IBC2026
Atomos (Booth: 7.B19) announced the integration of its cloud platform “ATOMOSphere” with “PlanckOne FoleyAI,” developed by Mocha Chai Laboratories. This will allow video producers and sound engineers to send video stored in the ATOMOSphere workspace to PlanckOne FoleyAI and generate Foley material synchronized with the video.
PlanckOne FoleyAI can analyze footage ranging from single shots and scenes to entire edited works. It identifies actions, materials, and environments within the frame and generates Foley sounds synchronized with the footage, such as footsteps, rustling clothes, and impact sounds.
ATOMOSphere is a cloud-based production ecosystem designed for video professionals, photographers, and content creators, supporting media uploading, storage, editing, sharing, reviewing, and collaboration.
This integration allows users to send videos directly from the ATOMOSphere workspace to PlanckOne FoleyAI for analysis. This process does not require audio stems (separated audio material) or reference sounds, and uses only the visual information contained in the video.
PlanckOne FoleyAI can generate a starting point for Foley footage synchronized with video in just a few hours. This significantly reduces the process, which previously sometimes required weeks of preparation and securing a Foley stage. The generated audio is provided as separate, individually editable layers for sounds such as footsteps, clothing rustling, and impacts.
Peter Barber, CEO of Atomos, commented as follows:
“Foley is an essential element in creating an authentic and immersive soundtrack, but it’s also one of the most time-consuming processes in post-production.”
“Integrating PlanckOne FoleyAI with ATOMOSphere allows filmmakers and sound designers to obtain the foundational material for Foley production early in the production process. At the same time, the creative final decision-making power remains with the producers building the mix.”
Chai Yiwei, CEO of Mocha Chai Laboratories, commented as follows:
“Foley has always been a craft built on time, patience, and in a room full of props, and this partnership will not change its essence.”
“Making PlanckOne FoleyAI available from within ATOMOSphere allows users to utilize the tool directly from the environment where their video footage is stored. Existing ATOMOSphere users can try out the version that is scheduled for official release in their projects without installing new software or moving files. The feedback received will guide future development.”
Even after generation, sound designers can adjust each Foley element individually. They can change the timing, weight, and texture of each layer without affecting other parts of the mix, allowing for fine-tuning to match character movements or the creative direction of the work.
Atomos states PlanckOne FoleyAI is built using open-source large-scale language models (LLMs) and trained using sound libraries and real-world post-production data owned by Mocha Chai Laboratories. It is designed to integrate directly into existing editing pipelines, allowing production teams to adopt it without significantly altering their current workflows.
Lee Jun-Han, Technical Lead at Mocha Chai Laboratories, commented as follows:
“With PlanckOne FoleyAI, you can easily add Foley effects during the offline editing stage, giving texture to offline cuts that would normally sound monotonous.”
“This collaboration will allow ATOMOSphere to directly connect cloud-based video footage with AI-powered audio post-production, in addition to media storage, review, and collaboration. Editing and sound teams will be able to move more quickly from video footage to create editable Foley sessions, and obtain audio material that can be used for subsequent creative work.”
Key Features of the PlanckOne FoleyAI Integration to ATOMOSphere
- Send videos directly from the ATOMOSphere workspace to PlanckOne FoleyAI for analysis.
- Generates sound effects such as footsteps, rustling clothes, and impact sounds synchronized with the video.
- Audio stems and reference sounds are not required.
- Generate an editable Foley starting point within hours.
- Individual audio layers can be adjusted independently.
- Can be integrated into existing editing pipelines.
PlanckOne FoleyAI is currently in beta. As such, its integration with ATOMOSphere is a technical preview. Plans are underway to offer a native version and a subscription service in the future.
