While 4K replaces HD in our homes, manufacturers are unveiling interesting marketing jargon, such as "Ultra HD upscaling" (UHD). However, scaling is not a unique feature: it simply allows 4K TVs to work with lower resolution video formats, such as 1080p and 720p.
All TVs are on the rise
Scaling means that low-resolution content will fill your entire TV screen. Without it, a low-resolution video takes up less than half the surface of the screen. This is a typical feature on all TVs. Even the 1080p TVs had them: they could switch to high-end content in 720p and display it in full-screen mode on a 1080p screen.
UHD upscaling is what makes your 4K TV work like any other. It can take lower resolution content and display it on the entire 4K screen.
1080p content on a 4K screen is often better than 1080p content on a normal 1080p screen. However, scaling is not magic: you will not get the sharp image you will get from native 4K content. Here is how it works.
Resolution exists at the physical and visual level
Before going to the upscaling, we must understand the concept of image resolution. At a glance, it's a relatively simple concept. An image or video with a high resolution looks "better" than an image or video with a low resolution.
However, we tend to forget some key aspects, namely the difference between the physical resolution and the optical resolution. These aspects work together to create a good image and form the basis of understanding the upscale. We will also cover the pixel density, but do not worry, we will keep things short and enjoyable.
Physical Resolution: On a TV technical sheet, the physical resolution is simply called "resolution". This is the number of pixels in a screen. A 4K TV has more pixels than a 1080p TV and a 4K picture is four times larger than a 1080p picture. All 4K displays, regardless of their size, contain the same number of pixels. While high-resolution physical televisions can use their extra pixels for extra detail, it does not always work that way. Physical resolution is at the mercy of optical resolution.
Optical resolution: That's why your old disposable photos with a disposable camera look better than the sophisticated photos taken with a digital camera. When a photo is crisp and sharp dynamic rangeit has a high optical resolution. Televisions sometimes squander their high physical resolution by displaying videos with poor optical resolution. This leads to blurry images and contrast. Sometimes this is the result of the move upmarket, but we will come back to it in a minute.
Pixel density: The number of pixels per inch on a screen. All 4K displays contain the same amount of pixels, but on smaller 4K displays, the pixels are closer to each other, giving them a high pixel density. A 4K iPhone, for example, has a higher pixel density than a 70-inch 4K TV. We mention this to reinforce the idea that the size of the screen is not the same as the physical resolution and that the pixel density of a screen does not define its physical resolution.
Now that we are all aware of the difference between physical and optical resolution, it's time to scale up.
Enlargement makes a picture "bigger"
Each TV contains a mess of interpolation algorithms, used to increase the quality of low resolution images. These algorithms actually add pixels to an image to increase their resolution. But why would you need to increase the resolution of an image?
Remember, physical resolution is defined by the number of pixels on a screen. It has nothing to do with the actual size of your TV. A 1080p TV screen includes only 2,073,600 pixels, while a 4K screen contains 8,294,400. If you are viewing a 1080p video on a 4K TV without upconversion, the video will only take up a quarter of the time. 39; screen.
For a 1080p image to fit a 4K screen, it must gain 6 million pixels through the upward conversion process (at this point, it will become a 4K image). However, the move upmarket relies on a process called interpolation, which is actually only a game of glorified guilding.
Upscaling reduces optical resolution
There are several ways to interpolate an image. The most basic one is called the "nearest neighbor" interpolation. To perform this process, an algorithm adds a mesh of "blank" pixels to an image, then determines the color value that each pixel should have by looking at its four neighboring pixels.
For example, an empty pixel surrounded by white pixels will become white; while an empty pixel surrounded by white and blue pixels can come out in light blue. It's a simple process, but it leaves a lot of digital artefacts, blur, and robust outlines in an image. In other words, the interpolated images have poor optical resolution.
Left: unedited image. Right: After interpolation of the nearest neighbor. Dean Drobot / Shutterstock
Compare these two images. The one on the left is not edited and the one on the right is the victim of the process of interpolation of the nearest neighbor. The image on the right is terrible, even if it's the same physical resolution as the one on the left. This happens on a small scale whenever your 4K TV uses the nearest neighbor interpolation to scale an image.
"Wait a minute," could you say. "My new 4K TV looks like nothing like that!" Well, it's because it does not rely entirely on interpolation by the nearest neighbor: it uses various methods to improve the images.
Attempts to move upmarket also try to attack the optical resolution
Ok, so the interpolation of the nearest neighbor is imperfect. This is a brutal method to increase the resolution of an image that does not take into account the optical resolution. This is why televisions use two other forms of interpolation in parallel with the interpolation with the nearest neighbor. These are called bicubic interpolation (smoothing) and bilinear interpolation (sharpness).
Left: An example of bilinear interpolation. Right: An example of bicubic interpolation. Dean Drobot / Shutterstock
With bicubic interpolation (smoothing), each pixel added to an image looks at its 16 neighboring pixels to take a color. The result is a resolutely "sweet" image. On the other hand, bilinear interpolation (sharpness) looks only at its two closest neighbors and produces a "clean" image. By combining these methods (and applying certain contrast and color filters), your TV can generate an image without significant loss of optical quality.
Of course, interpolation remains a guessing game. Even with proper tweening, some videos may become "ghosted" after resizing – especially if your cheap TV fears an upgrade. These artifacts also become more apparent when very low quality images (720p and lower) are converted to 4K resolution, or when images are resized on extremely large TVs with low pixel density.
The image above is not an example of upward conversion from a TV. It is rather an example of the move upmarket performed for the release of Buffy The Vampire Slayer DVD HD (from a video test of Passion of the nerd). This is a good (even extreme) example of how bad interpolation can ruin an image. No, Nicholas Brendon does not wear waxy vampire makeup, which is what happened to him in the process of going upmarket.
While all TVs offer upscaling, some may have better upscaling algorithms than others, which will give a better picture.
Scaling is necessary and rarely noticeable
Even with all its faults, scaling is a good thing. This is a process that usually goes off without a hitch and allows you to watch various video formats on the same TV. Is it perfect? Of course not. This is the reason why some movie purists and video games prefer to enjoy old art on its intended support: the old-fashioned TVs. But, for the moment, the move upmarket is not something exciting. Nor is it annoying.
It should be noted that 8K, 10K and 16K video formats are already supported by some of the material we use every day. If large-scale conversion technology fails to catch up with these high-resolution formats, it will result in a much greater loss of quality than we have a habit of living.
Manufacturers and streaming services still being shuffling their feet towards 4KHowever, we should not be worried about 8K at the moment.