Feedback Loops

A Feedback Loop track is a special track type where the previous frame’s output is mixed back into the current frame. This creates persistent visual memory - trails, smear, echo, and accumulation effects.

How it works

Each frame is processed in this order:

1. Input (from pre-feedback effects or blank)
       ↓
2. Pre-feedback effects
       ↓
3. Feedback blend  ←── previous frame's output (attenuated)
       ↓
4. Main effects (post-blend, pre-store)
       ↓
5. Store  ───────────► stored as "previous frame" for next tick
       ↓
6. Post-feedback effects
       ↓
7. Output

The Feedback parameter (on the track) controls how much of the previous frame is mixed in. At 0% there is no memory. At 100% nothing decays — the image accumulates forever.

Feedback Effect shaders

Use shaders with a red badge (Feedback Effect kind) inside feedback loop tracks. These shaders are designed to operate on the accumulation buffer - they often include:

• Decay - gradually fade the accumulated image to black
• Zoom / Drift - slowly scale or translate the buffer each frame, creating motion blur
• Color shift - tint or hue-rotate the accumulated image

Caution

Standard Effect shaders (purple badge) work in feedback loop tracks too — place them in the main effects slot (between the feedback blend and the store), so they process the blended result each frame. Pre-feedback effects run before the blend, affecting what is fed into the loop. Post-feedback effects run after the previous frame is stored, so they shape the final output but do not affect what is stored.

Building a feedback patch

A typical setup:

Slot	Content
Pre-feedback	A Layer shader to inject fresh content each frame
Feedback	~70–90%
Main effects	Transformations applied each frame (color, distortion, etc.)
Post-feedback	A Decay feedback effect + optional color/blur

Experiment with low Feedback values and fast decay for subtle trails, or high values with slow decay for dense accumulation.

Tips

• Feedback + displacement is a powerful combination — displacement shaders in a 3D track can also read the feedback buffer as an input texture
• Reduce the Feedback amount to 0% to “flush” the buffer and start fresh
• Map Feedback to an LFO for rhythmic smearing