Adobe starts New generative AI movie tools which provide amusing ways to create sound effects and control generated video results. In addition to known text hints, which usually allow you to describe what AI Firefly Adobe models should make or edit, users can now exploit voice recordings similar to onomatopoeia to generate custom sounds, and exploit reference materials to lead movements in films generated by Firefly.
The Generate Sound tool tool that starts in the beta version in the Firefly application It can be used with a recorded and generated material and provides greater control over sound generation than Google Veo 3 video tool. The interface resembles a video editing schedule and allows users to match the effects that create on time with the recorded recording. For example, users can play a video of a horse walking along the road and at the same time record the sounds of “Clip Clop” in time with hooves, along with a description of the text with the inscription “Kopyta on concrete”. Then the tool will generate four sound effect options to choose from.
This is based on the Project Super Sonic experiment, which Adobe showed Max at his event in October. It does not work on speech, but supports the creation of influence sounds like branches, steps, lock effects and others, as well as atmospheric sounds such as the sounds of nature and the atmosphere of the city.
Modern advanced controls also appear to the Firefly Text-to-Video generator. The reference to the composition allows users to send a video along with text monitor to reflect the composition of this material in the generated film, which should facilitate the achievement of specific results compared to the multiple introduction of the descriptions of the text. Key cultivations will allow users to whip and send images of the first and last frames, which Firefly can exploit to generate video between and recent preliminary settings, and recent stylish settings ensure the selection of visual styles that users can quickly choose, including anime, vector art, clay and many others.
These preliminary settings are only available with your own AI Firefly model from Adobe. The results leave something to be desired, if the live demo I saw was any clue – the “Clayation” option looked like 3D animation from the early 2000s. But Adobe still adds support for competitive AI models in his own tools, and the generative manager Ai Adobe Alexandru Costin said The Verge That similar control and pre -preliminary elements can be available for exploit with AI models of another company in the future. This suggests that Adobe is fighting to maintain his place at the top of the original FoodCain software when AI tools are growing popularity, even if he lasts behind OpenAI and Google in the generative models themselves.