Netflix Finds That Royalty-Free VP9 Video Codec Can Beat HEVC At 1080p Resolutions Or Higher

Status
Not open for further replies.
I'm guessing 4k testing wasn't done because of lack of samples, time and resources (spare cloud processing). I really wouldn't be surprised if it was in the works. I'm more and more impressed by Netflix, seems like they are really improving.
 
We'll see how long it stays royalty-free. Patent trolls might find something it infringes. Compression is such a patent minefield that it's hard to do anything (good) with no risk of infringement.

BTW, I don't know why they even bothered with PSNR or PSNRMSE. I hope those didn't skew the results.
 
Does this mean there's any hope of them ditching Silverlight? I find it a bit frustrating that my mid-range gaming computer can only do stereo and 720p, yet my far weaker phone could inexplicably do 5.1 and 1080p because it uses the app instead.
 
So VP9 is *sometimes* better. The headline made it sound like they'd made some breakthrough, because at most resolutions and content HEVC wins even with the x265 open source encoder. Given how long it took for it's open source predecessor x264 to catch up to commercial encoders, I would imagine there's pay solutions that produce even better results, further distancing HEVC from VP9 in commercial encodes. For most purposes though VP9 is still good enough.



Well if you run Windows 10 you just download the Netflix app and everything works perfectly at any resolution, even UHD where content exists for 4k. Problem solved.
 
You took Netflix's blog post, which explained that x265 outperformed VP9 under most circumstances, and wrote an article with a clickbait title that says exactly the opposite. Nicely done.
 
Not all researchers agree with the statement that VP9 and HEVC are comparable.
For example, the results reported in the paper "Contemporary Video Compression Standards H.265/HEVC, VP9, VP10, Daala" by M.P. Sharabayko, N.G. Markov reveals that HEVC is still superior to VP9.
 
Not all researches agree with the statement that VP9 is comparable to HEVC.
For example in the paper "Contemporary Video Compression Standards
H.265/HEVC, VP9, VP10, Daala" (by M.P. Sharabayko, N.G. Markov, 2016) reported that HEVC is still superior to VP9.
 


It's not your computer's fault...
"5.1 surround sound is not currently supported while streaming on a computer using Microsoft Silverlight or HTML5."

"The HTML5 Player may limit the quality of the stream while the Silverlight player does not. HD resolution streams are only available if the Internet connection supports at least 5 Megabits per second, but that is not the only restriction."

As for the 1080p resolution you could probably blame that on your ISP, not your computer.
 
Last I checked x265 still needed a bunch of tuning, but I don't know how that's going. I did have a wtf situation when x265's lossless mode created a larger file than x264's, but that's for lossless so ymmv. I still don't get the x265 crf values either as they scale differently from x264 where I usually use crf 20-22, but my Athlon II x4 640 is dead slow with x265 so I haven't done much experimentation. I'll play with it some more after I get my hands on a Zen processor.
 


If you're on Win10, get the Netflix app. Otherwise you'll have to deal with it until the HTML5 player matures some day.



Yeah x265 still needs work across the board. Better make that Zen purchase an 8C/16T model!
 
And now that the study was presented at the SPIE conference, we see that x265 was significantly more efficient than libvpx (VP9).
 
Status
Not open for further replies.