Google VP9 "Next Generation Open Video" information posted

[GTPVHD]

Google killed 4K H.264 encodes for VP9, you shouldn't be surprised they're killing 8K VP9 encodes for AV1, bandwidth consumption is enormous for Youtube.
This will force laggards like Apple to adopt AV1 just like they did with VP9 support finally in Apple products this year, 3 years after Google stopped 4K H.264 encodes for VP9.

[Blue_MiSfit]

Good luck "forcing" Apple to do anything

[benwaggoner]

Originally Posted by utack

Is Google now actively killing 8k VP9?
A lot of videos only have AV1 as 8k option available, and I found reference to one a year back in a forum that clearly showed VP9 was still available at the time.

The bigger challenge is that reality is killing 8K video because of the limits of human perception. I challenge you to find video content that looks better encoded at native 8K than downscaled to 4K at the same bitrate when viewed at enough distances that the edges of the screen are at a minimally acceptable viewing angle. No one in Hollywood is doing any post production in 8K. Heck, most visual effects are still being done in 2K for cost/speed reasons, and because 2K looks just fine if there's some motion blur and/or grain.

And with everyone working from home, even experimenting in 8K is nigh impossible. There literally aren't any monitors that are 8K, HDR, and have DisplayPort. Monitoring 8K HDR means using a (WAY too big for a desk) TV and a $3K AJA Kona 5 with experimental and finicky HDMI 2.1 firmware.

[takla]

With libvpx-vp9 v1.9.0 auto-alt-ref only works in 2pass. was this changed or was this always the case?

[Beelzebubu]

Originally Posted by takla

With libvpx-vp9 v1.9.0 auto-alt-ref only works in 2pass. was this changed or was this always the case?

Yes (sadly).

[edit]
At --cpu-used=4 using VBR, it works also in one-pass mode in some cases. But at other speeds or in CRF, it works in 2-pass only.

See this code.

[takla]

Originally Posted by Beelzebubu

Yes (sadly).

[edit]
At --cpu-used=4 using VBR, it works also in one-pass mode in some cases. But at other speeds or in CRF, it works in 2-pass only.

See this code.

Thanks for clarifying. But I'm confused. Why was it changed? Was it not working correctly before?

[Beelzebubu]

Originally Posted by takla

Thanks for clarifying. But I'm confused. Why was it changed? Was it not working correctly before?

I meant: yes, it was always the case. Sorry for the confusing answer.

[takla]

Originally Posted by Beelzebubu

I meant: yes, it was always the case. Sorry for the confusing answer.

Ah alright then. Thanks.

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 10.2.0):

VPx v1.9.0-156-g24bd0733e

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 10.3.0):

VPx v1.10.0-61-g4808d831d

[takla]

2021-09-27
v1.11.0 "Smew Duck"

This maintenance release adds support for VBR mode in VP9 rate control
interface, new codec controls to get quantization parameters and loop filter
levels, and includes several improvements to NEON and numerous bug fixes.

- Upgrading:
New codec control is added to get quantization parameters and loop filter
levels.

VBR mode is supported in VP9 rate control library.

- Enhancement:
Numerous improvements for Neon optimizations.
Code clean-up and refactoring.
Calculation of rd multiplier is changed with BDRATE gains.

- Bug fixes:
Fix to overflow on duration.
Fix to several instances of -Wunused-but-set-variable.
Fix to avoid chroma resampling for 420mpeg2 input.
Fix to overflow in calc_iframe_target_size.
Fix to disallow skipping transform and quantization.
Fix some -Wsign-compare warnings in simple_encode.
Fix input file path in simple_encode_test.
Fix valid range for under/over_shoot pct.

Source

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 12.1.0):

VPx v1.11.0-218-g9f1329f8a

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 12.1.0):

VPx v1.12.0-72-g59acf6739

SVT-VP9 0.3.0-d9ef3cc

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 12.2.0):

VPx v1.12.0-153-ga46ca4b6b

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 12.2.0):

VPx v1.12.0-214-gddca3dec3

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 12.2.0):

VPx v1.12.0-228-gd998bd823

[LigH]

New upload (MSYS2; MinGW32 / MinGW64: GCC 13.1.0):

VPx v1.13.0-398-g9ad950a9c