How to Use AI Face Age Transformer in DaVinci Resolve

Face filters have been around for a long time but are usually limited to apps like Snapchat, TikTok and Instagram, however Davinci Resolve has recently added a brand new feature called "AI Face Age Transformer". It's a generative effect that lets you change the visual age of a person in a video, making them look younger or older while preserving facial structure and motion tracking. Since it's a new feature you probably don't know where to find, we'll guide you through everything.

How to Use AI Face Age Transformer in DaVinci Resolve

Before you do anything else, you'll need to make sure that you have Davinci Resolve 21 or newer installed. If you are using an older version you will not have this feature.

• First go to the Color page
• Open the Effects Library
• Search for Face Age Transformer
• Drag it onto your node

Once applied, the effect becomes part of your node tree and can be adjusted in real time. Just like any of the other effects.

After applying the effect you'll want to do a couple more things in order for it to work properly.

• Run face detection on the frame

• The software will identify the subject and draw a tracking box
• Track the face through the clip

This step will make sure that the age transformation stays locked onto the subject as they move. If you don't do this it will not work.

Adjusting the Age Transformation (Younger or Older)

The effect requires you to define the subject’s real age:

• Enter the estimated actual age of the person in the shot and this will give the model a reference point for realistic transformation. However, instead of directly setting a new age, you can also adjust an offset:
• Move the slider to make the subject appear younger
• Or increase it to make the subject appear older

Lower values produce smoother, younger looking skin with softened features. Higher values add wrinkles, skin texture changes, and other aging details.

Prepare for a BIG Performance Hit!

This effect is computationally heavy, especially in high resolution projects. So performance depends on your system and timeline settings. In 4K projects, playback may not be real time even on powerful hardware. So if you are struggling you can try a couple of basic performance boosting changes.

• Lower playback resolution
• Disable overlays during playback
• Allow time for AI processing to complete tracking

You can find more performance boosting steps here.

Once it has fully processed, the effect updates across the tracked face throughout the clip and the transformation can look surprisingly realistic, if the recorded footage was good quality with good lighting. Otherwise it can look ridiculously out of place. At this point in time the feature is a bit of a fun concept rather than something serious.