Input Upscaling Overview
It does not matter if you are a person to watch old DVDs or binge-watch TV shows on Netflix. All incoming content will have a certain standard of definition. The TV you buy at the same time will also have a native resolution that it can, essentially, stretch the incoming content to. For instance, a standard 1280×720 resolution incoming to a TV with a native 4K resolution would look very weird surround by black bars on all sides if not upscaled to fit the 4K screen.
With that being said the 4 most common incoming resolutions are:
- 480p (640x480px)
- 720p (1280x720px)
- 1080p (1920x1080px)
- 4K (3840×2160px)
Each of these has its own algorithms set in place to stretch content to fit the entire screen. So now, the real question here isn’t why, but how. There are a large series of ways this is done and I’ve listed some down below.
Different Upscaling Methods
1. Nearest-Neighbor Interpolation:
Replacing every pixel with multiple pixels of the same color. This results in an image that is larger than the original, and preserves all the original detail but has jaggedness.
2. Bilinear and Bicubic Algorithms:
By interpolating pixel color values, this introduces a continuous transition into the output even where the original material has discrete transitions. This reduces contrast though, which is not ideal.
3. Box sampling:
The target pixel is considered as a box on the original image, and then samples are taken of all pixels inside the box. This ensures that all input pixels contribute to the output. Though with this method it is hard to produce sharp results.
A mipmap is a prescaled set of downscale copies of the original image or scene. When downscaling the nearest larger mipmap is used as the origin. This ensures no issues when scaling the images or scenes making this algorithm fast, and easy to optimize.
5. Edge-Directed Interpolation:
Edge-directed interpolation algorithms aim to preserve edges in the image after scaling, unlike other algorithms which can introduce staircase artifacts
Deep Convolution Neural Networks:
This method uses Machine Learning for more detailed images such as photographs and complex artwork. Very unrealistic that TVs will use this technology, but it is available.
The upscaling to this resolution will matter most for standard definition TV, DVDs. Most other content has a resolution higher than 480p.
The upscaling to this resolution will matter most for HD channels while watching broadcasted TV, some streaming videos and streaming devices. Most other content has a resolution higher than 720p.
The upscaling to this resolution will matter most for Blu-Ray movies, streaming videos, video files saved to a computer, and video games that are not 4K. Most other content will not have a resolution higher than 1080p.
The upscaling to this resolution will matter most for streaming HD/UHD videos, UHD Blu-rays, HDR content, and some video files saved to a computer. Most other content will not have a resolution of 4K quality.
Simply put, you want to be able to enjoy any kind of content as if it just came out on Ultra HD quality. Input upscaling aims to help you with that without you having to do any of the heavy lifting. There are a few common ways it gets done and more often than not, you won’t even notice it’s taking place.
Thanks for sticking with me and I hope this helps you find the right TV!