All of Netflix’s HDR video streaming is now dynamically optimized | by Netflix Technology Blog | Nov, 2023

0
199
All of Netflix’s HDR video streaming is now dynamically optimized | by Netflix Technology Blog | Nov, 2023


It is essential to notice that the fixed-ladder encodes had fixed length group-of-pictures (GoPs) and suffered from some inefficiency on account of shot boundaries not aligning with Instantaneous Decoder Refresh (IDR) frames. The DO encodes are shot-based and so the IDR frames align with shot boundaries. For a given rate-quality working level, the DO course of helps allocate bits among the many varied photographs whereas maximizing an general goal operate. Also because of the DO framework, inside a given rate-quality working level, difficult photographs can and do burst in bitrate as much as the codec degree restrict related to that time.

We A/B examined the fastened and optimized ladders; at the start to guarantee that gadgets within the discipline can deal with the brand new streams and serving new streams doesn’t trigger unintended playback points. A/B testing additionally permits us to get a learn on the advance in high quality of expertise (QoE). Overall, the enhancements may be summarized as:

  • 40% fewer rebuffers
  • Higher video high quality for each bandwidth-constrained in addition to unconstrained classes
  • Lower preliminary bitrate
  • Higher preliminary high quality
  • Lower play delay
  • Less variation in delivered video high quality
  • Lower Internet knowledge utilization, particularly on mobiles and tablets

Yes, we’re dedicated to supporting the open-source neighborhood. The present implementation, nonetheless, is essentially tailor-made to our inside pipelines. We are working to make sure it’s versatile, secure, and easy-to-use for the neighborhood. Additionally, the present model has some algorithmic limitations that we’re within the technique of enhancing earlier than the official launch. When we do launch it, HDR-VMAF may have increased accuracy in perceptual high quality prediction, and be simpler to make use of “out of the box”.

Thanks to the arrival of HDR-VMAF, we have been capable of optimize our HDR encodes. Fixed-ladder HDR encodes have been absolutely changed by optimized ones, lowering storage footprint and Internet knowledge utilization — and most significantly, enhancing the video high quality for our members. Improvements have been seen throughout all gadget classes starting from TVs to mobiles and tablets.

We thank all of the volunteers who participated within the subjective experiments. We additionally wish to acknowledge the contributions of our colleagues from Dolby, specifically Anustup Kumar Choudhury, Scott Daly, Robin Atkins, Ludovic Malfait, and Suzanne Farrell, who helped with preparations and conducting of the subjective exams.

We thank Matthew Donato, Adithya Prakash, Rich Gerber, Joe Drago, Benbuck Nason and Joseph McCormick for all of the fascinating discussions on HDR video.

We thank varied inside groups at Netflix for the essential roles they play:

  • The varied shopper gadget and UI engineering groups at Netflix that handle the Netflix expertise on varied gadget platforms
  • The knowledge science and engineering groups at Netflix that assist us run and analyze A/B exams; we thank Chris Pham particularly for producing varied knowledge insights for the encoding workforce
  • The Playback Systems workforce that steers the Netflix expertise for each shopper gadget together with the expertise served in varied encoding A/B exams
  • The Open Connect workforce that manages Netflix’s personal content material supply community
  • The Content Infrastructure and Solutions workforce that manages the compute platform that allows us to execute video encoding at scale
  • The Streaming Encoding Pipeline workforce that helps us orchestrate the era of assorted streaming belongings

Find our work fascinating? Join us and be part of the wonderful workforce that introduced you this tech-blog; open positions:

[1] L. Krasula, A. Choudhury, S. Daly, Z. Li, R. Atkins, L. Malfait, A. Mavlankar, “Subjective video quality for 4K HDR-WCG content using a browser-based approach for “at-home” testing,” Electronic Imaging, vol. 35, pp. 263–1–8 (2023) [online]
[2] A. Choudhury, L. Krasula, S. Daly, Z. Li, R. Atkins, L. Malfait, “Testing 4K HDR-WCG professional video content for subjective quality using a remote testing approach,” SMPTE Media Technology Summit 2023

LEAVE A REPLY

Please enter your comment!
Please enter your name here