Our Pixel 11 Pro XL sample gallery is now live and covers the newest camera modes on Google’s flagship smartphone, including Magic Capture and Camera Looks. We shot photos using all three rear cameras on the Pixel 11 Pro at their maximum 50MP resolution. 

If you’re considering a Pixel 11 Pro or Pro XL for its camera upgrades, this gallery offers real-world output files you can evaluate before you buy.

We’ve been shooting a Pixel 11 Pro XL production unit since launch day and have done our best to capture photos that reflect real-world scenarios across a range of lighting conditions and scenes. Since the Pixel 11 Pro XL and the smaller Pixel 11 Pro share the same triple-camera system, the results here are representative of what you would get from either model.

Dale Baskin

In addition to JPEG files, we have included downloadable Raw files for every mode that supports them. That means you can pull the files into Lightroom, Capture One or any other Raw-capable editor and draw your own conclusions rather than relying solely on the out-of-camera samples. However, it’s worth keeping in mind that on any modern smartphone, the manufacturer’s computational and AI technology is an important part of the image processing pipeline.

The gallery includes several examples captured using Google’s new Magic Capture mode and Camera Looks.  Magic Capture lets you capture video while AI analyzes the footage to select what it considers the best frames, converting them into 12MP stills. Since Magic Capture works from video frames, it doesn’t produce Raw files, so those samples appear in the gallery as JPEGs only.

Camera Looks lets you dial in a visual style before you shoot rather than applying a filter after the fact.  We photographed two distinct scenes, one rural and one urban, using every available Camera Look, so you can see how each style performs across different types of subject matter. For more information on how both features work, see our hands-on look at Magic Capture and Camera Looks.

Sample gallery

Please do not reproduce any of these images on a website or any newsletter/magazine without prior permission (see our copyright page). We make the originals available for private users to download to their own machines for personal examination or printing; we do so in good faith, so please don’t abuse it. 

Google’s new Pixel 11 smartphones include two headline camera features that represent different approaches to computational photography. Magic Capture aims to take the stress out of capturing the perfect moment by letting AI identify the decisive moment for you, while Camera Looks gives you greater control over your creative output by letting you define a visual style before you ever press the shutter button.

We’ve been testing both features on a Pixel 11 Pro XL to see how they work in practice, and whether the results hold up. The short answer is yes, but not without trade-offs.

What Magic Capture actually does and what the sample photos reveal

Magic Capture replaces the act of pressing a shutter button at the right moment. You roll video instead of shooting stills, and Google’s AI analyzes the footage to identify what it believes are the most compelling frames, then turns them into photos. The feature goes beyond simply extracting a frame from video and saving it as a still image: according to Google, the system analyzes roughly 400 frames in real time, combining on-device processing with Gemini AI technology.

After stopping a recording, you can see the phone’s initial frame selections quickly, but they look flat and unfinished. Once the Gemini AI processing completes, which took around a minute or so in our testing for a 15-second clip, the selected frames become noticeably improved, popping with more detail, and with HDR scenes in particular looking significantly better. The final output is a 12MP still image.

The sample images below come from a 14-second Magic Capture clip of a surfer riding a standing wave in a river. Google’s AI did a pretty good job of selecting frames that captured the highlights.

Image quality exceeds what you would typically expect from simply extracting a frame from a 4K video shot on a smartphone; it’s crisper, with better colors, and displays in HDR (when viewed on an HDR display).  It’s not quite up to the quality I’m used to seeing from a standard still photo shot on a modern 12MP smartphone, but it gets the job done, particularly if it’s mostly going to be viewed on other mobile devices.

For the best results, Google recommends enabling Video Boost before using Magic Capture, which ensures the underlying footage is captured in 10-bit 4K. That setting is worth locating before you go out to shoot, not after. Processing time is manageable for a single clip, but if you shoot several Magic Capture sequences back-to-back, the phone queues them for analysis, which can take a few minutes to complete.

One thing we haven’t fully assessed: how consistently Google’s AI selects the frame a photographer would have chosen. The feature’s premise, that the AI identifies the peak moment better than you can, is a pretty strong claim. Whether it matches a photographer’s instincts for capturing the decisive moment would require more testing to answer.

How Camera Looks works and what the photos look like

Camera Looks gives the Pixel 11 camera a selection of color and tone styles you can apply before you shoot. Google says the goal is to produce images that feel less processed and more like a deliberate photographic choice. The system is built around two layers that work together: base styles and creative presets.

In addition to Original (Google’s standard photo processing pipeline), there are three alternative base styles: Natural, Shadows and Vanilla, which sit at the top of the processing pipeline. They aren’t filters. They change how the camera collects and processes light before an image is generated, affecting how aggressively HDR merges multiple frames, how dark shadows are allowed to stay and how edge sharpening is handled. Think of them as choosing the core processing pipeline the camera runs on.

On top of those base styles, you can apply one of six creative presets: Editorial, Velvet, Classic, Digi, Black Tie and Minimal. These apply LUTs meant to emulate film chemistry, specific eras of digital photography or stylized monochrome profiles. The combination you pick matters: applying Velvet on top of the Original base produces a different output than Velvet layered over Vanilla.

The three alternative base styles can be set as your default shooting mode and can be customized. The six creative presets live in a separate carousel, and the interface can take a moment to get used to. The distinction between the two layers isn’t obvious at first glance, and understanding which choices you can adjust versus those you can simply apply requires some exploration. Once you’re oriented, the workflow makes sense, though.

Camera Looks fits a trend that’s already well established among both dedicated cameras and smartphones, where manufacturers offer in-camera film simulations, LUTs or custom picture profiles that let photographers customize their own looks. Google is applying the same logic to the Pixel 11 series, which means photographers who already think in terms of a preferred style will find the concept familiar.

Who are these features for?

Magic Capture and Camera Looks take different philosophical approaches to computational photography. Magic Capture removes a decision – when to fire the shutter – and replaces it with an AI judgment call. Camera Looks adds a decision, letting you define a visual style before you shoot rather than correcting for it afterward. One reduces your control while the other expands it.

Magic Capture makes the most sense for scenarios where keeping up with fast, unpredictable action is genuinely hard: kids, sports, pets, or anything where the peak moment is difficult to catch manually. Or, as Google pitches, it lets you “live in the moment” by just rolling video and not stressing about capturing the decisive moment. If you really care about which exact frame gets selected, though, you’re trading that choice for convenience. Google’s AI selections in our sample photos were solid, but the feature works best when you’re willing to accept its judgment.

Camera Looks, by contrast, gives photographers something to bite into and exert creative control. Setting a default look before you shoot is a reasonable workflow for anyone who knows the aesthetic they want and doesn’t want to recreate it in editing every time.

Both features reflect where Google is taking the Pixel camera: toward a model where AI handles more of the technical and even creative work. Whether that’s a net gain depends on how much of that work you wanted to do yourself.

If you’ve ever stood on an observation platform at Niagara Falls, at the edge of the Grand Canyon, or overlooking the Golden Gate Bridge, chances are pretty good you’ve seen a Tower Optical binocular viewer, and maybe even dropped a quarter into one, even if you’ve never heard of the company that made it.

These heavy, chrome-plated viewers are bolted to observation decks and overlooks across the US and Canada. But somewhere along the way, the viewers themselves became icons. And, for the first time in its history, the company is selling a small number of them to private buyers.

Tower Optical began in a Norwalk, Connecticut, machine shop in 1933, and remained in the same family until 2025. Between the 1930s and 1980s, it manufactured roughly 2,000 of the machines, with every component hand-assembled in the company’s factory in Norwalk, where the original machinery from the 1930s and 40s still stands. Today, the company has about 1,800 machines spread across roughly 400 locations in North America.

Tower Optical

In many places, these viewers have become part of the landscape as much as a means of viewing it. As someone with an interest in optics, what’s always fascinated me are the conditions these devices endure: freezing winters, blistering summers, rain, snow, ice, sand and every other environmental hazard you can throw at them. Yet the 10×42 binoculars inside keep delivering a sharp image, decade after decade.

Adam Rice, one of the company’s current owners, says the location and history of every viewer has been meticulously tracked by hand, decade after decade, as they moved around the continent.

Tower Optical / Adam Rice

“There are two push-pin maps at headquarters that are pretty outdated now, but they were used to track all the different sites, color-coded by location. Beyond that, we have a physical card for every single machine detailing its maintenance and whereabouts. I scanned thousands of these cards with a high-speed scanner to digitize the company’s historical records.”

Looking at one of these cards can be like a walk through history, Rice tells us. “If someone’s interested in the history of a machine, we’ll look at the maintenance cards… For example, machine #507 was manufactured in the 1950s. Our records show it went to the Staten Island Ferry in 1961, Battery Park in New York City in the ’80s, the North Carolina Zoo in the ’90s, and, most recently, Coit Tower in San Francisco.”

Tower Optical

One thing the company isn’t doing today is building new units, and the original manufacturing process may be lost to time.

“I’ve tried to understand the original manufacturing process, but no one is really left who knows how they were originally made. In our warehouse, there’s about 5,000 square feet of space filled with lathes and grinders from the ’30s and ’40s with overhead shafts to operate the pulleys. Our machinist estimated that if you had to make one of these domestically today, it would probably cost $15,000 to $20,000. The viewer heads are cast bronze and chrome-plated, and the rest of the components are iron.”

The glass inside has a pedigree of its own. “Interestingly, the metal plates on the viewers used to say ‘Bausch & Lomb Optical Viewer,'” he continues, “because the internal binoculars were manufactured by Bausch & Lomb. Many units still have the original Bausch & Lomb optics inside. At one point, they had a partnership where if binoculars needed to be serviced or replaced, we would send them directly to Bausch & Lomb, and they would fix them.”

Not much has changed over the years. The price crept up from a nickel, and the coin mechanisms were revised and hardened against people who tried sawing or drilling their way to the quarters inside.

Tower Optical

The biggest challenge the company faces today is one the founders probably never saw coming: “No one carries quarters anymore, but people still love using these machines, so our goal from day one was to retrofit them with tap-to-pay,” says Rice.

“Another challenge was that every machine was built by hand, so the dimensions aren’t exact CAD measurements. The engineering firm we worked with did an outstanding job engineering retrofit brackets that fit regardless of slight variations in screw hole placement between a machine built in the ’30s versus one built in the ’50s.”

What the company was unwilling to change was the viewers’ nostalgic appearance.

“Our mandate to the engineering firm from day one was that we did not want the look of the machines to change beyond the payment terminal itself. The terminal is very minimal, just a small black box, and we’re even considering putting a chrome foil on it to blend in further. The only other physical modification was drilling a small hole in the top shell for a cellular antenna,” Rice explains.

While many have encountered Tower viewers in their travels, it’s never been possible to own one. Tower Optical’s viewers are placed through revenue-share arrangements with host venues, but for the first time, the company has decided to sell a small number of the historic units.

Tower Optical

“It allows us to finance the engineering work and expand our network, since retrofitting and deploying machines isn’t cheap. We receive a ton of inbound interest from people who want to own one. It makes sense to pull the lever on sales for units we can’t retrofit,” says Rice.

“In an ideal world, we probably don’t want to sell more than 50 to 100 of them.”

Owning a piece of history won’t be cheap. According to the company’s website, they start at around $9500. “That includes full refurbishment, the original historical maintenance card, and a new pair of optics,” Rice tells us. “We work with OM Digital to supply those binoculars. Prices may carry a premium for units originating from iconic locations like the Empire State Building, Rockefeller Center, or the Statue of Liberty.”

So next time you’re at a scenic overlook in the US or Canada, look out for a Tower viewer. Rice envisions a future in which they might even get more upgrades than its founders could have imagined.

“Moving forward, we’re exploring ways to layer in digital experiences, like augmented reality at high-traffic sites like Rockefeller Center, while keeping the core viewing experience intact.”

Want to stay up to date on the latest product news and releases? Add DPReview as a preferred source to ensure our independent journalism makes it to the top of your Google search results.

We recently got our hands on the new Google Pixel 11 Pro XL smartphone, and it’s time to do some testing.

In addition to shooting a sample gallery to test the new cameras, we’re also planning to test two new features Google rolled out alongside the Pixel 11 series: Camera Looks, which let you customize how your photos look at the moment of capture, including presets Google claims will look “less processed,” and Magic Capture, a mode that captures photo and video simultaneously and uses AI to automatically select the best shots and a highlight clip once you’re done.

But before we dive in too far, we wanted to know what questions you have about Google’s newest Pixel Pro smartphones. Let us know in the comments, and if there’s something you’d like to see us test, let us know that too.

Want to stay up to date on the latest product news and releases? Add DPReview as a preferred source to ensure our independent journalism makes it to the top of your Google search results.

Skylum has announced an update to Luminar – its photo editing suite that became synonymous with AI sky replacement – which arrives under the pitch of protecting human talent and creative vision.

According to Skylum, the update is arriving in two phases. Phase 1 shipped on August 5 as version 1.28.0, and includes user interface and performance improvements. The Catalog and Edit modes have been merged into a single workspace, with Presets, Tools and Edits gathered into one reorganized panel on the right. Skylum claims exports are now up to 27% faster, preset previews up to 3.8 times faster, and masking uses up to 33% less memory.

The release also adds camera support for the Leica SL3-P and M11-P, the Panasonic Lumix DC-L10 and the Sony a7R VI, plus compressed Raw support for the a7 V. Additionally, the company says it has rebuilt its generative model to produce results that blend more naturally into the original photograph.

Skylum

We also noticed a quiet change: “Neo” appears to be going away. Skylum’s August release notes are headed “Luminar 1.28.0” whereas June’s were headed “Luminar Neo 1.27.1,” and the press materials don’t mention the word. Skylum’s own website hasn’t caught up, and still refers to the app both ways.

Phase 2 is slated to arrive this fall and will focus more on feature improvements. The company says a revised Crop tool will evaluate an image and propose several finished compositions, weighing subject, balance, leading lines and visual weight, and the Atmosphere Fog tool is being rebuilt to respond to a scene’s light, color and depth rather than laying a flat layer over the top. An improved Enhance tool is designed to reduce the risk of overprocessing, and the AI Assistant is being updated to take a description of the look you’re after and apply a combination of Luminar tools to deliver the result you want.

Finally, Skylum says it’s introducing a new web editor to support cross-platform editing.

Skylum

Skylum’s earlier marketing leaned heavily into AI as a selling point, through Luminar AI’s launch and Luminar Neo’s debut. This announcement reads a bit differently: Skylum says the update arrives as the industry grapples with fears that AI might “overshadow human talent” and bring an end to what it calls a nearly 200-year-old craft, with CEO Ivan Kutanin adding, “Creative vision comes first; technology is here only to let it shine.” 

That’s an interesting position from the company that popularized one-click sky replacement, but Skylum seems to be drawing a distinction between craft and documentary photography; effects that prove convincing in a scene versus ones that look pasted in. (The fact that the software also lets you reshape a subject’s body in a portrait makes that a tougher sell.)

Price and availability

Luminar is available as a one-time purchase in three versions. A desktop-only version for Mac and Windows sells for $129. The All Platforms version adds support for iOS, Android and ChromeOS apps, cross-device editing and access to Luminar on the web when it arrives this fall; it retails for $259, but is on sale for $159 through August 31. The Max version adds a year of the company’s Luminar Prime service and retails for $399, but is on sale for $199 through August 31.