.”
Bhardwaj says his company works with hundreds of musicians to create 10-second guitar riffs and 15-second piano melodies for the platform. Beatoven.ai’s database contains 50,000 samples of him collected from 200 artists.
“This is not a tool for musicians,” explains Bhardwaj. “I can’t upload any kind of sound… It’s an instrumental background music tool for videos and podcasts.”
“You can’t train an AI to make great music,” he says. “The biggest problem with music is that unlike text and images, copyright is strictly enforced,” he says, making it nearly impossible to get historical music data to train his AI model. explains why. “Nobody can do that. That’s why we even choose our own datasets.”
AI can teach you how to compose music, help freelancers mix and master, and help artists generate ideas, but it cannot replace musicians. can not. Many recordings of his artists have experimented with using AI to generate lyrics, the most famous being David Bowie’s song “Hello Spaceboy.” In 2018, musician and tech researcher Francois Pachet released Hello, World, his first pop album composed of artificial intelligence.
“For creative endeavours, like music, it becomes a collaboration between humans and AI,” says Bhardwaj. “AI has already created something for you, but it is your job to curate it and customize it to your needs. I need to,” he adds.
Bhardwaj believes AI is not only cost-effective, but has the potential to increase engagement and improve storytelling culture with compelling background music. “Everything you hear on our platform is not synthetic or AI-generated, but always human-composed music,” he said.
Bhardwaj’s company makes money through a subscription-based model, with a base tier of $20 per month and a top tier of $100 per month. He has over 28,000 users, mostly YouTubers. However, Bhardwaj says he has seen someone use the tool in the background to create his score and video he also uses it in games.
According to Bhardwaj, the demand for royalty-free music in India is huge, with searches for such AI tools estimated at 5-6 lakhs. To convert more users, Bhardwaj’s company is gathering data and working to build AI models, accelerating its efforts to focus on Indian classical and royalty-free regional music. In the coming months, Bhardwaj plans to launch another AI-based tool that analyzes videos, reads scene changes and scene moods, and tries to serve the best music for the same.