Disclosure: DPReview attended Adobe Max, and Adobe covered travel and lodging expenses.
Adobe used its Max conference to bring more AI into the Premiere timeline. Color Mode, a premiere color mode for grading inside the timeline, is becoming generally available, Generative Remove is entering beta, and an AI assistant can now carry out editing tasks. Several of these features first appeared in beta, while others remain previews.
The premiere color mode leaves beta
The premiere color mode becomes generally available after its public beta launched in April. It gives Premiere a dedicated grading workspace with a simplified interface, separate from the standard timeline view. Adobe designed premiere color mode for editors who want to grade without the skills of a professional colorist.
PetaPixel reports that premiere color mode replaces the Lumetri Color panel, which Adobe has used for grading for several years. The publication says the new system appears to be a direct response to DaVinci Resolve. Adobe describes the controls as bi-directional, with mouse movement up and down adjusting one value and left and right adjusting another.
Adobe says the Style Presets are more flexible than a LUT. They are editable, stackable collections of adjustments, so editors can change individual aspects of a look instead of accepting a fixed transformation. Adobe also says, “No more copying and pasting attributes. Actually understand how color is flowing through your timeline.”
Presets let editors judge how a look affects the whole sequence at a glance. Editors can then select single clips or groups that need extra adjustment. Adobe’s approach suits edits where most shots match but a few need attention.
Generation moves onto the timeline
Adobe is expanding the Generative Media Tool, which creates assets inside Premiere rather than in outside services. Editors can generate video, sound effects, soundscapes and music. During the briefing, Adobe also demonstrated image generation.
PetaPixel says the models behind the tool can include Adobe Firefly, Google, Kling, Runway and Luma.
Music and soundscape generation are not new. Adobe introduced both in beta in September. Generate Soundscape analyzes footage to create accompanying audio, while Generate Music offers controls for shaping a track to fit a project. In the briefing, the soundscape tool produced four separate tracks that editors can adjust individually.
More control over audio
Premiere’s audio updates include improved Enhance Audio, Separate Crosstalk and Separate Sound Source. Adobe says these tools clean up recordings and separate their components. Source 2 notes that Premiere can now split audio sources into stems that can be edited one at a time.
In the keynote, Adobe demonstrated a clip of a speaker on a boat, with a loud engine and music in the background. Enhance Audio reduced the background noise so the speaker was far easier to hear. The feature may help most with field recordings, where dialogue and ambient sound arrive together.
Separate Crosstalk handles recordings where people talk over each other. In Adobe’s example, an interview subject began speaking before the interviewer finished a question. The tool placed each speaker on a separate track, so the editor could mute the interviewer and keep the subject clear.
Adobe announced expanded audio separation in September. At last year’s Max Sneaks, the company showed Project Clean Take, an experimental tool that split a recording into voice, background sounds and music.
If Premiere’s audio separation tools have you thinking about capture hardware, RodeCaster’s 4K video update and expanded connectivity shows how production gear is changing too.
An AI assistant that acts
Adobe is expanding the AI Assistant in Premiere, a conversational tool powered by its Creative Agent. Instead of only explaining a task, the assistant performs editing work in response to natural-language requests. It first entered public beta in June and remains in beta.
Adobe highlights several new abilities. The assistant can apply effects across every clip in a sequence, track masked subjects, and adjust color and exposure across an entire sequence. In the briefing, Adobe also showed it synchronizing cameras and audio for multicam clips and flagging shots that may have been missed during color grading.
Editors can save repeatable instructions as custom “skills” and run them with a slash command. Adobe describes this as recording how an editor prefers a task handled, so the same setup does not need to be explained on every project.
Generative Remove enters beta
Generative Remove lets editors delete unwanted objects from footage and fill the gap with generated detail. Adobe says the tool uses automated masks to identify the object, then removes it nondestructively.
Adobe’s example involved a surfer interview in which a crew member walked into the frame. Presenter Dacia Saenz used automatic masking to select the crew member, then applied Generative Remove to take them out of the entire clip. Adobe’s examples of typical targets include boom microphones, logos and background distractions.
The demo looked convincing, but the keynote did not test complex movement or busy backgrounds. Those conditions will determine how useful the tool is on real projects. Editors should test it on their own footage before relying on it.
Premiere Mobile and licensed music
Premiere Mobile is getting the Creative Agent in private beta. The agent can turn footage into editable first cuts, accept changes written in natural language, and let the editor refine the result. Adobe says agentic editing will come to Premiere Mobile “soon.”
Adobe is also adding hundreds of thousands of licensed music tracks from Universal Music Group and Warner Music Group to adobe premiere on iPhone. Android support is coming later. These tracks are licensed for personal, non-commercial videos. Adobe says, “For projects that require commercially licensed music, you can continue to access Adobe Stock Audio.”
Licensing questions around AI-assisted content reach well beyond background music, and AI-generated characters and the copyright fight explores how ownership of machine-made work is being contested.
Third-party agents and project files
Adobe is previewing local Model Context Protocol, or MCP, integrations. The integrations let third-party agents work directly in native Premiere, Photoshop, Illustrator and After Effects projects. An MCP is a standard that lets an AI agent connect to and act within external software.
Bins and sequences stay editable. An agent can make changes inside the project, and the editor can inspect or revise them before export. The integrations are a preview and not yet generally available.
What editors will test first
Most of these tools are still in beta or preview, and Adobe has not published timelines for every rollout. The premiere color mode is the clearest case of a feature moving into everyday use. The rest will depend on how they handle real footage.
The real test is how much of the edit editors are willing to hand to the software. Adobe’s previews show where the Premiere timeline is heading. Working editors will decide how far it goes.
H/T dpreview.com
Frequently asked questions
What is Premiere’s new Color Mode and who is it for?
Color Mode is a dedicated grading workspace in Premiere with a simplified interface, separate from the standard timeline view. Adobe designed it for editors who want to grade without the skills of a professional colorist.
Which AI models power Premiere’s Generative Media Tool?
According to PetaPixel, the models behind the Generative Media Tool can include Adobe Firefly, Google, Kling, Runway and Luma. The tool creates assets inside Premiere rather than in outside services.
What can Premiere’s AI Assistant do for editors?
The AI Assistant performs editing work in response to natural-language requests rather than only explaining tasks. It can apply effects across every clip in a sequence, track masked subjects, and adjust color and exposure across an entire sequence.
What does Separate Crosstalk do in Premiere?
Separate Crosstalk handles recordings where people talk over each other. It places each speaker on a separate track, so an editor can mute one person, such as an interviewer, and keep the other clear.
More in
Technology
All Technology
More like this
Four underrated web tools every video creator should know
4 Best Royalty Free Music Sites of 2021
Adobe's Acquisition of Topaz Labs Signals Big Changes for AI-Powered Creative Tools
How to Use Promo Videos to Grow Your Creative Business
What’s royalty free music and why content creators should use it
