Back to Threads

Fusion Merge inputs explained: background, foreground and mask

Q&A Posted by Clipizo • Oct 3, 2026

A standard 2D Merge places its foreground over its background. Its blue mask input limits that operation; moving nodes around the graph does not change the connections.

What the three inputs do

  • Background, shown as yellow/orange: the image underneath. This input also establishes the standard Merge's output image dimensions.
  • Foreground, shown as green: the image placed on top, using its alpha channel and the Merge's compositing settings.
  • Effect Mask, shown as blue: an optional mask that limits where the Merge affects the background.

Use the input label or hover tooltip as well as the color. Inputs can change their position around a node as you rearrange the graph. The leftmost pipe is not automatically the background.

Example: put a title over a video

  1. Open a video clip on the Fusion page. Its MediaIn supplies the clip image, and MediaOut sends the result back to the timeline.
  2. Add a Text+ node and enter a short title.
  3. Add a Merge. Connect MediaIn to its Background input and Text+ to its Foreground input.
  4. Connect the Merge's output to MediaOut.
  5. View MediaOut so you are checking the final composition rather than just the Text+ image.

The image path is MediaIn → Merge → MediaOut, with Text+ → Merge Foreground. A normal Text+ image has transparency around its letters, so the video remains visible there.

If the order is wrong, the opaque video can cover the title. With the relevant Merge selected, Ctrl+T on Windows or Command+T on macOS swaps its foreground and background connections. You can also reconnect the pipes manually.

Example: limit where the title is visible

Add a Rectangle mask and connect it to the Merge's blue Effect Mask input. Adjust the rectangle over the part of the composition where you want the title to appear.

In an ordinary non-inverted mask, white permits the Merge operation, black prevents it, and gray allows a partial effect. Masking this Merge limits the title's contribution while leaving the background video visible. It does not automatically cut a transparent hole through the entire finished composition.

That distinction matters when building reveals: masking the Text+ node, masking the Merge, and masking a later node act at different points in the image chain.

Why is the result still wrong?

  • Nothing changes in the viewer: make sure you are viewing the Merge or MediaOut, not an earlier node.
  • The video disappears: an opaque foreground may be covering it. Inspect the foreground's alpha and the input order.
  • The title is missing: check its frame range, the Merge's Blend, and any connected mask.
  • The image has an unexpected size: inspect the Background input's dimensions and frame-format settings.
  • The Fusion viewer looks right but the Edit page does not: check that the final image is connected to the intended MediaOut.

A black viewer background alone is not proof that an image is opaque. Inspect the alpha channel or composite the image over a contrasting background to check its transparency.

Continue with the same setup

The Text+ Write On guide adds a typewriter reveal to the title. The slide-and-hold animation guide explains how to move it without moving the background video.

Reference: Blackmagic Design's Beginner's Guide to DaVinci Resolve 20, Lesson 9, especially the Merge discussion on printed pages 509–510. This article concerns the standard 2D Merge, not Merge3D or MultiMerge.

Discussion

No replies yet. Be the first to jump in.

You must be logged in to participate in this discussion.

Login to Reply