I had a deadline last week. Campaign shoot, eight portraits, client wanted finals by Thursday morning. Somewhere around image four, I hit that familiar wall where Photoshop’s masking behavior was doing something slightly unpredictable around fine hair strands, and I found myself rebuilding a selection from scratch that I had already built twice. Not a catastrophe, just the low-grade friction that compounds across a long edit day and eventually makes you question your tool choices.

That same afternoon, I pulled up Matt Kloskowski’s breakdown of what Adobe shipped in the April 2026 Photoshop update. I’d been following Matt’s tutorials for a while now, because he has a talent for cutting past the press-release framing and showing you what actually changed in practice. What I found was genuinely useful, and a few of these updates would have saved me a real chunk of time on those eight portraits.

The Masking Refinements That Change How We Work With Hair and Skin Edges

The headline update for retouchers is what Adobe has done to refine edge detection inside Select and Mask. Matt walks through how the edge detection algorithm now handles transitional zones more intelligently, particularly where fine hair crosses a lighter or medium-toned background. Previously, that zone required a lot of manual brushwork in Refine Edge or a separate luminosity pass to recover what the algorithm flattened. The new behavior reads contrast shifts across that transitional area more accurately, which means fewer passes to get a clean result.

The practical step here: open Select and Mask the way you normally would, apply your initial subject selection, then move to the Refine Edge Brush and work it along the hair boundary as usual. What you’ll notice is that you can use a slightly larger brush than you might have defaulted to before, because the detection is picking up finer strand detail without the bleed onto the background that used to require you to scrub back down to a tiny brush tip. Matt demonstrates this on a portrait with loose, windblown hair against a mid-gray backdrop, and the difference between the old and new output is visible without squinting.

Generative Fill Behavior When Working on Skin Texture Areas

Matt also covers changes to how Generative Fill handles selections on skin. This is one of those features I have complicated feelings about, but I use it for specific, narrow tasks: filling small blemish areas on body skin where frequency separation feels like overkill, or patching a distraction at the edge of a frame. The update tightens how the fill samples surrounding texture, particularly in areas with directional lighting across skin.

The workflow he demonstrates is straightforward. Make a tight selection around the problem area, feather it two to three pixels depending on how close you are to an edge, invoke Generative Fill with a blank prompt, and let Photoshop generate options. What’s new is that the sampling radius seems to weight closer pixels more heavily, which gives you results that match skin direction and pore scale better than the previous version did. Matt generates three options on camera and shows how the texture integration has improved, especially on cheek areas where the light is raking across the surface.

Neural Filters: What Got Quietly Better

Buried a bit further into the tutorial is something I almost missed the first time through. The Skin Smoothing neural filter received a backend update that affects how it handles the boundary between skin and non-skin areas, specifically around eyebrows, the hairline, and lip edges. Those transitions used to require a mask almost every time, because the filter would bleed a softening effect into areas you wanted to keep sharp.

Matt’s demonstration shows the filter now doing a more reasonable job of self-limiting at those edges. His suggestion, and one I’ve now tested myself, is to run the filter at a lower intensity than you might have previously, somewhere in the 30 to 45 range rather than pushing it higher and masking back. The new edge behavior means a more moderate application reads as natural more often, without the plastic-looking result that used to make me distrust this filter for anything beyond a quick proof.

Where I’d Push Back, Or at Least Adjust

I want to be honest about one place where my workflow diverges from what Matt demonstrates. The improved subject selection and edge detection is genuinely better, but I’m cautious about leaning into it too heavily on very high-resolution medium format files, where I do a lot of my current brand work. At that file size and pixel density, even a well-refined automatic edge can introduce a subtle antialiasing artifact that’s invisible at screen resolution and shows up in print. My habit is still to use the improved selection as a starting point and do a final edge check on a Curves adjustment layer clipped to the mask, viewing the mask itself at 100 percent zoom along the boundary. It takes three extra minutes and has saved me from several print surprises.

This isn’t a criticism of the update. It’s a reminder that any automated improvement still lives inside the larger conversation between the tool and the specific file you’re working with.

The One Thing Worth Remembering From All of This

Photoshop updates tend to be incremental, and the temptation is to either over-celebrate them or ignore them completely. What Matt’s tutorial makes clear is that the April 2026 update is meaningfully useful for portrait and beauty retouchers in specific, testable ways, and the edge refinement improvements alone are worth taking an hour to work into your existing process.

Watch the full tutorial to see these features demonstrated visually, because some of the before-and-after comparisons are the kind of thing that sticks better when you see it in motion: Matt Kloskowski’s April 2026 Photoshop Update on YouTube.