Adobe presented Project Sound Lifta tool powered by artificial intelligence that allows separate a voice recording into different tracks. This will allow us to solve common problems that arise when recording videos, such as background noise or wind interference.
According to Adobe, Project Sound Lift is a one click solution which helps users effortlessly manipulate audio recordings in a variety of scenarios. The application uses artificial intelligence to control and transform speech and sound independently. The user you just need to upload the file, select the item you want to separate and AI will take care of the rest.
The company presented this tool during its Adobe MAX conference in Japan. Adobe made three demos in which Project Sound Lift attacks a different problem. The first of them is a clip from the keynote, where artificial intelligence separates the presenter’s voice, applause and background music with a precision never seen before.
Judging by the demo, Project Sound Lift’s interface is quite intuitive. It is only required upload or drag the sound or video file to the box, wait for it to be analyzed and select the element to separate. Artificial intelligence can separate applause, laughter, background noise, pets, birds or the annoying clicking sound on mechanical keyboards. By choosing it, the separation process is automatically activated and generates an audio track that can be downloaded.
Adobe confirmed that Project Sound Lift integrates Enhance Speech, a filter for Premiere Pro that enhances voice recordings using AI. This technology can eliminate background noise and adjust the person’s voice for greater clarity.
What is Adobe Project Sound Lift for?
The three Project Sound Lift demos show various scenarios where we could separate sounds from a file. Adobe’s artificial intelligence identifies elements such as music, voice, applause, street sounds or people in the background. After download them independentlythe user can bring them into Adobe Premiere, Audacity or other editing software.
One detail to consider is that Project Sound Lift separates only one element per operation. If you want to extract background noise, music, or voice from a video, you will have to analyze and download each one separately. Another element that is not shown in the demo, but can be seen in the videos, is an option called “Final improvement with Project Sound Lift”, which could be a file retouched with Enhance Speech.
The tool is part of Adobe Sneaks, a series of prototypes developed by Adobe engineers to show the scope of the technology. Project Sound Lift will likely be integrated into Adobe Premiere or Audition in the future, along with other Sneaks such as Project Stardust, Project See Through or Project Dub Dub Dub.
“We are entering a new era of creativity, where generative AI is expanding access to powerful new workflows and unleashing our most imaginative ideas. “Adobe Sneaks offer a “sneak peek” of what is to come,” Adobe said.