PDAF (Phase Detection Autofocus) is the dominant autofocus technology in modern smartphones, action cameras, and mirrorless cameras. It works by splitting incoming light into pairs, comparing how the two halves align, and calculating exactly how far and in which direction the lens needs to move — all in a single measurement. Unlike contrast-detect AF, which hunts back and forth, PDAF knows the answer before it moves. On current flagship phones, on-sensor PDAF covers up to 100% of the image area and can lock focus in under 0.1 seconds.
How Phase Detection Autofocus Works
Phase detection borrows a principle from rangefinder cameras. Dedicated photodiodes on the image sensor are masked so each one in a pair sees light from only one side of the lens. When the subject is in focus, the two light paths converge perfectly and the paired signals align. When focus is off, the signals are shifted — and the direction and amount of that shift tell the AF system exactly how to correct.
This is fundamentally different from contrast detection, which only knows focus is wrong, not which way or how much. CDAF has to try different lens positions and compare edge sharpness — a serial, iterative process. PDAF gets the answer in one reading, which is why it’s faster. The processor sends a single, precise command to the lens motor: move this many steps in this direction. Done.
From Dedicated Pixels to All-Pixel Coverage
Early smartphone PDAF implementations (around 2013-2015) embedded a limited number of phase-detect pixels across the sensor — sometimes just a cross-shaped pattern or a few hundred points. These worked well for center-focused subjects but struggled when the subject was near the edges.
Samsung’s Dual Pixel AF, introduced with the Galaxy S7 in 2016, was a breakthrough: every pixel on the sensor was split into two photodiodes, making every single pixel a phase-detect point. This meant autofocus worked anywhere in the frame with equal speed and accuracy. Sony followed with similar on-sensor approaches, and by 2020, full-sensor PDAF coverage became standard in flagships.
Current implementations go further. Apple’s “Focus Pixels” cover 100% of the sensor. Google’s Pixel phones use dual-pixel data not just for AF but for depth estimation in portrait mode. Samsung’s 2023+ sensors use “Dual Pixel Pro” with 2×2 pixel structures that detect phase in both horizontal and vertical directions, improving accuracy on diagonal patterns and complex textures.
PDAF vs. Contrast Detection vs. Laser AF
Contrast-detect AF (CDAF) is the simplest and oldest method. It works by analyzing the captured image, adjusting the lens, and checking if contrast improved. Repeat until contrast peaks. It’s accurate — actually more accurate than PDAF at final position — but slow because it’s trial-and-error. You see it as that annoying back-and-forth “hunting,” especially in low light where contrast is low.
Laser AF measures subject distance using a time-of-flight infrared pulse. It’s fast at close range (under 2 meters) and works regardless of light or contrast, but it can’t handle distance subjects or track moving ones. It’s largely disappeared from flagships as PDAF improved.
Most modern phones use a hybrid system: PDAF for speed and initial positioning, with contrast detection for final fine-tuning. Some mid-range phones still rely on CDAF alone — which is why budget phones often feel sluggish when focusing in dim environments compared to flagships with full-sensor PDAF.
Why PDAF Matters for Mobile Photographers
Speed is the obvious advantage, but it’s not the only one. Full-sensor PDAF enables real-time subject tracking — the phone can continuously update focus as a subject moves across the frame. This is what makes modern burst mode and action photography viable on phones. Without PDAF, continuous autofocus during video would be a stuttering mess.
PDAF also performs dramatically better in low light than contrast detection. When there’s barely enough light to see, CDAF can’t find edges to analyze. PDAF still gets a phase reading as long as some light reaches the sensor — which is why flagship phones can focus in near-darkness while cheaper phones with CDAF-only give up or hunt endlessly.
For action cameras like GoPro, on-sensor PDAF is what enables reliable autofocus during high-motion activities where the camera is mounted and can’t be manually adjusted. Drone cameras similarly depend on PDAF for maintaining focus during flight when subjects change distance rapidly.
PDAF’s Role in Computational Photography
Dual-pixel data has become an input for computational photography far beyond basic focus. The slight perspective difference between the two halves of each pixel creates a miniature stereo pair — essentially, your sensor sees depth. Google was among the first to exploit this for single-camera portrait mode depth maps on the Pixel 2 and 3.
Modern phones use PDAF depth data for bokeh rendering, subject segmentation, and even low-light noise reduction (knowing what’s at the same depth helps denoise more intelligently). It’s also used to improve HDR stacking by understanding scene depth and adjusting alignment accordingly.
PDAF went from being an autofocus feature to being a fundamental sensing layer. When someone asks what separates a flagship phone camera from a budget one, on-sensor phase detection coverage is one of the biggest technical answers — it affects focus speed, tracking accuracy, portrait mode quality, video stability, and low-light performance all at once. It’s the unglamorous hardware feature that makes most of the glamorous software features possible.
