NVIDIA, RED Digital Cinema solve 8K bottleneck

Source: NVIDIA blog.  Video processing.

While 4K video resolution is going mainstream, professionals are working on 8K. This HD standard, which contains four times the pixels of 4K, provides stunning clarity on large screens, but requires massive computing resources.

The extra pixels from an 8K camera give the cinematographer more creative choices in post-production for work that is distributed in 4K. For example, there is more flexibility for image stabilisation, or panning and zooming to reframe a shot without losing image quality in the final delivery format. For visual effects, high resolution can provide more detail for tracking or keying. Downsampling high-resolution video can help reduce noise as well as maintain a high level of quality.

Although state-of-the-art cameras can capture 8K video, the sheer amount of data captured means that video editing can become a challenge. An 8K camera that captures video at 8,192×4,320 resolution, or more than 35 million pixels per processed frame. That is some 250 billion pixels of data for five minutes of a video shot at 24 frames per second (fps), and exponentially more for the hours of footage that would be typically shot.

This is where NVIDIA's Turing GPU architecture and new Quadro RTX GPUs come in. At camera maker RED Digital Cinema, Turing makes it possible for video editors and colour graders to work with 8K footage in full resolution — in real time — reaching greater than 24 frames per second using just a single-processor PC with one Quadro RTX GPU. 

This solution costs less than half the price of a workstation that post-production professionals would typically use, and puts 8K within reach of RED camera users. “RED is passionate about getting high-performance tools in the hands of as many content creators as possible,” said Jarred Land, President of RED Digital Cinema. “Our work with NVIDIA to massively accelerate decode times has made working with 8K files in real time a reality for all.” 

8K video has also led to compromises in viewing files as some workstations lack the processing power required. Video applications like Adobe Premiere Pro, Blackmagic’s DaVinci Resolve and Autodesk Flame are already capable of working with 8K footage captured from cameras like RED’s. This includes footage stored in its REDCODE RAW file format. However, some hardware cannot handle the footage at full resolution. This could cause software to drop frames or delay playback while buffering. Preprocessing footage into a more manageable format is possible, but takes time and disc space. 

By offloading all of the computationally intensive parts of the REDCODE processing to a Turing GPU, NVIDIA and RED are giving post-production professionals access to 8K footage at full resolution in real time. The acceleration technology will also substantially increase REDCODE processing performance on other NVIDIA GPUs.

New capabilities will also be possible with the NVIDIA RTX Tensor Cores and RT Cores available with Turing. Editors will gain from new functionality like artificial intelligence (AI)-enabled upscaling, which will let them intermix archival footage or zoom in beyond 8K resolution with the best possible results.