Introduction:
A new firm named Tape It came in to fill the hole left by Apple’s Music Memos app, which musicians used to generate song ideas. Tape’s software uses artificial intelligence to detect the instrument and annotate the recording automatically. With the Deliver Studio-Quality Noise Reduction Via AI noise reduction algorithm driven by AI that works on any audio, not just speech, that startup is moving forward in its quest to revolutionize the audio-recording process.
To license the technology to vendors in the future, the AI denoiser was released this week as a free web application. The company’s flagship Tape It app will subsequently incorporate it, according to the company.
Thomas Walther and Jan Nash, two musicians and friends, founded Tape It in 2020 to create an iOS recording software for musicians. Walther spent three and a half years at Spotify before starting Tape It after selling his audio recognition business, Sonalytic, to the music streaming service. While Nash is an opera vocalist with a background in classical music, he is also a bassist and an engineer.
The two were first motivated to create Tape It because they wanted a tool for themselves as bandmates that would be as user-friendly as Apple’s Music Memos but more potent thanks to the application of AI.
The initial app was able to automatically identify the instrument and then annotate the recording with a visual cue so that users could more easily locate those recordings by looking for the bright icon. Additionally, musicians could add their markers, notes, and images to the files for subsequent examination.
This resulted in the startup’s most recent creation, an AI-powered denoiser developed for the previous two years. The issue with recordings, according to the business, is background noise. Musicians use sophisticated software and record in studios to minimize background noise and electrical interference. Tape aims to use AI to offer a less expensive option. Their program automatically eliminates background noise, such as hums and hisses, as well as spoken word, to produce field recordings, single-instrument tracks, and studio-quality songs.
In the previous 15 years, professional recording studios have used denoising software, which Walther describes as “what we developed”—a fully automatic version.
Tape It is releasing an academic study with a scientific listening test that verifies its findings and demonstrates the software’s ability to compete with leading denoisers on the market.
Deliver Studio-Quality Noise Reduction Via AI:
Deliver Studio-Quality Noise Reduction Via AI [Source of Image: Techcrunch.com]
The company argues in a video that although voice enhancement systems have made considerable advancements, they often only function for speech and mutilate or corrupt music signals. Professional denoising methods, on the other hand, demand expert users to manually handle complicated software. Tape Its technology entails coupling a neural network controller with a noise reduction algorithm based on signal processing.
According to Walther, “Those [professional systems] haven’t been automated because you can’t traditionally put them into a neural network… you can’t train such a system.” We are the first to train such a system. Thus, we are particularly excited about this more extensive area.
He continues that the academic community will be more interested in how they made the denoising product function this way than in the denoising product itself because of its implications for other uses of automating studio software.
Nevertheless, the denoising program already has a few interested potential users, including a significant studio software vendor and a significant hardware maker. In some circumstances, enterprise pricing will be made available, although less expensive plans will be provided for smaller firms.
Of course, other businesses are turning to AI to produce podcasters’ audio like a professional recording studio, so it is not just artists for whom AI technologies are utilized to decrease background noise. As an illustration, Podcastle debuted its Magic Dust AI, a generative AI tool that reduces background noise and broadens its dynamic range.
The group’s five members are based in Berlin, London, Los Angeles, and Stockholm, and they include Christian Crusius, a designer and musician who was formerly with the Accenture-acquired design company Fjord. Most of the denoising software was developed by Christian Steinmetz, a PhD candidate in audio and AI.
Given the velocity of the AI business, the company is currently considering raising financing and is having those discussions, but formal decisions have yet to be taken.
