Menu
Codecs are essential for the entertainment industry, allowing efficient playback and transmission of high-quality videos. Imagine you have a puzzle at home that you assemble and disassemble each time you use it. Depending on the type of puzzle, you need different available resources.
A codec is software or hardware for assembling and disassembling this puzzle. It compresses (disassembles) and decompresses (assembles) video files using certain rules for playback. Some codecs use more CPU (processor) and others more GPU (graphics card). This is important when handling multiple outputs per server or using multiple servers. With higher resolutions in the industry, such as 4K projectors or LED screens with a tight pitch (1.9, 2.9), maintaining synchronization and proper playback can be challenging. Therefore, it is crucial to understand codecs and how they affect system resources, disk storage, VRAM, etc.
A codec is responsible for compressing and decompressing the video, while a container (like MP4, MOV, MPG, MXF) is a file format that stores video, audio, and other data (such as subtitles) in one file.
| Codec | Usage (CPU / GPU) | Storage per Frame (4K) | Colorimetry | Max Support 8/10 bits |
|---|---|---|---|---|
| TIFF | CPU | ~35MB | 4:4:4 | 10-bit |
| PNG | CPU | ~35MB | 4:4:4 | 10-bit |
| TGA | CPU | ~35MB | 4:4:4 | 10-bit |
| DDS | CPU | ~35MB | 4:4:4 | 10-bit |
| JPEG | CPU | ~5MB | 4:2:0 | 8-bit |
| DXV2 | GPU | ~10MB | 4:2:2 | 8-bit |
| DXV3 | GPU | ~10MB | 4:2:2 | 8-bit |
| Hap | GPU | ~10MB | 4:2:2 | 8-bit |
| HapQ | GPU | ~15MB | 4:2:2 | 8-bit |
| HapR | GPU | ~20MB | 4:2:2 | 8-bit |
| Animation | CPU | ~10MB | 4:4:4 | 10-bit |
| ProRes | CPU/GPU | ~20MB | 4:2:2 | 10-bit |
| MPEG-2 | CPU | ~5MB | 4:2:0 | 8-bit |
| H.264 | CPU | ~5MB | 4:2:0 | 8-bit |
| H.265 | CPU | ~5MB | 4:2:0 | 10-bit |
| Notch LC | GPU | ~20MB | 4:2:2 | 10-bit |
| VP9 | CPU | ~5MB | 4:2:0 | 8-bit |
Each codec has limitations in terms of maximum resolution and bandwidth. For example, codecs like H.264 can support up to 8K, but their performance depends on the hardware and compression profile used. Lossless codecs like TIFF or ProRes can handle higher resolutions, but at the cost of requiring more storage space and processing resources.
Modern GPUs from both Nvidia and AMD have significantly advanced in terms of resolution limits. Currently, high-end graphics cards can support resolutions up to 16K. For example, graphics cards like the Nvidia RTX 4090 and AMD Radeon RX 7900 XTX are designed to handle these extremely high resolutions, with capabilities of up to 16384×16384 pixels, making them ideal for entertainment applications and professional use in multi-screen and high-resolution setups.
Modern video cards are equipped with advanced technologies and greater VRAM (video memory) capacity to handle the large amount of visual data required for these resolutions. Therefore, when choosing a graphics card, it is important to consider not only the maximum supported resolution but also the VRAM capacity, monitor refresh rate, and system power and cooling requirements.
The choice of codecs and server configuration will depend on the specific needs of each project. Each type of codec has its advantages and disadvantages, and the right choice will depend on the project’s goals and requirements. For example, for a series of concerts with 3.9 pitch LED screens where the front row is 8 meters away, it may be relevant to choose a graphics card with more VRAM that allows synchronization and adaptability to different opportunities during the tour, using codecs like DXV3 or Hap. For a fixed installation with fewer content changes, it may be better to use codecs like Animation, ProRes, or Notch LC where people are closer and maximum resolution needs to be maintained.
Some live video playback software allows you to overcome the 16384×16384 pixel per GPU limitation, but not all do.
Understanding the capabilities and limitations of each codec allows you to optimize your workflow and ensure that your audiovisual content plays as smoothly as possible on different hardware and software configurations.
All rights to trademarks and trade names mentioned in this comparison belong to their respective owners. There is no intention to infringe such rights. This comparison is provided for informational and educational purposes only.
Copyright © 2020 Plug and Pixel All rights reserved.
