Video Wall Image Scaling: Why it's Important and How it Works
Userful The Visual Networking Platform
Why it's important
With video walls, source content resolution is typically substantially smaller than the combined pixel output resolution offered by the displays in the video wall. The vast majority of existing source content is 1080p with most new content being created now at 4k.
Video walls typically can offer 6k or even 16k pixel output resolution, so it is not uncommon to scale up content four or even sixteen times its original pixel count for output on the video wall.
Scaling on a 2x2 video wall:
With four 1080p displays combined into a 2x2 video wall, you actually have the capacity to display 4k resolution. Hence, the source content needs to be upscaled roughly four times—to 4K—to output to the 2x2 wall (four displays). This scales up the 1080p source file stretching it across four displays. Though the algorithms are typically more complex, in the simplest approach (linear scaling), it takes one pixel and expands it into four.
Scaling is common in video walls and there are many different scaling algorithms that vary in terms of perceived visual quality—basically, a viewer's ability to see pixelation—and CPU demands. One way to achieve scaling is to intelligently utilize the built-in scalers within the display (these hardware-based scalers add no performance load to the server and generally create very good image quality).
One way to do this is by sending a 720p source to a 1080p display. In a network video wall, this “display scaling” can often create quite good results as it places less demand on the network and host CPU. Additionally, the built-in scalers in most displays provide more than decent visual scaling quality.
Another approach is to use a performance-optimized scaling algorithm on the CPU. Userful offers an optimized mix of “server scaling” and “display scaling” to balance performance—available bandwidth/CPU—with visual results. A third approach is high-quality cubic/Lanczos scaling and high-performance server-side scaling, leveraging the optional offloading card.
This enables us to use the best scaling algorithms, and scale on the server-side up to full output resolution without bogging down the CPU. Userful automatically optimizes the scaling algorithms based on a user's setup and content, but also provides the option to manually specify if a user wants to substantially improve visual image quality on heavily scaled video walls.