Automatic Foley: Krotos Video-to-Sound now also with footsteps and cloth sounds
[18:09 Mon,26.January 2026 by blip]
Since last autumn, Krotos—specializing in sound design software and effect libraries—has been offering the Video-to-Sound feature, which is intended to automatically and precisely add Foley sounds to a video. Initially, only ambient sounds were supported, but now footsteps and cloth sounds have been added. However, the entire feature is still in beta.
Krotos Video-to-Sound
The Video-to-Sound feature is AI-based, but according to Krotos, generative AI is not used—instead, the clips are visually analyzed by AI algorithms. In a second step, genuinely recorded sounds from the library are then adapted to movements, tempo, and scene context and added to the images.
This is intended to work for short social media clips as well as 10-minute films, although the quality also depends on the respective video material. Users retain creative control, as the resulting audio material can subsequently be further processed in a DAW or an NLE.
If, for example, footsteps are needed, a suitable surface can first be selected, such as tiles, gravel, wood, etc., whereupon Krotos professionally recorded footsteps are adapted accordingly. Prerequisites for this are visible human figures in the frame with clear posture and movement, full or partial visibility of the body, moderate speed of movement, and well-lit subjects.
The footsteps can then be exported as MIDI to refine the timing within Krotos Studio or another DAW. A similar process is intended for rustling clothing.
The sound design can be used commercially without concern, as only specially created and licensed sound snippets are used. However, one should not forget that genuine Foley sound design is an art in itself—see also our report on sound design and Foley.
The Video-to-Sound platform is browser-based and supposedly available for subscribers of all tiers of Krotos Studio; however, we actually only see the feature listed for the Pro and MAX subscriptions at 23.33 or 83.33 Euros, with 100 videos of 5 minutes or 200 videos of 10 minutes, respectively. New users receive one test generation for free.
