StaxRip

[JohnLai]

- cbrhq ? Any additional information?

The "HQ" means two pass encoding.

edit:
Source from SDK 7.1.9 in case people are skeptical
http://i.imgur.com/RCHXHa1.png

edit2:
I guess my Constant Quality deduction back when SDK7.0 was released ....is right on the money. From 7.1.9 itself.
Target quality: This mode is specified by setting rateControlMode to one of the
VBR modes and desired target quality in targetQuality . The range of this target
quality is 1 to 51, roughly corresponding to the range of possible QP values. In this
mode, the encoder tries to maintain constant quality for each frame, by allowing the
bitrate to vary subject to the bitrate parameters specified in maxBitRate and
averageBitRate. If maxBitRate and averageBitRate are not specified, the
encoder will use as many bits as needed to achieve the target quality. However, if
both parameters are set, they will form the upper bound on the actual bitrate. The
bitrate will become constrained, resulting in the desired target quality possibly not
being achieved.
In my previous method....I actually proposed to max both averageBitRate and maxBitRate to 17500 kbps which produced more or less the same result as it is not constrained by bitrate and is able to reach defined target quality. Hmm.......

[mcjordan]

Thanks, John.
And... hmm... today again:

NVEnc 3.04
To be able to output lossless even with HEVC.
It was restricted so as to be impossible.

[Magik Mark]

Stax,

New x265 CLIs:

--ssim-rd
--aq-motion
--multi-pass-opt-rps
--multi-pass-opt-analysis
--multi-pass-opt-distortion

[mcjordan]

Another day, another version:
NVEnc 3.05
To be able to output lossless even with HEVC.
It was restricted so as to be impossible*

* Updated to 3.05 (2017.01.11)
It was corrected again as it could not be fixed by 3.04...
I'm sorry I updated it repeatedly.

P.S. Fast test with new edition (StaxRip command line) - vbrhq vs cbrhq (in combination with simple avs r2376 Sharpen (0.2,0.2):
"%app:NVEncC%" --cbrhq 4428 --codec h265 --bframes 5 --ref 4 --lookahead 32 --qp-init 1 --aq --level 4.1 --output-buf 128 --cabac --cu-max 64 --device 1 --vbr-quality 1 --enable-ltr --fullrange -i "%script_file%" -o "%encoder_out_file%"
and same but --vbrhq (--vbr2).
The resulting image in vbr mode in comparision with cbrhq is slightly better (GTX960). On GTX1080 the image is constantly more blurry (may be because default SAO?). In all cases -> High lookahead value -> more blurred result. And all of this in static frame comparision. Playback of the final video doesn't contain significant differences (vs original source) or visible artefacts. John - advise me, please? Where I'm wrong (or better ;-) ?
A... and in my mind - Is it possible inclusion of hardware edge enhancement into NVEncC like AsVideo converter (by steps from 1 to 100; working on both GTX960 and 1080)?

[JohnLai]

Another day, another version:
NVEnc 3.05
To be able to output lossless even with HEVC.
It was restricted so as to be impossible*

* Updated to 3.05 (2017.01.11)
It was corrected again as it could not be fixed by 3.04...
I'm sorry I updated it repeatedly.

P.S. Fast test with new edition (StaxRip command line) - vbrhq vs cbrhq (in combination with simple avs r2376 Sharpen (0.2,0.2):
"%app:NVEncC%" --cbrhq 4428 --codec h265 --bframes 5 --ref 4 --lookahead 32 --qp-init 1 --aq --level 4.1 --output-buf 128 --cabac --cu-max 64 --device 1 --vbr-quality 1 --enable-ltr --fullrange -i "%script_file%" -o "%encoder_out_file%"
and same but --vbrhq (--vbr2).
The resulting image in vbr mode in comparision with cbrhq is slightly better (GTX960). On GTX1080 the image is constantly more blurry (may be because default SAO?). In all cases -> High lookahead value -> more blurred result. And all this in static frame comparision. Playback of the final video doesn't contain significant differences (vs original source) or visible artefacts. John - advise me, please? Where I'm wrong?

..............use unconstrained* VBR Constant Quality mode instead.

http://i.imgur.com/qaujIkT.jpg

Two pass encoding is pointless because there is "lookahead" rate control being used.

If you take a look on nvenc sdk list of presets, you will notice HQ (two pass encoding) mode is only used for low latency video (eg, videoconferencing)

EDIT:
Now...the problem with your command line for nvenc hevc encoding.
--bframes 5 = option useless because there is no b-frame support
--ref 4 = useless because nvenc hevc only make use of 1 single reference frame
--output-buffer 128 = if you use SSD, maybe 128mb. However, if you use HDD, better stick with sane value like 32mb or 64mb (matching HDD cache, you know, HDD has fast DRAM cache)
--cabac = -.- no need to specify this, HEVC use cabac by default
--cu-max = -.- NVENC encoding chip is limited to 32x32 max, just let it auto.
--enable-ltr = useless because there is only one active reference frame in DPB
--fullrange = give this one up....nvidia has color reproduction bug https://devtalk.nvidia.com/default/topic/958132/video-technologies/nvenc-hevc-with-full-range-colors-/

EDIT2: (rephrasing)
-> High lookahead value -> more blurred result? Are you analysing it based on PSNR?
First reason = SAO
Second reason = spatial adaptive quantization(SAQ). Probably SAQ algorithm decides to raise QP factor after analyzing more frames (higher lookahead value) and determine those frames are heavy in motion.
From SDK :
http://i.imgur.com/j3ANeDb.png

[mcjordan]

Yes, thank you very much, John. I know more of this things.
But strange - StaxRip log indicate that 64x64 cu-max is activated?! It's real or not (most likely)?
Encoding with NVEnc is a real pain - must guess all things into smoke on the water ;-)

[JohnLai]

Yes, thank you very much, John. I know more of this things.
But strange - StaxRip log indicate that 64x64 cu-max is activated?! It's real or not (most likely)?
Encoding with NVEnc is a real pain - must guess all things into smoke on the water ;-)

Nope....it is not.
If you specify 64x64, it automatically fallback to "auto"...which mean 32x32.
It is mentioned in SDK documentation and I also verify it through hevc bitstream analyzer.
http://i.imgur.com/xwIASap.png

[mcjordan]

John, thank you again for responsiveness and sorry for disturbing.
Things, a little bit, began to explain to me. Obviously I need to read more (SDK, I mean)
and especially - more carefully. Obviously, there is still much more to be done in development of NVEnc.
Pity, this is one of the fastest hardware encoders, that I know. I'll do some more tests, but generally, soon,
if Ryzen prove what I guess - I'll assemble such kind of configuration and will go on pure x265.
With AVX512 support, may be? (as rumor has ;-)

[JohnLai]

John, thank you again for responsiveness and sorry for disturbing.
Things, a little bit, began to explain to me. Obviously I need to read more (SDK, I mean)
and especially - more carefully. Obviously, there is still much more to be done in development of NVEnc.
Pity, this is one of the fastest hardware encoders, that I know. I'll do some more tests, but generally, soon,
if Ryzen prove what I guess - I'll assemble such kind of configuration and will go on pure x265.
With AVX512 support, may be? (as rumor has ;-)

Not sure about Zen AVX512, but so far, Zen can't do single cycle AVX and AVX2 256bit unlike Intel.
Zen needs two cycles (128bit each) to execute AVX/AVX2 instruction.

Anyway, back to topic....stax76.....have you add NVENC NPP resizers into the GUI and NPP library itself (maybe provide a link to NPP library as it is 37MB) to the test-build yet?
Note: I am still using v1.4.0.0-stable.....

[mcjordan]

John, NPP library (on Rigaya shareplaces / working like a charm in command line mode by the way /) -
https://onedrive.live.com/?cid=6bdd4375ac8933c6&id=6BDD4375AC8933C6%212293&lor=shortUrl
or
https://drive.google.com/drive/folders/0BzA4dIFteM2dS1ZUT1FjTnF3Q0E?usp=sharing

P.S. Concerning Zen - in this case the price is a determining factor...

[JohnLai]

John, NPP library (on Rigaya shareplaces / working like a charm in command line mode by the way /) -
https://onedrive.live.com/?cid=6bdd4375ac8933c6&id=6BDD4375AC8933C6%212293&lor=shortUrl
or
https://drive.google.com/drive/folders/0BzA4dIFteM2dS1ZUT1FjTnF3Q0E?usp=sharing

P.S. Concerning Zen - in this case the price is a determining factor...

I know about using command line.
Just wanna make thing easier for newbie with GUI implementation......

Or maybe implement the custom preset for unconstrained VBR Target Quality rate control method. If one ask me about UVBRTQ, I think it is quite similar to software encoder CRF method.


EDIT: I just tested staxrip testbuild StaxRip-x64-1.4.0.2-test, but apparently, ffms2 as source filter doesn't play well with nvencc
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Script
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

LoadPlugin("C:\Users\Anonymous\Desktop\Staxrip\Apps\Plugins\both\ffms2\ffms2.dll")
FFVideoSource("C:\Users\Anonymous\Desktop\test_track1_und.h264", cachefile = "L:\test_track1_und_temp\test_track1_und.ffindex")

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Script Properties
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

source frame count : 2184
source frame rate : 23.976024
source duration : 00:01:31.0910000
target frame count : 2184
target frame rate : 23.976024
target duration : 00:01:31.0910000

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Encoding using NVEncC 3.02 x64
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

@echo off
C:\Users\Anonymous\Desktop\Staxrip\Apps\NVEncC\NVEncC64.exe --cqp 20:26:30 --codec h265 --ref 6 --aq --level 5.1 --lookahead 32 -i L:\test_track1_und_temp\test_track1_und.avs -o L:\test_track1_und_temp\test_track1_und_out.h265

cmd.exe /C call "L:\test_track1_und_temp\test_track1_und_NVEncC.bat"

avs: invalid colorformat.
Failed to open input file.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Error Encoding using NVEncC 3.02 x64
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Encoding using NVEncC 3.02 x64 failed with exit code: 1 (0x1)

The exit code might be a system error code: STATUS_WAIT_1

The exit code might be a system error code: Incorrect function.

avs: invalid colorformat.
Failed to open input file.

Start: 10:00:45 PM
End: 10:00:45 PM
Duration: 00:00:00

Other source filter such as LSMASHSource works well. But I need FFVideoSource filter for exact timing.

EDIT: Turn out AviSynth+r2294 is the problem. Uninstall 2294 and install AviSynthPlus-r1576 somehow works.
EDIT2: Hmm...apparently AviSynth+_v0.1.0_r1825-MT works too. Source video is 10bit h264. R2172 doesn't work too.

[mcjordan]

Try these updates:

ffms2 have a new version ->
https://github.com/FFMS/ffms2/releases/download/2.23/ffms2-2.23.1-msvc.7z
Chagelog:
Updated FFmpeg APIs used (Daemon404)
Added new API that can read/write indexes to a memory buffer (Daemon404)
Added support for high bitdepth output in Avisynth+ (Myrsloik)
vapoursource: Add nfMakeLinear flag to better performance in most cases (Myrsloik)
Remove bad colorspace matrix guessing for YUV (Myrsloik)
Tell libavformat to discard packets we don't want (Daemon404)
Restore original FFCopyrightInfringement behavior (Myrsloik)
Make FFMS_Init threadsafe (Daemon404)
Support latest libav* APIs (Daemon404)

And -> latest Avisynth+ is r2380
http://www.mediafire.com/file/6pyic6yoto80vtq/avsplus-r2380.7z
Includes support for 10-bit formats & various color space conversions.

LSMASHSource -> r921
https://www.dropbox.com/sh/3i81ttxf028m1eh/AAABkQn4Y5w1k-toVhYLasmwa?dl=0

P.S. Specifically I use as source filter DGDecNV r2052 (2017.01.09 is latest)

[JohnLai]

StaxRip-x64-1.4.0.2-test uses ffms2 2.23.1
Latest Avisynth+ r2380 still have the same "avs: invalid colorformat." error.

LSMASHSource doesn't exactly work well for multichapters plus separate MKV content. Only FFVideoSource can properly index and load multichapter mkv files.
Example, load chapter1.mkv, then at 1:30, load OP.mkv, next at 3:20, load chapter2.mkv.
LSMASHSource and LWLibavVideoSource actually skipped the OP.mkv resulting in out of audio sync with video.

Not a big deal since older avisynth+ works fine.

[mcjordan]

Thank you for response.
I will do test to reproduce a problem, but tomorrow.
I had a quite hard day in work.

[pinterf]

StaxRip-x64-1.4.0.2-test uses ffms2 2.23.1
Latest Avisynth+ r2380 still have the same "avs: invalid colorformat." error.
Beginning with r2290 Avisynth+ has native 10 bit support. Your 10 bit video is probably opened as real 10 bit by ffms2. Thus the output is also 10 bits which is not handled well by the encoder at least not without appropriate parameters. Try ConvertBits(8) line before the end of the script, or right after opening the video, or try to use a 8 bit colorspace parameter directly with ffms2.

[Dark Eiri]

Are you planning to support HEVC thru AMD VCE on Polaris \ Vega GPUs soon?

[NikosD]

Are you planning to support HEVC thru AMD VCE on Polaris \ Vega GPUs soon?

http://forum.doom9.org/showthread.php?p=1790780

[Khun_Doug]

Using the latest 1.4.0.2 build to encode with X265. I have the latest encoder that supports the limit-tu option. But the test build of Stax appears to create a command string "--limit-tu level 2", but it should really be "--limit-tu 2". The word "level" needs to be removed from the command string.

PS - Sorry to read Stax76 is going to leave the project. This is surely going to be a loss sorely felt.

[dracore]

im new to all this and ive been trying to add lookahead for my encodes in staxrip .. could someone please help me how to .... thank you

[JohnLai]

im new to all this and ive been trying to add lookahead for my encodes in staxrip .. could someone please help me how to .... thank you

You didn't even mention which encoder.
If I didn't bother to check your post at videohelp forum, nobody will even bother to help you.

http://i.imgur.com/h84owC1.png



Or if you use latest Staxrip beta.
http://i.imgur.com/Y2HThzR.jpg

[drpepper]

Hey folks, I'm fairly new here but I already have a problem, haha.

So I was trying to encode a video that I created using Vegas Pro 13.0 with staxrip. But when I try to load my source file (in this case, the video file is about 350 GB) and choose AviSynth+ as source filter I get an error: "Failed to open source, try another filter? System exception - Access violation (C:\Users\Jan\Videos\2016_1_temp\2016_1_Source.avs, line 2)".

Now, this is peculiar because I cannot seem to find the cause of this problem. My AviSynth script is actually very simple and I didn't write it myself, it was generated automatically by staxrip.
LoadPlugin("C:\Users\Jan\Documents\Zeug\Programme\staxrip\Apps\Plugins\both\ffms2\ffms2.dll")
FFVideoSource("C:\Users\Jan\Videos\2016_1.avi", cachefile = "C:\Users\Jan\Videos\2016_1_temp\2016_1.ffindex")
ConvertToYV12()

I tried using other source filters, but none of them worked. This left me a little puzzled since I don't understand any of this. I don't know if this is relevant, but here are the MediaInfo of my source file:
General
Complete name : C:\Users\Jan\Videos\2016_1.avi
Format : AVI
Format/Info : Audio Video Interleave
Format profile : OpenDML
File size : 337 GiB
Duration : 58 min 11 s
Overall bit rate : 830 Mb/s
TCOD : 0
TCDO : 34916800000

Video
ID : 0
Format : YUV
Codec ID : UYVY
Codec ID/Info : Uncompressed 16bpp. YUV 4:2:2 (Y sample at every pixel, U and V sampled at every second pixel horizontally on each line). A macropixel contains 2 pixels in 1 u_int32.
Duration : 58 min 11 s
Bit rate : 829 Mb/s
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 25.000 FPS
Color space : YUV
Chroma subsampling : 4:2:2
Compression mode : Lossless
Bits/(Pixel*Frame) : 16.000
Stream size : 337 GiB (100%)

So, can anybody shed some light on this please :)?

[dracore]

You didn't even mention which encoder.
If I didn't bother to check your post at videohelp forum, nobody will even bother to help you.

http://i.imgur.com/h84owC1.png



Or if you use latest Staxrip beta.
http://i.imgur.com/Y2HThzR.jpg



thank you

[chummy]

I dont know if that is the best place to talk about this but since the noticed issue started with Youtube i would say here. A guy noticed than when uploading Nvenc source to Youtube the reencoding results were worse than feeding with x264 source. So i tried samething and find same problem. Then i decided to reencode at low bitrate with my own x264 software from both NVENC and x264 high bitrate sources which you will not notice quality difference and i see than samething happens, when reencoding at low bitrate like youtube (1080p@5mbit) NVENC source give worse result. With Quicksync happened samething. Is that something was already known?

[spon]

Hi everyone. JohnLai posted some recommended QSVEncC HW encoding settings for me a few months back that have been helpful. http://i.imgur.com/OaGdzPs.jpg

I just updated to the latest Staxrip and Rigaya QSVEncC and was wondering if any of the recommendations have changed. Thanks for any advice!

[Yups]

Hi everyone. JohnLai posted some recommended QSVEncC HW encoding settings for me a few months back that have been helpful. http://i.imgur.com/OaGdzPs.jpg

I just updated to the latest Staxrip and Rigaya QSVEncC and was wondering if any of the recommendations have changed. Thanks for any advice!



His recommendation is flawed. B-frames over 4 results in playback issues when converting with a low bitrate, TU4 and even more TU1 is dead slow under H265 because Intels GPU is too slow, it's a bad tradeoff. You may try TU4 which is slow, but forget about TU1. Ref frames 5 - never saw improvements in real world over 3. MBBR is enabled by default, no matter if you disable or enable it, it won't change anything. It runs enabled even if you try to disable it. ICQ Bitrate mode has lower quality than VBR, yes it has some benefits over VBR in some cases but it can be really poor in static scenes for example. Better go with VBR as a starting point. The non QSV decoder is much slower with more CPU use, only use it if you need scenechange (doesn't work with QSV decoding)

[JohnLai]

-.- Spon wants the best possible out of Intel HEVC encoding. Those options are the best possible options.

B-frame over 4 only problematic for smartphone or low end devices. Beside, Intel is using GPB, which mean, IBBBBBBBBBBBBBBBB, might as well as max it and leave it to 'scenechange' to insert new I-frame when it is necessary......
TU1 --> Remember 32x32 CU? Efficiency?
REF --> Real world, not really, in synthetic PSNR, there is an impact.
Oh, I think you are the one who provide me the HEVC sample last time for Intel. Technically, the reference top out* at 2 frames. The third reference frame only account for around 1-5%.

Example from your Intel HEVC sample last time with 3 reference frame. (Note, Frame number 40 is being analysed)
Reference pictures % Pixels
29 1.26 26240
33 39.80 831488
37 77.37 1616320

But, for animated content, high count of reference frames are very useful.

ICQ ---> which is why 'scenechange' enabled to minimize static scene artifact. Note, VBR is worse than ICQ in my test.
Decoder ---> That why I wrote only use QSVENC as decoder if one is certain the source file can be QSV-decoded.

[spon]

Yups & JohnLai -

Thank you both for your responses. It's interesting to see the differing opinions. I'm going to try them both and see which one looks better to my eyes.

One question - I can't seem to find the TU1/TU4 that you reference. Which menu is is located in - Basic, Slice, Rate Control, Motion Search, Porfile, Performance, Bitstream, VPP, VUI, Deinterlace or Other? Also, can the latest version of QSVEnnC encode 10bit? I saw a reference to it on Rigaya's site, but don't see anything in Staxrip.

Thanks again

[JohnLai]

It is at Basic --> Quality/Speed

Best = TU1
.
.
.
Fastest = TU7
You know the rest.

EDIT: Latest Qsvencc can encode HEVC* 10bit just fine....if you have kaby lake cpu......

[spon]

It is at Basic --> Quality/Speed

Best = TU1
.
.
.
Fastest = TU7
You know the rest.

EDIT: Latest Qsvencc can encode HEVC* 10bit just fine....if you have kaby lake cpu......

Ahhh ok, thanks John. I don't mind the slowness of the TU1 - it's still exponentially faster than Handbrake.

I have a skylake CPU, so probably no-go on 10but.

Thanks again

[NikosD]

@Stax76

Hello Frank.

A brand new VCEEncC v3.01 HW encoder for AMD is here:
http://rigaya34589.blog135.fc2.com/blog-entry-891.html?sp

It has a few new options for H.264 and it's the first one with H.265 HW encoding support for Polaris cards and newer.

Rigaya has also added some other useful features like HW decoding, --check-features etc

[Yups]

.

B-frame over 4 only problematic for smartphone or low end devices.


I'm running a i7-7700k, would you call it a low end device? B-frames over 4 requires some amount of minimum bitrate, a 1080p video with 3 Mbps isn't going to work. It's clear you didn't test anything in real world. And ICQ is below VBR in image quality, confirmed by Intel (and me).

[JohnLai]

I'm running a i7-7700k, would you call it a low end device? B-frames over 4 requires some amount of minimum bitrate, a 1080p video with 3 Mbps isn't going to work. It's clear you didn't test anything in real world. And ICQ is below VBR in image quality, confirmed by Intel (and me).

Duh.....I don't have skylake/kaby based system, being saying this since forever. The only QSV HEVC sample I have and tested was from you.

Beside, I can only test lotta B-frames with Haswell H264 (IPBBBBBBBBBBBBBBB). I am certain ICQ is better than VBR for H264 from PSNR and MSSIM analysis. Some old smartphone has problem decoding H264 with a lot of B-frames, but generally, PC discrete GPU can decode such video just fine.

I assume it would be the same for intel HEVC.

Anyway, I had been requesting QSV HEVC sample with many b-frames and refs from you for some time now......

[NikosD]

ICQ for H.264 is A LOT better than VBR.

No idea regarding HEVC though.

After lots of test with Haswell, my opinion is that plain ICQ is the best mode for Haswell H.264 HW encoding.

[Magik Mark]

Guys,

Which among the resizer is the best for 2160p->1080p?

[JohnLai]

Guys,

Which among the resizer is the best for 2160p->1080p?

Downscaling?
Stick with avisynth default BicubicResize.

BicubicResize default value parameters b (0) and c (0.5) = Cubic-Catmull-Rom spline.

Hmmm.....for you to learn more about b and c as well as resampling filter.....you must visit http://www.imagemagick.org/Usage/filter/

I propose you to read from the beginning. Critical reading for your question is at Mitchell-Netravali Filter section

*Note: don't skip the reading at imagemagick site. Some links at the site had become rotten due to time passing.......May the information can be a help for your question. :helpful:

[Magik Mark]

Thanks john! That's what I did. I used the default settings in staxrip. Do you happen to know as well if the default setting extends to retaining the original aspect ratio?

[JohnLai]

Thanks john! That's what I did. I used the default settings in staxrip. Do you happen to know as well if the default setting extends to retaining the original aspect ratio?

:confused:

I thought staxrip has Resize's DAR, SAR and PAR settings plus Output Mod for retaining original aspect ratio right next to the "Filters"?

[Magik Mark]

Yeah, there is. Not sure how to interpret it. Dunno if the values there indicate same aspect ratio

[JohnLai]

Yeah, there is. Not sure how to interpret it. Dunno if the values there indicate same aspect ratio

As you move around the resize slider and selecting the resolution, make sure the DAR,PAR,SAR match the original source DAR,PAR,SAR

One can click on PAR for more selection:
http://i.imgur.com/h64b5jY.png

Back in the day when H264 just made it first appearance on satellite/terrestrial broadcast, some broadcaster broadcasted video with frame size 1440x1080 (for bandwidth cost reason), but for video to be properly displayed, it need to be stretched to 1920x1080. Kinda similar method for anamorphic DVD-Video too.

I remembered I screwed up with my DVD archives after resizing without noticing the DAR. The good old day of XVID encoding......

[Magik Mark]

Thanks John :helpful:

[Magik Mark]

JohnLai,

Can't believe you're still doing GPU encoding. I gave up on that. It could only do so much. Aren't you moving to CPU based? Quality is a lot better and it has more options to tweak

[JohnLai]

JohnLai,

Can't believe you're still doing GPU encoding. I gave up on that. It could only do so much. Aren't you moving to CPU based? Quality is a lot better and it has more options to tweak

True, but I simply don't have the time to wait......
plus those Intel HEDT or Xeon multi socket stuff costs a bomb for me.
I only earn USD 451.31 (after conversion from my nation currency) per month. ~.~""
Let see if AMD Zen can change everything.

x265 = 1fps with 3570k 4.4ghz....not exactly good idea....
NVENC HEVC = 150 - 180 fps
Intel H264 = 300 - 400fps

As long as hardware encoder quality is 'acceptable'.

[Magik Mark]

In ebay john there's a lot of used xeon. That's where I got mine

[Magik Mark]

Is there a way to convert bt2020 to bt709 color matrix? That's 4kuhd to 1080p

[JohnLai]

In ebay john there's a lot of used xeon. That's where I got mine

Used Xeon still unaffordable for me. ~.~

Is there a way to convert bt2020 to bt709 color matrix? That's 4kuhd to 1080p

Note: I never try this before.
http://avisynth.nl/index.php/Dither_tool

Dither_convert_yuv_to_rgb
then
Dither_convert_rgb_to_yuv

The string matrix (undefined) for both dither_convert

Convert BT2020 to rgb, then convert rgb to BT709.

Not sure if this one works.

[Zetto]

@JohnLai, since you appear to be an expert on using nvenc, could you please post a guide of sorts? I'm sure a lot of people would appreciate an explanation of different options for nvenc encoding, pluses and minuses. I tried searching for something of the kind to no avail.

[JohnLai]

@JohnLai, since you appear to be an expert on using nvenc, could you please post a guide of sorts? I'm sure a lot of people would appreciate an explanation of different options for nvenc encoding, pluses and minuses. I tried searching for something of the kind to no avail.

All big three hardware encoders info is available from their own published SDKs.

For instance:
Info on Lookahead, Spatial AQ, Temporal AQ, 10bit encoding can be found at Chapter 8 of https://developer.nvidia.com/nvenc-programming-guide

Some info on nvidia two-pass and other rate control:
http://on-demand.gputechconf.com/gtc/2014/presentations/S4654-detailed-overview-nvenc-encoder-api.pdf

Nvidia NVENC architecture back in kepler day.
http://on-demand.gputechconf.com/gtc/2013/presentations/S3379-High-Performance-GPU-Video-Encoding.pdf

EDIT: I am still waiting for nvidia new nvenc hardware revision for proper HEVC B-frames. I mean, back in 2015, nvidia said HEVC B-frame support was in their roadmap.....
http://i.imgur.com/nk8FcSv.png
With Pascal.....nvidia had implemented motion estimation only, SAO, 10bit encode, lossless, new rate control, 8K, AQ
Missing part --> 12bit HEVC encoding and HEVC B-frames.
I wish for b-pyramid support too.......

[Zetto]

I'm not that proficient to wade through those technical documents from nvidia. I'm looking more for a guide in layman's terms, so that anyone could gain at least a basic understanding what nvenc does. Right now I'm just blindly wandering through it, poking at different options with little to no understanding what they really do.

[Pengaban]

I'm wondering why on the latest Stax1.4.0.0 with NVEnc 3.02 I cant find the deinterlace function in Encoder Options?
The info "help.txt" for 3.02 lists the deinterlace function.
Tried to use the commandline "--vpp-deinterlace adaptive" only gives me NVEnc errors.

Is this a known bug? Or did I broke something on my setup?
I'm using 376.06 NV driver on Win7 x64 with a GK106.

With Stax1.3.4.0 with NVEnc 2.04 "--vpp-deinterlace adaptive" works as supposed.


@JohnLai
For me another advantage of GPU(ASIC) encoding over the CPU is power draw. Would be interesting to see how much watt a Xeon CPU eats up in comparison to NVEnc. And switching a GPU takes 10 minutes including Driver Update. In contrast trying to change a motherboard would take hours on the minimum(days are more realistic) and probably needs to clean reinstall the OS also various driver Update sessions and so on. Not talking about the money side. The cheapest current gen NV/AMD encoding GPU 100,-?

[JohnLai]

I'm wondering why on the latest Stax1.4.0.0 with NVEnc 3.02 I cant find the deinterlace function in Encoder Options?
The info "help.txt" for 3.02 lists the deinterlace function.
Tried to use the commandline "--vpp-deinterlace adaptive" only gives me NVEnc errors.

Is this a known bug? Or did I broke something on my setup?
I'm using 376.06 NV driver on Win7 x64 with a GK106.

With Stax1.3.4.0 with NVEnc 2.04 "--vpp-deinterlace adaptive" works as supposed.


@JohnLai
For me another advantage of GPU(ASIC) encoding over the CPU is power draw. Would be interesting to see how much watt a Xeon CPU eats up in comparison to NVEnc. And switching a GPU takes 10 minutes including Driver Update. In contrast trying to change a motherboard would take hours on the minimum(days are more realistic) and probably needs to clean reinstall the OS also various driver Update sessions and so on. Not talking about the money side. The cheapest current gen NV/AMD encoding GPU 100,-?

"--vpp-deinterlace adaptive" option only works IF the decoder is set to use 'avcuvid'. Now here the problem, the video has to be hardware decoded by GPU.

This means if your video is 10bit H264, then --vpp-deinterlace will never work as avcuvid decoder can't decode the video in hardware (Nvidia PureVideo hardware limitation, never see any 10bit H264 hardware decoder in the market).

https://s88.imgup.net/NVENCf7d2.PNG
https://m53.imgup.net/NVENC2f2a6.PNG