You can now generate automatic realistic footsteps and cloth Foley for your video clipsKrotos – the Edinburgh-based company specialising in sound design software and effects libraries for film, TV and video games – has expanded its Video to Sound platform, now enabling users to add automatic footsteps and cloth Foley to video clips.
Using AI-assisted visual analysis to interpret movement, pacing and scene context, Krotos is able to place realistic-sounding footsteps and cloth sounds in sync with a video clip – whether it’s a 10-second social media short or 10-minute film. And creative control is preserved; the resultant audio can then be refined in a DAW or NLE.

READ MORE: Fender forays into consumer audio with two Bluetooth speakers and a set of wireless over-ear headphones

Named after pioneer Jack Foley, Foley refers to the art of adding everyday sound effects – like footsteps or clothes rustling, for example – to visual media like films, TV, or video games in order to enhance realism.
It has long entailed painstaking manual work, from recording each necessary sound and lining it up with a video, to making constant timing adjustments as video edits change. But Krotos seeks to offer a massive workflow and time-saving improvement with the addition of automatic footsteps and cloth Foley to its platform.
When applying footsteps to a video clip, users can select an appropriate walking surface – like tile, gravel or wood – and Krotos applies “professionally recorded footstep sounds that match the pacing and action of the scene”. These footsteps can subsequently be exported as MIDI in order to refine timing within Krotos Studio or any other DAW.
Credit: Krotos
The system works the same way in analysing video clips for cloth Foley; it interprets motion and changes in movement, adding context-aware cloth sounds “that reflect subtle shifts, turns, and layered action in a single pass”.
In terms of availability, automatic footsteps and cloth Foley – part of the expanded Video to Sound platform – is available as part of Krotos Studio. Subscribers across all tiers of Krotos Studio have access to the feature, while Krotos Studio Max users can upload videos up to 10 minutes in length, and process up to 200 videos per month.
Krotos also makes the point that user-uploaded videos “are not used for AI training purposes”. 
We spoke to Krotos last year, when the company stressed the importance that “creative control remains firmly in the hands of the user”.
“When we say our AI engines are ‘non-generative’, we mean they are not creating entirely new sounds from scratch. Instead, they are designed to assist sound designers by processing and enhancing existing high quality audio assets that have been recorded by professionals,” Krotos told MusicTech.
“…The quality of the sound is always high unlike the results of the existing generative models. Our commitment to ethics is reflected in how we train our AI. We exclusively use professionally recorded assets that we own or have full rights to… The goal is to empower users, not replace them.”
Learn more about Video to Sound and Krotos Studio at Krotos.
The post You can now generate automatic realistic footsteps and cloth Foley for your video clips appeared first on MusicTech.