By: Marine Sorin, Senior Product Marketing Manager

“Best, high, top, premium, flawless video quality”—superlatives seem to dominate every marketing brochure in the video streaming industry. But concretely, what do these words mean? Let’s explore how is video quality perceived by four key stakeholders: the content producer, the end viewer, the streaming provider, and the engineer.

For the content producer: No compromises in video quality

For those responsible for the first capture and distribution—whether it’s a live event, a blockbuster film, or a sports game—there is no room for compromise. The goal is to preserve as much information as possible, ensuring that every creative option remains available during the production phase.

In this phase, ultra-high-definition (UHD) resolutions ensure a maximum number of pixels, while low compression for color sampling, such as 4:2:2, retains color fidelity. Recording with a wide color range, like 10-bit depth, captures richer, more nuanced shades, while metadata for high dynamic range (HDR) ensures bright highlights and deep shadows are faithfully represented.

At this stage, the content is raw but powerful, serving as the foundation for everything downstream. The better the source, the more room there is to enhance, adapt, and produce a visually stunning final product.

For the end viewer: Noticing the unnoticed

For viewers, video quality is most real when it’s bad. We’re far from concerns about color sampling or resolutions here; as long as the experience is smooth, no one pauses to admire perfect textures or a pristine twilight sky.

But when something goes wrong—a blocky shadow (dithering), a wavy green field (wobble), jagged diagonal edges (aliasing), or dancing specks around players (noise)— the video streaming quality breaks the illusion. Suddenly, the viewer is pulled out of the story, the match, or the meeting.

Great video quality is invisible. It keeps you immersed without asking you to think about it. The moment a flaw is spotted, video quality becomes a conscious experience, and not in a good way.

For the engineer: Video quality by the numbers

For engineers, video quality is a measurable entity—a series of metrics that tell the story of fidelity and degradation. Whether it’s the latest blockbuster or a live webinar, engineers assess how closely the processed and encoded video matches the original source.

The engineer’s job is to minimize degradation, squeezing the most visual clarity from every pixel. Measuring success is both art and science: while subjective human assessments are accurate, they aren’t scalable. So they rely on metrics like PSNR (Peak Signal to Noise Ratio), SSIM¹ (Structural Similarity Index), and VMAF² (Video Multimethod Assessment Fusion). These algorithms, rooted in human visual perception, are the benchmarks for evaluating quality.

For the streaming provider: Balancing video quality and cost

For service providers, video quality is a financial equation—striking a balance between user satisfaction and delivery costs. And the connection? Bandwidth.

Higher bandwidth equals better video quality, but it also means skyrocketing CDN (Content Delivery Network) costs. Operators seek the perfect balance: keeping viewers happy without breaking the bank. That’s where techniques like bitrate optimization and smarter encoding come in, allowing them to deliver high-quality video while keeping bandwidth usage—and costs—under control.

The many faces of video quality

For content producers, video quality starts with preserving every bit of information for later creativity.
For viewers, video quality is about seamless experiences.
For engineers, it’s a quest for precision and measurable improvement.
For streaming providers, it’s the art of cost-efficient delivery.

Depending on who you’re talking to, “video quality” can mean pristine HDR pixels, a flawless sunset scene, a peak metric score, or a well-balanced budget. Superlatives may tell the story in brochures, but in reality, it’s about understanding the nuances of what video quality truly means.

At MediaKind, we are the engineers behind those seamless viewing experiences. Learn more about how we tackle video quality at www.mediakind.com.

References

1. [1] Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, “Image Quality Assessment: From Error Visibility to Structural Similarity,” IEEE Transactions on Image Processing, vol. 13, no. 4, pp. 600–612, Apr. 2004 ↩︎
2. [2] Zhi Li, Anne Aaron, Ioannis Katsavounidis, Anush Moorthy and Megha Manohara, “Toward A Practical Perceptual Video Quality Metric”, https://netflixtechblog.com/toward-a-practical-perceptual-video-quality-metric-653f208b9652 ↩︎