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. |
28th October 2019, 10:55 | #7141 | Link | |
ffx264/ffhevc author
Join Date: May 2007
Location: /dev/video0
Posts: 1,844
|
Quote:
In my encoding, I no longer use it as it's slow here too. --hme seems to run faster here, at least for me on a Core i7 7700K on Linux HME is a bit beneficial for 1080p. But it really shows its strength for 2160p and higher |
|
28th October 2019, 13:12 | #7142 | Link |
Pig on the wing
Join Date: Mar 2002
Location: Finland
Posts: 5,733
|
In this case, VMAF could be useful in determining if it's worth it or not. Just need to make sure that the compared streams are decoded frame accurately so I would probably encode a sample clip, mux to mkv and index and decode with DGDecNV tools.
__________________
And if the band you're in starts playing different tunes I'll see you on the dark side of the Moon... |
28th October 2019, 22:29 | #7143 | Link | |
Moderator
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,771
|
Quote:
If doing CRF, it'll just make the file larger. In general, those values will reduce compression efficiency of most content in most scenarios. |
|
1st November 2019, 06:50 | #7145 | Link | |
Pig on the wing
Join Date: Mar 2002
Location: Finland
Posts: 5,733
|
Quote:
If your only problem is that your computer becomes unusable, run the process at a lower priority. You can do that in command prompt by using "start /belownormal ...your encoder stuff here..."
__________________
And if the band you're in starts playing different tunes I'll see you on the dark side of the Moon... |
|
1st November 2019, 11:01 | #7146 | Link | |
Registered User
Join Date: Jun 2013
Posts: 32
|
Quote:
Furthermore, I use htop to see CPU utilization. This is my third run, and the exact same 4 threads are used. Does it imply that ffmpeg only runs on those same 2 cores (or 4 cores) over and over? I know next to nothing about CPU architecture & organization, but if it implies so, does it have a long term effect on those cores? I thought that the OS should shuffle the workload among cores (between run), still I want to confirm. Last edited by nghiabeo20; 1st November 2019 at 11:08. |
|
1st November 2019, 11:09 | #7147 | Link |
RipBot264 author
Join Date: May 2006
Location: Poland
Posts: 7,816
|
Why don't you try runing x265 in idle priority? What you do is not optimal.
__________________
Windows 7 Image Updater - SkyLake\KabyLake\CoffeLake\Ryzen Threadripper |
1st November 2019, 11:29 | #7148 | Link |
Pig on the wing
Join Date: Mar 2002
Location: Finland
Posts: 5,733
|
The OS scheduler does do that, unless you set the process to use only specific cores (=CPU affinity). But as I and Atak_Snajpera said, set the process priority lower (idle or just below normal) and let the encoder use whatever resources are available to it after other processes.
__________________
And if the band you're in starts playing different tunes I'll see you on the dark side of the Moon... |
1st November 2019, 11:51 | #7149 | Link | |
Artem S. Tashkinov
Join Date: Dec 2006
Posts: 345
|
Quote:
Linux: man taskset OS X: You're f*cked. |
|
1st November 2019, 18:21 | #7151 | Link | |||
Registered User
Join Date: Jun 2013
Posts: 32
|
Quote:
Quote:
Quote:
|
|||
3rd November 2019, 19:19 | #7153 | Link |
Registered User
Join Date: Jan 2019
Location: Russia
Posts: 105
|
I've done tests, 1080p with bitrate 22mb, x264 10bit (placebo, subme 11, bframes 8, etc) and x265 10bit (slow, no-rect, subme 4, bframes 8 and some other tweaks). In the end I get almost the same encoding speed. And x265 looks much better. I don't know why many people say that at high bitrate x264 beats x265.
Last edited by redbtn; 3rd November 2019 at 19:32. |
3rd November 2019, 23:01 | #7154 | Link | |
RipBot264 author
Join Date: May 2006
Location: Poland
Posts: 7,816
|
Quote:
__________________
Windows 7 Image Updater - SkyLake\KabyLake\CoffeLake\Ryzen Threadripper |
|
3rd November 2019, 23:31 | #7155 | Link |
Registered User
Join Date: Jan 2019
Location: Russia
Posts: 105
|
I know that 22 is very high for 1080p, I did it specially for test. After encode I upscaled FHD to UHD and compared them. I don't say that x265 MUCH better, but it significantly better at preserving small details. And I can definitely say it's not worse anyway.
|
4th November 2019, 20:45 | #7156 | Link |
Derek Prestegard IRL
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,989
|
It's definitely not impossible that x265 would produce better results than x264 at 22 Mbps for 1080p. I think it's fair to say that for MOST content the difference would be quite negligible, especially at the same encoding speed. Even if you can get a small improvement when encoding several times slower, is it really worth it?
In many use cases, probably not. But for some? Absolutely. |
4th November 2019, 22:53 | #7157 | Link | |
Registered User
Join Date: Jan 2019
Location: Russia
Posts: 105
|
Quote:
But IMO, if x265 better at low bitrates and better at high bitrates, I don't see why I should use x264 at any cases. |
|
5th November 2019, 09:30 | #7158 | Link |
Registered User
Join Date: Dec 2014
Posts: 666
|
I'm getting this warning. I'm trying to test the new frame duplication switch:
x265 [warning]: Frame-duplication require NAL HRD and VBV parameters. Disabling frame duplication How do I address this?
__________________
Asus ProArt Z790 - 13th Gen Intel i9 - RTX 3080 - DDR5 64GB Predator - LG OLED C9 - Yamaha A3030 - Windows 11 x64 - PotPlayerr - Lav - MadVR |
|
|