The first time a client told me my edits looked “plastic,” I wanted to close my laptop and never open Photoshop again. I’d spent three hours on her portrait. The eyes especially, where I’d gone in with the Dodge tool like I was excavating treasure, burning whites to pure white and irises to this glassy, oversaturated turquoise. She looked like a doll. Not a good one.

That was eight years ago, and I still keep that file in a folder called “Humbling Moments.” It’s the best teacher I’ve ever had.

Eye enhancement is where most portrait retouching either earns trust or destroys it. Get it right, and people say the photo is “stunning” without knowing exactly why. Get it wrong, and something feels off in a way that’s hard to articulate but impossible to unsee. The eyes are where we read emotion, aliveness, and authenticity, which means they’re also where heavy-handed edits register first and fastest.

Why Eyes Are So Unforgiving to Retouch

The human eye, in a photograph, contains more tonal variation than almost any other facial feature. You’ve got the deep shadow of the pupil, the layered color of the iris, the translucent sclera with its faint pink vascular network, the wet highlight that tells the viewer there’s a real person in the frame, and the fine texture of the limbal ring, that dark circle around the iris edge that naturally fades with age and fatigue.

When we over-process any one of these, the relationship between them breaks. Whiten the sclera too aggressively and it no longer reads as tissue. Saturate the iris and it disconnects from the natural color grading of the rest of the image. Remove the wet highlight and the subject looks dead behind the eyes, literally. Retouching eyes well is fundamentally about preserving relationships between tones, not improving individual elements in isolation.

The Layer Stack I Use on Every Single Portrait

I work in Photoshop, and my standard eye enhancement stack uses five layers, each one doing a specific job at low opacity. None of them are destructive. Everything is masked to the eye area.

Start with a Curves adjustment layer, masked to the iris only. Lift the midtones slightly, maybe 10-15 points on the RGB curve. This adds luminosity without flattening the depth of the pupil. Clip this layer to a group so it doesn’t spill into the whites.

Next, add a Hue/Saturation layer over the same iris mask. Boost saturation by 8 to 12 points, not more. Then use the targeted color range selector to nudge the specific hue toward something more natural and saturated rather than simply brighter. This is where you’re working with the subject’s actual eye color, not against it.

For the sclera, create a new blank layer set to Screen mode at 20-30% opacity. Paint with a soft brush using a warm white, around #f5f0e8, rather than pure white. This lifts shadows in the whites while keeping them looking biological rather than photographic.

A Dodge and Burn layer, 50% gray filled and set to Overlay mode, handles the limbal ring and catchlight. Burn the limbal ring slightly to restore definition, and dodge the existing catchlight just enough to give it a little more punch. I use a 20-pixel soft brush at 3-5% exposure here. Single passes only. If you’re going over the same spot twice, you’re already overdoing it.

Finally, for texture, I run a High Pass sharpening layer at 0.8 pixels, set to Soft Light at 40% opacity, masked to the iris. This brings back the fine detail in the iris texture that gets lost in compression and skin softening, and it keeps the eye reading as something real.

The Catchlight Problem Nobody Talks About

One thing I see constantly in workshop submissions: retouchers removing or painting over secondary catchlights. The thinking is usually that multiple light sources look messy or distracting. But a natural catchlight is evidence of a real environment. Removing it too aggressively creates that same doll-eye flatness I accidentally inflicted on my client years ago.

If a catchlight shape is genuinely distracting, I’ll use the Clone Stamp at 30% opacity to soften its edges and reduce its intensity, not erase it entirely. There’s a difference between cleaning up a reflection and erasing proof that your subject exists in physical space.

Matching the Enhancement to the Shot’s Intention

Editorial beauty work and personal portraiture are not the same job, and your eye enhancement shouldn’t be either. For a magazine campaign, I’ll push iris saturation to the upper end of that range, around 15 points, and sharpen more aggressively because the image will be viewed large and the heightened intensity is expected. For a family portrait or a personal branding headshot, I pull everything back by about half. Those clients want to look like themselves, just on their best day.

I learned this the slow way, by sending the same level of enhancement to a corporate headshot client as I’d applied to a beauty brand campaign. She emailed back asking why her eyes looked “too intense.” She wasn’t wrong. I’d treated her like a product when she was a person.

The opacity values matter less than your ability to read what the image is actually asking for before you touch a single slider.

The One Thing That Changes Everything

Every technique in this workflow is in service of the same principle: the goal is not to make eyes look better in isolation. It is to make them look more alive within the context of the whole image. That is the only metric that matters, and checking it means zooming out to 50% view after every major adjustment, every time, before you decide you’re done.