Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Video Encoding > High Efficiency Video Coding (HEVC)

Reply
 
Thread Tools Search this Thread Display Modes
Old 17th December 2019, 11:09   #341  |  Link
ReinerSchweinlin
Registered User
 
Join Date: Oct 2001
Posts: 454
I gave myself a christmas present and got a RTX 2060. First tests with NVENC in different GUIs give me good enough quality for my 1080p recodes when speed is important (around 140fps).
So far, I only used default settings. Like pacuro, I was wondering what settings I could use to max out quality (enabling b-frames gave me a bigger filesize than defaults in hybrid). Is there some detailied documentation of the NVENC in RTX Cards I could look into?
ReinerSchweinlin is offline   Reply With Quote
Old 29th December 2019, 07:41   #342  |  Link
tyee
Registered User
 
Join Date: Oct 2001
Posts: 416
I'm using this command line app and it has very nice quality and I also get around 150fps for 1080p, about 50fps for 4k

https://github.com/rigaya/NVEnc

Is John still around the forums? I saw his settings somewhere, he was using Staxrip.
tyee is offline   Reply With Quote
Old 14th March 2020, 10:11   #343  |  Link
pacuro
Registered User
 
Join Date: Sep 2015
Location: Opole, Poland
Posts: 20
@ReinerSchweinlin you probably doing something wrong if enabling B-frames makes your encodes bigger. In my case just enabling b-frames + b as ref each makes files lighter - as it should be in theory.

__________________
LIT BRO: i5-9500, Z370, RTX2060
BIG BRO: i5-10500, H570, GTX1660S
HTPC: i7-11700K, Z570, Topping D10s • MPC-HC + LAV, foobar2000

Last edited by pacuro; 14th March 2020 at 10:18.
pacuro is offline   Reply With Quote
Old 18th April 2020, 14:30   #344  |  Link
craigpro
Registered User
 
Join Date: Feb 2003
Posts: 68
These are the settings I'm using and my 720p encodes are always a few hundred MB larger than similar shows from the scene HEVC encodes. Does anyone have any suggestions for better settings please? Thank you.

--cqp 18:20:22 --codec h265 --preset quality --level 5.1 --output-depth 10 --qp-init 20 --qp-max 22 --qp-min 18 --max-bitrate 5000 --aq --aq-temporal --gop-len 240 --lookahead 16 --slices 2 --multiref-l0 4 --multiref-l1 4 --strict-gop --nonrefp --weightp
craigpro is offline   Reply With Quote
Old 20th April 2020, 02:06   #345  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by craigpro View Post
These are the settings I'm using and my 720p encodes are always a few hundred MB larger than similar shows from the scene HEVC encodes. Does anyone have any suggestions for better settings please? Thank you.

--cqp 18:20:22 --codec h265 --preset quality --level 5.1 --output-depth 10 --qp-init 20 --qp-max 22 --qp-min 18 --max-bitrate 5000 --aq --aq-temporal --gop-len 240 --lookahead 16 --slices 2 --multiref-l0 4 --multiref-l1 4 --strict-gop --nonrefp --weightp
Having a QP range of just 18-22 seems like a really narrow band if your content is ever VBV constrained. 4 QP isn't enough for even a 2x difference in bitrate.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 20th April 2020, 05:38   #346  |  Link
RanmaCanada
Registered User
 
Join Date: May 2009
Posts: 328
Quote:
Originally Posted by craigpro View Post
These are the settings I'm using and my 720p encodes are always a few hundred MB larger than similar shows from the scene HEVC encodes. Does anyone have any suggestions for better settings please? Thank you.

--cqp 18:20:22 --codec h265 --preset quality --level 5.1 --output-depth 10 --qp-init 20 --qp-max 22 --qp-min 18 --max-bitrate 5000 --aq --aq-temporal --gop-len 240 --lookahead 16 --slices 2 --multiref-l0 4 --multiref-l1 4 --strict-gop --nonrefp --weightp
With current tech you will not get close to what software encoding can do. If you want better results, use a CPU. I know it's not what you want to hear, but it's the truth. GPU is quick and dirty.
RanmaCanada is offline   Reply With Quote
Old 22nd April 2020, 19:48   #347  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by RanmaCanada View Post
With current tech you will not get close to what software encoding can do. If you want better results, use a CPU. I know it's not what you want to hear, but it's the truth. GPU is quick and dirty.
Another way to frame that is that the quality @ perf for GPU is higher at very high speeds, but when encoding time/resources can be more, SW pulls ahead in quality as GPU encoding's peak quality is low.

GPU encoding is great for things like Twitch where it is recording screen activity while gaming. It takes minimal CPU away from the game, and is able to use the frames in the GPU as source without having to copy the pixels to main memory. Just writing 4Kp60 RGB frames to the CPU on top of everything else further stresse the memory bandwidth.

It's also good for very low power embedded solutions where a Tegra or Atom processor can be used with little CPU but with a HW encoder available.

There's been work done on hybrid encoders, which use the GPU for the first pass to be refined in software, which can speed up encoding 20-25%.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 29th November 2020, 00:41   #348  |  Link
Yups
Registered User
 
Join Date: Sep 2011
Posts: 362
I'm currently testing Quicksync on Iris Xe LP. Compared to Gen9.5 they have greatly improved the fixed function encoder, both quality and speed. On Gen9.5 graphics the fully fixed function encoder was more or less useless (no b-pyramid, no bframes), means the hybrid was the better choice and for H265 Gen9.5 didn't even support FF. On Iris Xe LP I don't really see a quality difference between hybrid and the fully fixed function (H265) encode, the difference is minor I would say but the speed difference is really huge. So at this point it makes the h265 hybrid somehow obsolete.
Yups is offline   Reply With Quote
Old 29th November 2020, 05:32   #349  |  Link
RanmaCanada
Registered User
 
Join Date: May 2009
Posts: 328
Quote:
Originally Posted by Yups View Post
I'm currently testing Quicksync on Iris Xe LP. Compared to Gen9.5 they have greatly improved the fixed function encoder, both quality and speed. On Gen9.5 graphics the fully fixed function encoder was more or less useless (no b-pyramid, no bframes), means the hybrid was the better choice and for H265 Gen9.5 didn't even support FF. On Iris Xe LP I don't really see a quality difference between hybrid and the fully fixed function (H265) encode, the difference is minor I would say but the speed difference is really huge. So at this point it makes the h265 hybrid somehow obsolete.
Are you using the Intel Media SDK? That is what is recommended to get the best encodes out of Quicksync. At least last time I checked.
RanmaCanada is offline   Reply With Quote
Old 29th November 2020, 11:51   #350  |  Link
Yups
Registered User
 
Join Date: Sep 2011
Posts: 362
Quote:
Originally Posted by RanmaCanada View Post
Are you using the Intel Media SDK? That is what is recommended to get the best encodes out of Quicksync. At least last time I checked.

As for the hardware encoder it doesn't matter, it just needs the graphics driver, it has all the required files. The Media SDK is required for the software encoder.
Yups is offline   Reply With Quote
Old 29th November 2020, 22:46   #351  |  Link
Yups
Registered User
 
Join Date: Sep 2011
Posts: 362
I finished my Tigerlake speed test and also some other results just for reference. QSVEnc/NVEnc (h265) used for the GPUs and Handbrake nightly for the x265 run. Test sample is a 1080p 24fps 25 Mbit video which is converted down to ~2500 Kbit.



Iris Xe LP 1100 Mhz GPU+fixed function
CBR slowest= 68 fps
CBR balanced= 131 fps
CBR fastest= 226 fps

Iris Xe LP 1100 Mhz fixed function
CBR slowest= 260 FPS
CBR balanced= 325 FPS
CBR fastest= 507 FPS


Iris Xe LP 1100 Mhz GPU+ fixed function
CQP slowest= 75 fps
CQP balanced= 144 fps
CQP fastest= 263 fps


Iris Xe LP 1100 Mhz fixed function
CQP slowest= 271 fps
CQP balanced= 344 fps
CQP fastest= 518 fps


For reference some other GPUs and x265 ultrafast:

HD 630 1150 Mhz GPU + fixed function
CBR slowest= 43 fps
CBR balanced= 100 fps
CBR fastest= 302 fps

i7-1165G7 2800 Mhz (Turbo disabled)
x265 ultrafast= 58 fps

GTX 1080 2000 Mhz
CBR quality= 317 fps
CBR fastest= 502 fps

Last edited by Yups; 29th November 2020 at 23:38.
Yups is offline   Reply With Quote
Old 3rd December 2020, 02:49   #352  |  Link
RanmaCanada
Registered User
 
Join Date: May 2009
Posts: 328
But what was the quality comparison to NVENC? Speeds mean nothing if the output is garbage (Looking at you AMD VCE) If you have not done a comparison, would it be possible to do one?
RanmaCanada is offline   Reply With Quote
Old 3rd December 2020, 13:53   #353  |  Link
Yups
Registered User
 
Join Date: Sep 2011
Posts: 362
In this speed test NVENC quality is worse than Iris Xe but I will use different settings when I go for quality on both, for example I didn't use lookahead on NVENC and keep in mind this isn't Turing. I will do a quality comparison later and also I have to test Turing which I don't have for now. I'm testing on Iris Xe mainly at the moment. As for Iris Xe CQP is the best bitrate control method when it comes to quality, better than ICQ and much better than CBR/VBR. CQP works very good with many bframes+bpyramid on Iris Xe unlike ICQ.
Yups is offline   Reply With Quote
Old 3rd December 2020, 21:49   #354  |  Link
RanmaCanada
Registered User
 
Join Date: May 2009
Posts: 328
Quote:
Originally Posted by Yups View Post
In this speed test NVENC quality is worse than Iris Xe but I will use different settings when I go for quality on both, for example I didn't use lookahead on NVENC and keep in mind this isn't Turing. I will do a quality comparison later and also I have to test Turing which I don't have for now. I'm testing on Iris Xe mainly at the moment. As for Iris Xe CQP is the best bitrate control method when it comes to quality, better than ICQ and much better than CBR/VBR. CQP works very good with many bframes+bpyramid on Iris Xe unlike ICQ.
Thanks!
RanmaCanada is offline   Reply With Quote
Old 5th December 2020, 01:06   #355  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by Yups View Post
In this speed test NVENC quality is worse than Iris Xe but I will use different settings when I go for quality on both, for example I didn't use lookahead on NVENC and keep in mind this isn't Turing. I will do a quality comparison later and also I have to test Turing which I don't have for now. I'm testing on Iris Xe mainly at the moment. As for Iris Xe CQP is the best bitrate control method when it comes to quality, better than ICQ and much better than CBR/VBR. CQP works very good with many bframes+bpyramid on Iris Xe unlike ICQ.
The Turing NVENC added B-frames, IIRC, so is quite a lot better.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 5th December 2020, 13:57   #356  |  Link
utack
Registered User
 
Join Date: Apr 2018
Posts: 63
I am honestly surprised how much praise nvenc gets "around the internet"
The H.264 encoder certainly isn't bad for a hardware encoder, but people comparing it to x264 medium or even slow need to tone it down a notch.

In a "non realtime" scenario with aq, slowest preset and full bframes it did not seem competitive to x264 medium preset in quality in some gaming footage I've tested
utack is offline   Reply With Quote
Old 6th December 2020, 02:22   #357  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 4,406
It is from comparing it to the previous hardware encoders, Turing was a big upgrade.

It is actually somewhat watchable now! Credit where it is due.
__________________
madVR options explained
Asmodian is offline   Reply With Quote
Old 6th December 2020, 02:35   #358  |  Link
aegisofrime
Registered User
 
Join Date: Apr 2009
Posts: 478
Does Ampere have any upgrades over Turing? I have a 3080 and 3060Ti coming, so I'm curious about this.
aegisofrime is offline   Reply With Quote
Old 6th December 2020, 16:39   #359  |  Link
foxyshadis
ангел смерти
 
foxyshadis's Avatar
 
Join Date: Nov 2004
Location: Lost
Posts: 9,558
Quote:
Originally Posted by aegisofrime View Post
Does Ampere have any upgrades over Turing? I have a 3080 and 3060Ti coming, so I'm curious about this.
Would you be willing to give the HEVC challenge a shot once they arrive? https://forum.doom9.org/showthread.php?t=175776 At least then we have a thing we can compare to. Staxrip can handle the encoding to nvenc quite well.
foxyshadis is offline   Reply With Quote
Old 6th December 2020, 16:46   #360  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,344
Quote:
Originally Posted by aegisofrime View Post
Does Ampere have any upgrades over Turing? I have a 3080 and 3060Ti coming, so I'm curious about this.
No, the encoder is the same. Only the decoder was enhanced to support AV1.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel is online now   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 21:29.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.