AutoCrop helps you reframe your sequence for social outputs without manual keyframing. It analyzes your work area, identifies the active speaker, and applies smart reframing keyframes back onto your Premiere timeline.
Access AutoCrop
- In Adobe Premiere Pro, open
Window>Extensions. - Launch
Brevidy AutoCrop. - In your sequence, set the section you want to process using
IandOpoints.
What AutoCrop Does
AutoCrop currently runs in Active Speaker mode and is designed for interview, podcast, talking-head, and dialogue-driven footage.
It can:
- Reframe your sequence for social aspect ratios.
- Follow the active speaker instead of staying on a fixed crop.
- Control horizontal and vertical crop movement separately.
- Adjust how quickly it switches between subjects.
- Zoom in more or less depending on how tightly you want the subject framed.
- Apply the final reframing back to your timeline automatically.
Output Options
You can choose from these output formats:
9:1616:91:14:5
Each format includes resolution options such as 720p, 1080p, and 4K where supported.
AutoCrop Controls
Choose these settings before running AutoCrop:
Output format: choose your target aspect ratio and resolution.Subject Switching:Stablefor fewer switches,Balancedfor most footage, andResponsivefor faster speaker changes.Motion Tracking: chooseTracked,Dynamic, orStaticseparately for horizontal and vertical movement.Zoom: chooseNo Zoomfor the widest framing,Standardfor moderate zoom when helpful, orClosefor tighter framing.
After that, click Run AutoCrop.
How It Works
When you run AutoCrop, Brevidy:
- Renders a proxy of the selected work area.
- Securely uploads it for processing.
- Analyzes scenes, speakers, and framing.
- Builds crop keyframes for the selected output format.
- Applies those reframes back into Premiere.
AutoCrop does not render a final delivery file for you. It applies the reframing directly to the sequence so you can keep editing.
Current Limits
- AutoCrop uses your
IandOpoints, not the entire sequence by default. - AutoCrop currently runs on work areas up to
8 minutes. - Extremely short ranges can fail preflight checks.
- It currently works best on footage with clearly visible people and speaker changes.
- The current production workflow is focused on
Active Speakerreframing.
For best results, use clear interview, talking-head, podcast, or dialogue-driven footage.