Depth of Field and Bokeh: Prompting Focus and Background Blur

Depth of field (DOF) is the range of distance within a scene that appears acceptably sharp in the final image. When that range is narrow — only a few centimeters in front of and behind the focused subject — we call it shallow depth of field. When the range is wide and nearly the entire scene is sharp, it is deep depth of field.

Three optical variables control DOF in real cameras: aperture (the size of the lens opening, expressed as f-stop), focal length (how much the lens magnifies the scene), and subject-to-camera distance. A wide aperture (low f-number like f/1.4 or f/1.8) produces shallow DOF. A long focal length (85mm, 135mm, 200mm) compresses the scene and separates subject from background. Moving the camera closer to the subject also reduces DOF.

AI models have absorbed millions of tagged photographs and understand these relationships implicitly. When you write "shallow depth of field, f/1.8, 85mm portrait lens" in a Floniks AI Image prompt, you are invoking a well-understood optical signature — the model knows to render a sharp subject against a smoothly blurred background, replicating the look of a professional portrait lens at wide aperture. You do not need to explain the physics; you just need the right vocabulary.

Bokeh — from the Japanese word for blur or haze — describes the visual character of out-of-focus areas rather than merely the fact of blur. Not all blur is equal. A lens with well-corrected spherical aberration produces smooth, round highlight discs often described as "creamy" or "buttery." A lens with specific optical design (like the Sony G-Master or Canon L series primes) produces almost perfectly circular bokeh. A budget kit lens may produce hexagonal or harsh-edged discs. In AI prompts, you can specify the quality directly.

Prompt phrases by bokeh character:

• "creamy bokeh, smooth background blur, 85mm f/1.4" — the classic portrait look
• "swirly bokeh, vintage 58mm f/1.4 Helios lens rendering" — the characteristic swirl pattern of Soviet-era lenses
• "nervous bokeh, harsh double-edge highlights" — consciously ugly, can feel gritty and analog
• "circular bokeh balls, neon light reflections in background, f/1.2" — cyberpunk / nightclub aesthetic
• "starburst bokeh, f/22, city lights at night" — achieved by small apertures, creates spike patterns on point light sources

In Floniks AI Image, pairing bokeh descriptors with the correct focal length and lighting context produces highly consistent results. Adding "out-of-focus background" or "background separation" as supplementary phrases reinforces the intent when the primary bokeh term alone is not sufficient.

Shallow DOF is the dominant choice for portrait photography, product close-ups, and emotional drama. It physically removes distracting background information and forces the eye to the sharp subject. In AI generation it also tends to produce more polished, professional-looking results because the background fill becomes a painterly wash rather than a competing scene element.

Prompt for shallow DOF: "medium close-up portrait, woman in botanical garden, shallow depth of field, 105mm, f/1.8, soft afternoon light, background flowers out of focus".

Deep DOF is essential for landscapes, architectural photography, group shots, and any scene where the environment is as narratively important as the subject. It communicates abundance of information — the viewer's eye can roam across a sharp, rich scene.

Prompt for deep DOF: "wide-angle landscape, mountain valley with wildflowers, deep depth of field, 24mm, f/11, golden hour, every blade of grass in sharp focus".

When to be explicit vs. let the model infer: Floniks AI Video and AI Image models will infer reasonable DOF from the shot type (close-up implies shallow; wide shot implies deep). Being explicit is most important when your intention contradicts the default — for example, a close-up product shot with everything in focus for catalogue clarity ("macro product shot, all-in-focus, f/16, infinity focus, flat lighting"), or a landscape with deliberate tilt-shift miniaturization effect ("aerial view, tilt-shift lens, shallow depth of field, toy city effect").

A rack focus (also called focus pull) is a deliberate shift of the focal plane during a shot — the camera starts focused on a foreground subject, then smoothly transitions focus to a background subject (or vice versa), revealing or concealing narrative information dynamically. It is one of the most cinematic in-camera transitions available.

In Floniks AI Video, you can request focus racks in the motion prompt: "rack focus from foreground coffee cup to couple in background, shallow depth of field, warm café light, 85mm". Not every model executes this with perfect precision, but the phrase meaningfully biases the motion toward focus transitions rather than camera movement.

A soft focus pull to black (racking focus until the subject blurs completely) works well as an emotional or scene-closing transition: "slow rack focus pull, subject blurs into soft bokeh, fade to black, cinematic, 85mm".

For multi-shot sequences in the Floniks Editor workflow, consider building a rack-focus opening node (background in focus, subject blurred) that then cuts to a sharp-subject node — this mimics the suspense device used in thriller cinematography where the threat arrives into focus.

Product and food photography have their own DOF conventions. Hero product shots typically use a very shallow DOF to isolate the product against a complementary background — the product is razor-sharp while the environment behind it dissolves into complementary color wash. Prompt: "product close-up, luxury skincare serum bottle, shallow depth of field, 100mm macro, white marble surface, background soft peach bokeh, studio lighting".

Flat-lay food photography flips the convention: the camera shoots straight down at a table arrangement, and the shallow focus plane runs horizontally across selected ingredients. Prompt: "flat lay food photography, avocado toast, overhead shot, selective focus on center toast, f/2.8, 50mm, natural window light, out-of-focus herbs around edges".

Beverage photography often exploits foreground bokeh — ice cubes or condensation beads placed between camera and subject create an intimate, immersive quality. Prompt: "beverage product shot, craft beer glass, foreground water droplets in soft focus, background bar lights as bokeh, 85mm, golden rim light, dark moody background".

These formulas map directly to Floniks AI Image prompts and to product-catalog workflows in the Floniks Editor, where you can batch-apply the same DOF signature across multiple product variants.

Use these ready-made phrase blocks to add directly to any Floniks prompt:

Portrait, maximum separation: "shallow depth of field, 85mm, f/1.4, subject sharp, creamy background bokeh"

Product isolation: "f/2.8, 100mm macro, subject tack-sharp, background dissolved into color"

Cinematic drama: "razor-thin focus plane, 135mm, wide aperture, background lights as circular bokeh balls"

Landscape, everything sharp: "deep depth of field, 24mm, f/11, hyperfocal distance, every element crisp"

Night portrait, bokeh street lights: "portrait, city night, neon reflections as bokeh discs in background, 50mm, f/1.2, ISO mood"

Tilt-shift miniature: "aerial cityscape, tilt-shift effect, selective narrow focus band, toy-like scale illusion"

Macro, near-total blur surround: "extreme macro close-up, single dewdrop on leaf, surrounding leaf texture blurred, 100mm macro, f/2.8"

Stack these phrases with shot type, lighting, and subject description for a complete, high-fidelity prompt architecture. In the Floniks Editor, save them as reusable node templates to apply consistently across a production batch.