The new machine learning-powered Auto White Balance
Pixelmator Pro includes one of the best collections of color adjustment tools in any app on any device. And we want to make them even better. In Pixelmator Pro 1.1 Monsoon, we added some super cool Auto buttons next to three color adjustments: White Balance, Lightness, and Hue & Saturation. Although those buttons may be small, there’s a lot of machine learning power behind them and we wanted to talk a little bit about one in particular — Auto White Balance.
Fixing white balance
When editing photos, fixing the white balance is almost always the first thing photographers do before moving on to any other edits. It’s important to get it right because all your later edits depend on it. And if you’re not used to doing it, it’s not all that easy — there are two different sliders that are influenced by each other and it’s not that obvious what they do. I mean, adding yellow cancels out a blue tint, how does that work?
So, in Pixelmator Pro 1.1 Monsoon, we’ve added the machine learning-powered Auto White Balance to make fixing the white balance in any photo as simple as clicking one little button. Behind that button, there’s a 200-layer-deep neural network trained on millions of image pairs and integrated via the awesome Core ML.
Did you know that neural networks are practically magic and no one knows how they really work? Basically, getting machine learning models to do what you want them to requires you to test them on really huge datasets. The models then make complicated, seemingly random connections based on your tests and their quality depends on how good your tests and datasets are.
In this case, we needed millions of images with bad white balance and good white balance. It probably goes without saying that there’s no giant repository of badly white balanced images to just download. Even if there was, we’d still need to get someone to fix each one and that’s not exactly efficient either. For that reason, we had to find a different approach. We decided to take millions of good images and create a separate algorithm to ‘break’ them in as natural a way as possible to mimic real-world white balance issues.
Neural network training
Once we had our image pairs, we could start training. Except there was another little problem. Most neural networks work with small, fixed-resolution images, which usually have to be square. That meant we needed to turn every photo into a uniform small square image. But if the photo is portrait or landscape, potentially important details at the edges of the image would be lost if we cropped them. If we simply stretched them, then they wouldn’t look like actual photos people edit.
To get around this, we took each image and applied content-aware scaling. This separate algorithm identified features in the original images (for example, a sky, some trees, or people) and scaled them according to their importance to the photo. The content-aware scaling makes for some pretty strange-looking images — people look a little like weird aliens — but the neural network likes them a little better. Now we could start training and getting our model to learn. And once that was done, we could start testing how accurate our approach was.
Quality and accuracy
Most image editing apps have some sort of non-machine learning automatic white balance correction tools and we compared our approach with the results you get in those apps. We measured quality using the industry standard CIEDE2000 metric and found our Auto White Balance tool to be at least 50% more accurate, and usually between 2 to 10 times better than those standard tools. And we think our approach passes the eye test even better. Just looking at the results, the Auto White Balance adjustment seems to be much more natural and visually pleasing.
One awesome feature of Auto White Balance is support for RAW images and any extended data within them. Our trained neural network uses extended-range information to correct white balance issues in RAW images with improved accuracy. Sweet!
Here are the Auto Color Adjustments getting a mention during the Platforms State of the Union at this year’s Worldwide Developer Conference in San Jose.
Better than humans
To round things off, here’s a fun fact. The new Auto White Balance correctly fixes the white balance in the infamous “dress” photo, which became an internet phenomenon in 2015. And that photo isn’t even in the training dataset! This is an especially good example of white balance and lighting issues which are tricky even for humans to solve. And Pixelmator Pro now makes it effortless with the help of breakthrough machine learning.
If you want to try out these awesome machine learning features on the dress or any of your own photos, you can download Pixelmator Pro 1.1 Monsoon — it’s available now from the Mac App Store and it’s a free update for everyone who has already bought Pixelmator Pro.
Pixelmator Team releases major update to Pixelmator Pro
The Pixelmator Team today released Pixelmator Pro 1.1 Monsoon, a massive update to the world’s most innovative image editing app for Mac. The update adds a modern way to easily prepare images for the web, brings machine learning-powered Auto Color Adjustments, support for the MacBook Pro Touch Bar, a new Slice tool, SVG support, a new Tutorials page to help with learning Pixelmator Pro, and much more.
“Pixelmator Pro 1.1 Monsoon is an absolutely massive update with tons of great new features,” said Saulius Dailide, one of the founders of the Pixelmator Team. “With the innovative new Export for Web features, it’s now easier than ever to prepare and optimize images for the web. And Pixelmator Pro continues to push the boundaries of image editing with the new Auto Color Adjustments, which take advantage of the latest advances in machine learning and computer vision to automatically improve photos.”
Pixelmator Pro now gives you everything you need to prepare and optimize your images for the web. You can use advanced compression techniques to compress images to the smallest size possible without losing quality. Export presets let you quickly optimize images with your favorite settings. As you adjust any export settings, Pixelmator Pro displays a live preview of the optimized image right in the Pixelmator Pro window. Quick Export lets you instantly export a web-friendly image optimized with your most-used settings at any time. And the new Slice tool makes it much easier to prepare web designs by letting you divide entire designs into smaller images, optimizing each one separately.
With the new machine learning-powered Auto White Balance, Auto Lightness, and Auto Hue & Saturation adjustments, you can automatically perfect the colors of a photo with just a click. The machine learning algorithm was trained on millions of images to identify the contents of a photo, finding the best colors, lighting, and contrast to make every image stand out. Integrated via Core ML and powered by Metal, the Auto Color Adjustments deliver lightning fast performance and stunning quality.
Full support for the MacBook Pro Touch Bar gives you quick access to all your favorite tools and all kinds of useful options. You can apply color adjustment or effect presets, select colors for your brushes, and even share your image, all right from the Touch Bar. And whenever you switch tools, the Touch Bar automatically changes to put the most useful options at your fingertips, which makes using Pixelmator Pro simpler and easier than ever.
Pixelmator Pro 1.1 Monsoon includes many more incredible new features, including support for opening and exporting vector images in the SVG format, live previews when formatting text and adjusting layer blending options, a Select Color Range tool that lets you make selections of similarly colored areas in an image, support for exporting HEIF images, a new Tutorials page to help with learning Pixelmator Pro, and much more.
Pixelmator Pro now available on the Mac App Store
The Pixelmator Team today released Pixelmator Pro, a brand new Mac app that redefines image editing on the Mac, providing professional-grade editing tools in an incredibly intuitive and accessible design. Pixelmator Pro 1.0, codenamed Whirlwind, includes a modern single-window interface, nondestructive, GPU-powered image editing tools, machine learning-enhanced editing features, and more.
“We’re incredibly excited to finally release Pixelmator Pro on the Mac App Store,” said Saulius Dailide, one of the founders of the Pixelmator Team. “Its innovative, user-centered design makes editing images on a Mac simpler than ever. And with its GPU-powered and machine learning-enhanced editing tools, Pixelmator Pro is built for the future.”
The modern, dark single-window interface of Pixelmator Pro has been created exclusively for working with images. A streamlined, macOS-inspired design provides a completely native Mac app experience and is fully consistent with the look and feel of macOS. And a reimagined, user-centered workflow design makes the professional editing tools in Pixelmator Pro especially accessible, even to first-time users.
Completely nondestructive color adjustments, effects, styles, and layouting tools in Pixelmator Pro give users the freedom and flexibility to go back and modify or delete individual changes at any point in the editing workflow. Thanks to the new presets feature, you can create endless combinations of multiple adjustments, effects, or styles, save them to your favorites and reuse them in any of your images. Drag-and-drop sharing also makes it a breeze to share presets with others.
Pixelmator Pro pushes the boundaries of image editing, using breakthrough machine learning to deliver more intelligent editing tools and features. Integrated via the new, blazing fast Core ML framework, machine learning lets Pixelmator Pro detect and understand various features within images, bringing a number of groundbreaking advancements, such as jaw-droppingly accurate automatic layer naming, automatic horizon detection, stunningly realistic object removal, and intelligent quick selections.
Designed exclusively for the Mac, Pixelmator Pro uses the Mac graphics processor in tandem with the latest Mac graphics technologies, such as Metal 2, Core Image, and OpenGL, to deliver blistering fast, GPU-powered image editing. Built with Swift 4, Pixelmator Pro is a next-generation app that takes full advantage of the latest technological innovations in hardware and software.
The Pixelmator Team has packed a wide range of features and innovations into Pixelmator Pro, such as full RAW support, including support for multiple RAW layers; HEIF file format compatibility; superior support for Adobe Photoshop images; complete macOS integration via full support for iCloud, Versions, Tabs, Full Screen, Split View, Sharing, Wide Color, and others; Metal 2-powered painting tools and dual texture brushes; powerful retouching, reshaping, selecting, type, shape and vector tools; and a lot more. In addition, a number of free updates adding major new features to Pixelmator Pro are already lined up for the very near future.
Pixelmator Pro 1.0 Whirlwind is available exclusively from the Mac App Store for $59.99.
Pixelmator Pro requires macOS High Sierra and a Metal-compatible graphics card. Full system requirements and more information on Pixelmator Pro can be found at www.pixelmator.com/pro/
Pixelmator Pro Color Adjustments Film
We’ve just released another awesome new film about Pixelmator Pro — this one’s all about the color adjustments.
Black Friday marks the start of the winter holiday and gift-giving season. And what better way to celebrate than by getting Pixelmator and Pixelmator for iOS for a huge 50% off. The sale starts today and will go on for two weeks until December 6th. If you already have and love both apps and know someone else who might enjoy image editing, let them know about the Black Friday sale. Or maybe even surprise them with Pixelmator as a gift. Learn more about the sale.
Art Deco is a style of visual arts characterized by sleek and streamlined forms, bright, vivid colors, and symmetric, geometrical compositions — discover more about it in the latest article on artistic styles.
Pixelmator for iOS gets iPhone X support
We love the new iPhone X. So we took our time to polish the Pixelmator for iOS interface and make sure it looks and feels great on the brand new iPhone X screen.
The update is now ready for you to download from the App Store. It also includes a number of bug fixes and performance improvements. And, as always, it’s free for all those who have already purchased Pixelmator for iOS.
View in App Store
Yup, November 29th. That is the day Pixelmator Pro will be released.
We can’t wait for you to get your hands on what might be the most beautiful, innovative, and inspiring app we’ve ever made. It really is that great.
We didn’t announce the Pixelmator Pro release date earlier because we wanted to take as much time as necessary to build a solid v.1.0. Pixelmator Pro is huge and there are just so many new things and technologies to take care of to make sure everything works as it should. But the current Pixelmator Pro 1.0 Whirlwind beta version I’m running is already rock solid so why not announce the release date is most likely to be November 29th.
Oh, and Pixelmator Pro 1.0 Whirlwind will launch with a price of $59 as we want as many of you as possible to be able to afford it. However, we don’t think the price will stay that low for too long. It will go up as soon as we ship free updates with major new features (that are not yet announced but are in the works already) as well as Pixelmator Pro for iPad (which is a great friend of Pixelmator Pro for Mac), Vectormator, and more. The intended original price for Pixelmator Pro is $99 and we plan to ship big updates quickly.
Get ready for November 29th by subscribing to our awesome newsletter or adding the date to your calendar.
Add to Calendar
And just so you know, you might well be taking part in image editing history here…
Illuminated manuscripts are carefully handcrafted, highly decorated medieval books — learn more about them in the latest feature on artistic styles.
Pixelmator Pro Automatic Layer Naming Film
Our latest film is about one very cool and handy use of machine learning in Pixelmator Pro — Automatic Layer Naming. See it in action below.