AV1 video codec specification presented - will save bandwidth

Any idea when pornhub will implement this?
Asking for a friend

 

Does this mean all TV's out there now won't support this because they are lacking the hardware?
Or does this mean that a TV company has to release an update so some TV's can use this with the existing hardware?
Do we all have to buy new ones when it is finally released to manufacturers?

 

same here

 

It cant be worse that when netflix said you needed the latest Intel CPU, win10 and Chrome, to watch their 4k

 

Week after nVidia introduces proprietary HW encoder for it with 20 times higher compression speed per watt than CPU does.
"More You Buy, More You Save."

 

This got my attention: "AV1 offers improved compression compared to VP9 and HEVC", that's really impressive.

 

What caught my eye is this:

AV1 offers improved compression compared to vp9 or hevc, the video bandwidth reduction can run upwards to 30 to 40 percent, without you seeing a difference.

Which is a superbly sneaky thing to say. Because of the implication that is saves 30-40%. Yet this comes at the expense of image quality -> "without you seeing a difference".
So it's not apples to apples, because it's not REALLY the same video quality, is it?

For all we know, it's totally possible HEVC/vp9 can save 30-40% compared to AV1, without us seeing a difference

 

If you don't want to see any difference you should not be using lossy compressing in the first place.

24-30-48-60-120 TIFF's per second.
Guess how much bandwidth does that consume ?

The entire point of lossy image/audio/video encoding is for the image or audio to be "good enough", perceptually the same as the uncompressed one.
If AV1 can show a video that is perceptually similar to HEVC with 30% less bitrate, it's win-win !

I'm guessing they are saving more bitrate on the motion encoding, so if you look at extracted screenshot from H264 or H265 and one from AV1, the AV1 might be worse. However, when it's playing, the difference might be completely imperceptible. We will see... (literally!)

 

Who even mentioned uncompressed?

The comparison is between AV1/HEVC/VP9 in lossy encoding. And it's a mess out there when it comes which one is the best. As I thought. Depends on who you ask.

No I don't want to see any difference between the compared samples. And if you're comparing the samples fairly your job is to minimize the quality difference (no need for lossless), not to optimize one for bandwidth, and then claim there is "almost" no difference, yet the bandwidth is saved. That's pure bias. You don't know what am I going to do with the video, am I going to edit it or zoom in, and then suddenly I AM SEEING the difference.

You could have just as well optimized HEVC for bandwidth and pretty much claim the same thing, if the codecs are of similar quality, and it seems that they are.

 

I wish I still worked there, so I could tell you..

 

This is possibly very good for people with slow internet and on phones. I tend to download the biggest x264 I can find so I'm not the target demographic but if they can get something with the same quality for smaller files, why not.

Chromecasts and other streaming devices exist, so that isn't much of a problem. Smart TV software and hardware usually sucks anyway so cheap streaming sticks and boxes should eventually support the new codec and provide a better experience.

 

I wonder how well tablets/smartphones will cope with this, especially older ones with slower CPUs. Probably won't be good for battery life either.

My old Q8400 was already getting 100% playing a 1080p HEVC sample.

 

My x5470 at 4GHz is around 60%, but it depends on bitrate, RAM/NB latency and CPU clock. A GPU decoder is a must have.

 

if you doing video editing, u want lossless source in first place

now regarding quality... what matter here is basically "file-size" right ? ... smaller file-size needed to transfer = less bandwith (either counted per-packet or whole size)

simply look back pass enconding
like with HEVC vs H264.... HEVC successful making file size smaller than H264 right
now do you see "quality difference" between HEVC and H264 (at same preset)?

so you think all newer enconder = better "faking" image quality ? which in otherway old encoder = better image quality but poor compression ?
like your claim "Yet this comes at the expense of image quality -> "without you seeing a difference"."

 

What you want, and what you have, are two different things. I am allowed to edit, even if I don't have access to uncompressed..?

At what bit rates?.
Someone already did it for me. I take HEVC when I can because it's been determined to be better by objective measurement, especially on lower bitrates.

What are we arguing about?
I said they used funny wording "without you seeing a difference", suggesting subjective evaluation has been used.
And it's a mess out there when it comes to evaluating and comparing these codecs, even with objective methods. There is no consensus. Not least because the codec is not finished, and they have delayed the bit freeze for the 3rd time.
And it's slow as fuk. The aim is 5x slower in encode and 2x in decode than VP9 by the end of the year.

Now where is 30-40%?
(latest version 1.0)
https://forum.doom9.org/showthread.php?s=6c90646d3f2ad068ea121e7fc73e4ee0&p=1837312#post1837312

Code:

some tests with foreman clip @250kbit/s on my i7-2600k:


x265 - 2.7+3-9086c8a3e76d:[Windows][GCC 7.3.0][64 bit] 8bit+10bit+12bit
preset : veryslow
real bitrate 254 kb/s size 311 KiB
time : pass 1 : 39s pass 2 : 39s total time : 78 s
metrics :
SSIM All:0.952898 (13.269586)
PSNR average:37.447079
VMAF score = 95.024192

libvpx-vp9 v1.7.0-213-gf4b1eca53
preset : cpu-used 0
real bitrate 238 kb/s size 291 KiB
time : pass 1 : 2s pass 2 : 68s total time : 70 s
metrics :
SSIM All:0.952918 (13.271442)
PSNR average:37.275722
VMAF score = 94.977788

aomenc 0.1.0-8871-g7a3c26460
preset : cpu-used 2
real bitrate 252 kb/s size 307 KiB
time : pass 1 : 2s pass 2 : 1254s total time : 1256 s
metrics :
SSIM All:0.951930 (13.181228)
PSNR average:37.565993
VMAF score = 95.79106

 

LOL correction

x265 :
SSIM All:0.956373 (13.602420)
PSNR average:38.090090
VMAF score = 96.459293

AV1 got its azz kicked handily by both VP9 and HEVC LOL

 

Glad to hear its free to use. Hopefully with adaption it will further be optimized.