Transcription
Summer 2007 at WWDC, Apple unveiled Leopard. Now, let's move on to our big cats, a new Mac OS with a completely redesigned desktop. It introduced features that would later surface across every Apple platform. And its centerpiece was the new dock. This wasn't just a fresh 3D shape. It sat on a glossy glass shelf with live reflections. The dock was shiny yet not blurred. Both its reflections and the desktop behind it remained razor sharp. The menu bar, however, carried a subtle background blur if you looked closely. That tiny flourish would soon trigger a major shift.
Leopard wasn't the first blur on the market. Earlier that year, Microsoft's Windows Vista shipped with Aeroglass, and the Compass Window Manager added a blur plugin for Linux, but it marked Apple's first systemwide leap toward real-time GPU effects, all in the same year. What on earth happened in 2007? Why did everyone suddenly embrace blur? 2007 was the first year when consumer GPUs learned to calculate Gaussian blur in 1/60th of a second. "It's perhaps the most thoroughly thought out OS release from Apple yet," wrote Wired in its October 2007 review of Leopard, noting design more than is usual for a review like this. The combination of translucent glass, sharp reflections, and subtle background treatments laid the groundwork for Apple's later material language and hinted that interface design was heading somewhere new and far more exciting very soon.
To understand how this changed everything, we'll explore three questions. What exactly is blur inside a computer? Why didn't it appear sooner? And why does it feel so natural to you? When we talk about human eyes, we fall into the trap of seeming simplicity. In everyday life, you can just defocus, turning the world into a soft haze. But in graphics, things get harder. There's no lens to adjust, no eye. Every pixel has to be calculated and computed a lot.
In essence, blur means mixing pixels. If we mix them randomly, the whole picture turns into pixel mush. So, we need to mix only a certain number of neighboring pixels. The mixing radius is exactly the size of the little box around each pixel inside which everything is mixed. Set the blur radius to five and each pixel involves about 100 neighbors. Blurring a full HD image is roughly 200 million calculations. A bit trickier than simply defocusing your gaze. This method is called box blur because such a box is formed around every pixel. The problem is that this calculation produces an unnatural blur. It doesn't look much like what you see all over today's interfaces.
A different approach is used: Gaussian blur. Instead of drawing a box around the pixel, you draw a bell-shaped radius. The closer the neighbors are to the pixel, the more they are mixed. The farther away, the weaker the effect. That gives a blur that looks far more natural, like real optical defocus. The catch is that computing such realistic blur is hundreds of times more demanding for the processor. And remember, we're talking about interfaces, about animation. So all those calculations must happen every single frame. No CPU could handle that load.
The most natural blur happens when the subject stays razor sharp and everything behind it fades smoothly into haze. Cameras and phones have tried to copy this for years, but the results still feel off. Shaky edges, broken detail, so designers still build true depth blur by hand, masks, layers, patience. While I was crafting this look for the video, I ended up with a clean smoky sphere. I liked it so much I printed it on a white tea. It just looks quietly pleasing. Hard to explain, but it works. If you'd like one, too, the links are in the description.
GPUs saved us from universal sharpness. The key difference between CPUs and GPUs is that GPUs can perform calculations in parallel rather than sequentially. One-off complex calculations favor the CPU, but tasks that need a huge number of simple simultaneous operations like pixel math are GPU territory. That's why in the 2000s, both Apple and Microsoft were working in parallel on GPU-based rendering systems. Microsoft's desktop window manager and Apple's Core Image played almost the same role. Up to 2007, Apple and Microsoft experimented heavily with blur, hunting for loopholes and hacks to make it both smooth and natural. But nothing really worked until GPUs of the needed level became ubiquitous. The same applies to phone hardware. Before iOS 7, Apple avoided blur in the iPhone UI, and even iOS 7 used it sparingly, disabling translucency and live blur on older devices.
But why was everyone so obsessed with blur? An interface's job is to show shapes and boundaries. Yet, blur seems to do the opposite. Because blur is not just another visual gimmick. It's a fundamental device comparable to negative space in design or to silence in music. It's a basic building block of human perception. Right now, you're looking at a monitor or phone. Behind the screen, something exists in physical space, but your eyes and brain are focused on the object of attention, while everything in front of or behind it falls into blur. Filmmakers grasp this immediately. The camera works partly like an eye, and optical blur is a basic way to guide attention, not merely a pretty effect. When a blurred background sits beneath a menu, it creates special depth inside the interface. And depth has been essential since the very first graphical user interface prototyped under DARPA, not Xerox. From version one, we had overlapping elements, meaning space was implied. Space without blur is a lifeless imitation. Thus, blur, which is essentially noise, chaos, fuzziness, is needed so that something else becomes sharp and captures center stage. That is why interfaces progressed toward realism in the basic principles of human perception.
And here lies the fundamental difference between Microsoft's and Apple's approaches. So when design had to grow up, abandoning skeuomorphism for flat, clean forms, a fork appeared that perfectly exposed that difference. Windows went first with an absolutely flat interface. But with flatness and the death of volume, they also discarded the basic cues of spatial perception, shadows, and blur. Meanwhile, the flat interface Tim Cook shocked the world with in iOS 7 actually pushed those basic principles to the forefront. Guess what happened to Windows? A few years later, they brought back background blur and shadows.
Starting in 2007, Apple has been moving step by step away from illustrative metaphors, yet keeps converging with the real world in fundamental principles. This reaches beyond visual effects to form itself. Did you know Apple icons are not rounded rectangles? They're rectangles with variable curvature. When iOS 7 debuted, it looked as if the corner radius changed from 10 pixels to 12. But in fact, the entire philosophy of shape shifted, ultimately changing the approach of the whole design industry. See our video about 12 pixels that change design forever for the full story. Oh, and if that blurry sphere still lingers in your mind, the tea is waiting in the description.