# Computational smartphone look (deep-focus HDR)

> The deep-focus HDR strand of phone photography: a processed wide view keeps foreground and background legible while retaining bright sky, windows or pale surfaces alongside shaded detail. Burst fusion tends to suppress noise, but does not guarantee a noiseless finish. This is a rendering choice, not every phone picture, Night mode result or synthetic-defocus portrait; it is not tied to one subject, pose or lighting rig.

- Period: 2014–present
- Where it began: A networked processing idiom rather than an organised movement
- Kind: Digital & social
- Also called: Computational HDR phone look, HDR+ look, Smart HDR look, iPhone look (processed HDR sense)
- Page: https://lumogeny.com/styles/smartphone-computational-look/

## How the look is made

- Use a phone's processed HDR capture rather than an unprocessed single-frame file; select a result with visibly retained bright-area detail, not merely an HDR-labelled mode.
- Include a bright window, sky or pale surface alongside shaded detail, retaining both where possible without requiring every shadow to become bright or every lamp core to retain texture.
- Keep the wide main-camera view sharp through depth, with no synthetic portrait blur; favour noise suppression over added grain, but reject neither real texture nor a little remaining noise.
- For dim scenes, use the available lamps without flash and hold the phone steady throughout the multi-frame capture; do not confuse the capture duration with one long exposure.
- For people, retain the surrounding scene in focus and obtain consent rather than treating synthetic defocus or face reshaping as part of this look.
- Check moving edges, hair and fine texture for merging errors, and choose a clear frame rather than treating every processing artefact as intentional.

## How it is lit

- Available daylight through a bright window with a shaded room or subject in front; computational exposure balancing opens darker detail rather than a physical fill light changing the scene.
- Existing street lamps, shop lights or household bulbs at night, with no flash; retain pale-surface detail and some warm-cool source separation while multi-frame capture reduces noise. A silhouette-heavy night result alone does not establish this HDR look.
- Outdoor daylight with a bright sky behind consenting people or terrain; keep sky structure as well as readable shaded faces and ground, without synthetic background defocus.

### A light plan from its choices (schematic)

- Key: a window, about 60° to the camera’s left, soft.

## Camera settings

24–28 mm · f/1.8–f/2 · 1/1000–0.3 s · ISO 50–1600 (on phone), inferred. Illustrative wide-main-camera starting envelope, not a measured population or settings recovered for these specimens. Physical phone f-numbers are not full-frame depth equivalents; exposure and ISO are automatically chosen. Documented handheld night-mode constituent frames can reach 333 ms, with longer supported captures; the total burst lasts longer than any single frame. This entry selects deep-focus output and excludes synthetic portrait defocus.

## Directing people

There is no defining body pose: many views contain no figures; in everyday scenes people stand loosely or concentrate on a task. Let consenting sitters choose a comfortable stance and keep their real facial proportions; leave the setting legible rather than prescribing a particular face, expression or beauty ideal or blurring it synthetically.

## Styling

Leave rooms, streets, clothes and objects as found, with lived-in surfaces and ordinary dress; elaborate sets, coordinated fashion and make-up are not required. Keep texture in fabric and walls rather than artificially perfecting the scene. The palette is illustrative of mixed-light examples, not a mandatory brand colour scheme.

## Before you shoot it

Phone snapshots can expose faces, private homes, locations and routines: obtain consent where appropriate, avoid publishing identifying location metadata, and do not infer permission from a photograph's presence online. The named contemporary photographers are evidence, not signatures to copy; this entry describes shared processing choices. Do not alter a sitter's complexion or anatomy without informed agreement. Multi-frame fusion and synthetic blur may change edges or movement, so preserve originals when an image is used as evidence. Never pass a generated image off as a real street encounter, news event, family record or photograph of an actual private dwelling; any illustration must be labelled and use invented, non-identifiable subjects. The chapel specimen is evidence about exposure rendering, not an invitation to reproduce sacred imagery or use a religious interior without permission.

## A prompt for an image model

No prompt for this style. Behind its pictures: everyday phone photographs record real people, homes and routines. A generated imitation could pass as a personal snapshot or documentary evidence of an actual encounter or place.

## Photographs

- Marc Levoy, Princeton University Chapel, HDR+ on, published 2014. Google Research. No file-specific open licence confirmed; in copyright, link only. In copyright, linked: https://research.google/blog/hdr-low-light-and-high-dynamic-range-photography-in-the-google-camera-app/
- Ryan Geiss, two people before a mountain landscape, HDR+ on, published 2014. Google Research and its linked comparison album. No file-specific open licence confirmed; in copyright, link only. In copyright, linked: https://research.google/blog/hdr-low-light-and-high-dynamic-range-photography-in-the-google-camera-app/
- Andrei Manuilov, laundry between apartment buildings, published 2020. Apple Newsroom, Night mode challenge. No file-specific open licence confirmed; in copyright, link only. In copyright, linked: https://www.apple.com/newsroom/2020/03/apple-reveals-the-best-night-mode-photos-shot-on-iphone/

## Sources

- Marc Levoy, Google Research, HDR+: Low Light and High Dynamic Range photography in the Google Camera App, 27 October 2014: https://research.google/blog/hdr-low-light-and-high-dynamic-range-photography-in-the-google-camera-app/ (burst alignment, equal short exposures, reduced noise, highlight retention, shadow recovery, authorship and Nexus 6 chapel example; early documented anchor, not a claim to invention).
- Marc Levoy and Yael Pritch, Google Research, Portrait mode on the Pixel 2 and Pixel 2 XL smartphones, 17 October 2017: https://research.google/blog/portrait-mode-on-the-pixel-2-and-pixel-2-xl-smartphones/ (HDR+ base image, small-sensor depth, segmentation and depth-dependent synthetic defocus; Matt Jones comparison).
- Marc Levoy and Yael Pritch, Google Research, Night Sight: Seeing in the Dark on Pixel Phones, 14 November 2018: https://research.google/blog/night-sight-seeing-in-the-dark-on-pixel-phones/ (available-light burst capture, motion-dependent constituent exposures, no flash, white balancing and the distinction between individual exposure and burst duration).
- Apple, Apple reveals the best Night mode photos shot on iPhone, 3 March 2020: https://www.apple.com/newsroom/2020/03/apple-reveals-the-best-night-mode-photos-shot-on-iphone/ (Manuilov and Zhang photographs, attributions, phone models and worldwide publication; maker hometowns are not established image locations).
- Apple Support, iPhone 11 Pro Max technical specifications: https://support.apple.com/en-us/111878 (wide camera f/1.8, Smart HDR, Night mode and portrait depth control; does not establish individual specimen exposures).
- Kyle Chayka, Have iPhone Cameras Become Too Smart?, The New Yorker, 18 March 2022: https://www.newyorker.com/culture/infinite-scroll/have-iphone-cameras-become-too-smart (publisher metadata confirms the author and 18 March 2022 publication date; the publisher's description explicitly identifies overprocessed iPhone 12 Pro and 13 Pro images. This is evidence of aesthetic criticism, not proof that every phone renders identically).
- Revision scope: select only the deep-focus, highlight-retaining HDR strand, not all phone capabilities. The 2017 Google source explicitly distinguishes the phone's broadly in-focus base image from added synthetic portrait blur; the latter is excluded here, which is why optics.deep remains defining. The four defining decisions are computational phone capture, HDR rendering, deep focus in the finished image and retained bright-area detail. finish.clean is characteristic only: denoising is a tendency, not a guarantee of visibly noiseless output. shading.open remains characteristic rather than universal; no flat-shadow or zero-clipping guarantee is made. The inspected Geiss HDR+ ON photograph provides daylight evidence alongside the chapel. The Manuilov Night mode image is retained only as probable: Night mode provenance does not independently establish HDR processing. Zhang's silhouette-heavy night photograph was replaced, not counted as confirmed HDR evidence. No direct ancestry from a film movement is established.

---

From Lumogeny (https://lumogeny.com/), a service of Van Riel Online Services. The shoot sheet for your own camera, lenses and lights is on the page: https://lumogeny.com/styles/smartphone-computational-look/#sheet
