# Shapes Module Once you know your values, the Shapes module asks: do those values actually read as clean, simple shapes? This module is about silhouette and mass, not detail. Every viewer in this module also has an **Enable Image Caching** switch. It's a technical convenience, not a creative choice — when on, Notana remembers the last processed result so flipping settings back and forth is instant. Leave it on unless you're short on memory. ## Simple Notan The classic two-shape (light/dark) silhouette study, stripped of everything but shape. - **Black-White Balance** — moves the line between "counts as light" and "counts as dark." Push it one way to grow the dark shapes, the other way to grow the light ones. - **Squint (Blur)** — softens the photo before splitting it into shapes, simulating the effect of squinting at your subject so only the big masses survive. - **Draw Outline** — adds a thin outline around each shape, useful when the light/dark shapes butt up against similar-toned neighbours. ## Block-In Breaks the photo into the big rough shapes an artist would block in first, before any refinement. ![Block-In applied to the reference photo at basic settings](examples/shapes-block-in.jpg) *Block-In at its default block size — the scene reads as a handful of organic, connected masses instead of fine detail.* - **Layout Pattern** — the underlying grid the blocks are distributed on: square grid, honeycomb, brick, diamond, radial rings, an organic "Voronoi" pattern, strips, or a random mix. Voronoi tends to feel the most natural; grid patterns feel more graphic and deliberate. - **Block Shape** — the shape of each individual block — fill (organic, follows the Voronoi regions), square, rectangle, circle, triangle, star, blob, and more. - **Block Size** — how large each shape mass is. Larger blocks give simpler, broader forms; smaller blocks keep more of the original detail. ## Value Bands Groups shapes by value step, similar to the Value module but focused on the resulting shape outlines rather than the tones themselves. - **Number of Bands** — how many distinct value bands to split the image into. - **Squint (Blur)** — smooths the transitions between bands so the result reads as shapes rather than a stepped gradient. - **Draw Outline** — traces the boundary between each band. ## Value Masses Merges neighbouring same-value shapes into single connected masses, so you can see if your composition reads as a few big masses or a scatter of small, disconnected ones. - **Edge Sensitivity** — lower reveals more of the original detail as separate shapes; higher merges everything down to only the major transitions. ## Light Basins Highlights the connected pools of light across the image, useful for checking whether your light areas form one confident shape or several competing ones. - **Detail Level** — lower captures more of the fine detail in how light pools together; higher shows only the broadest light masses. ## Edge-Bounded Builds shapes from where edges naturally close off, closer to how the eye actually groups a scene into objects. - **Edge Sensitivity** — lower picks up more edges (more, smaller shapes); higher only registers the strongest boundaries. - **Simplification** — the smallest a region can be before it gets folded into a neighbour — higher gives simpler shapes. - **Draw Outline** — traces the boundary around each detected shape. ## Shape Simplifier A general-purpose control for pushing a photo toward simpler, blockier shapes or backing off toward its real complexity. - **Edge Smoothing** — smooths uniform areas while keeping real edges sharp, like squinting at your subject. - **Value Steps** — how many distinct light/dark notes the shapes are built from; fewer steps means simpler shapes. ## How to use it Ask "does this still look like the subject if I squint?" If a shape breakdown falls apart or turns into confetti, ease the simplification back. If it still looks too busy even at the simplest setting, it's often a sign the original photo's composition doesn't have a strong shape story to begin with — worth checking in the Composition module.