At Adobe's annual MAX conference, the company introduced Project Super Sonic, a prototype that leverages AI to generate sound effects for video content. While creating compelling visuals is important in video production, the role of audio is equally crucial. Finding or creating suitable sound effects, however, can be a time-consuming task. Adobe’s new tool aims to simplify this process by integrating AI technologies, allowing users to generate sound effects quickly and easily.
Project Super Sonic allows users to create audio effects using a text-to-audio feature. While generating sounds from text prompts is a fun and useful feature, similar technologies are already offered by companies like ElevenLabs, making this aspect of the tool less revolutionary.
What sets Adobe’s project apart is the addition of two innovative modes. The first mode uses object recognition models, enabling users to click on any part of a video frame. The system then generates a sound based on that object. This approach combines AI models to streamline the sound creation workflow.
The most impressive feature, however, is the third mode, where users can record themselves imitating the desired sound. Project Super Sonic analyzes the recording and automatically generates the appropriate sound effect, matching it to the video’s timing. This feature allows for more expressive control over the audio design.
According to Justin Salamon, head of Sound Design AI at Adobe, the team initially developed a text-to-audio model but wanted to give users more creative control. The goal was to create a tool not just for casual users but also for professional sound designers. The vocal control feature, which analyzes the characteristics of a user’s voice, offers precise control over the generated audio’s energy and timing.
Though Project Super Sonic remains a demo, there is optimism that it could eventually become a part of Adobe’s product suite. The same team behind this project also contributed to the audio portion of Generative Extend, a feature of Adobe’s Firefly AI model, which extends short video clips by a few seconds along with their audio tracks.
As with many experimental features showcased at Adobe MAX, there’s no guarantee that Project Super Sonic will make it into Adobe Premiere or other tools. However, its potential to streamline the audio creation process makes it a promising addition to the company's AI-driven offerings.
