DSLR-style portrait blur, from any phone photo

Portrait-style background blur, in plain English

DSLR portrait lenses produce the characteristic sharp-subject smooth-backdrop look because of optical depth of field — wide-aperture lenses (f/1.4 to f/2.8) physically can only focus on a thin slice of the scene at the focal distance, with everything in front of and behind that slice falling into progressive blur. Phone cameras have small sensors and small aperture stops that physically can't reproduce this effect optically — which is why phones do it computationally instead. iPhone Portrait Mode and Google Pixel Portrait use a depth-aware two-pass technique: first they generate a depth map (via dual cameras, lidar, or single-camera ML inference), then they apply a synthetic blur to pixels at greater computed depth than the focus subject. The output is visually similar to optical DSLR blur for typical portraits, distinguishably worse for complex scenes (translucent objects, fine hair, cluttered backgrounds), but good enough that 'phone portrait mode' has become the default expectation for headshots and personal photos in 2026.

This editor implements the same computational portrait-blur technique post-hoc, on any photo, regardless of whether the original was shot in portrait mode or whether the source phone supports portrait mode at all. The cutout pipeline runs in-browser ONNX/WASM segmentation to identify the subject (the high-quality cutout we already use for background removal), then applies a Gaussian blur kernel only to the non-subject pixels. The intensity slider scales the blur kernel radius from 0 (no blur) to 100 (maximum dissolve). The edge-falloff treatment — soft transition between sharp subject and blurred background — eliminates the obvious cutout-on-cutout look that gives away naive fake-portrait-blur attempts. Distinct from our /blur-background/ tool which is the 'apply blur after background removal' workflow; this tool is the 'add portrait-style blur to any photo without removing the background' workflow.

Why a portrait-blur tool beats opening Photoshop or paying for a Lightroom plugin

The competitive landscape for portrait-style background blur is uneven and somewhat absurd given how common the use case is. Photoshop ($22.99/mo Photography plan) does it via Camera Raw + masking, but the workflow takes 10-15 minutes per photo and requires moderate Photoshop fluency. Lightroom plugins like Topaz Mask AI ($79.99 perpetual) or Luminar Neo's Portrait Bokeh AI ($99 perpetual) automate it but require a desktop install and a paid license. iPhone and Pixel Portrait Mode work natively but only for photos shot ON a supported device IN portrait mode at capture time — they don't help for older photos, imported photos, non-supported devices, or photos that lost depth metadata in transit (sharing via Messages compresses metadata; AirDropping with image-quality reduction strips it; uploading to most websites also strips it). This editor delivers the same portrait-blur look at $0 with no install, no license, no device requirements, no metadata dependency — works on any photo, any device, any browser. The license documented at /commercial-use-background-remover/ permits use of the result on commercial photography, paid editorial, brand creative, and any other revenue-generating context.