Re: mpv and estd and --vo=xv

2021-07-18 Thread Rhialto
On Wed 14 Jul 2021 at 21:38:51 +, RVP wrote: > 5. Make sure `hwdec=auto' is set in ~/.config/mpv/mpv.conf, and >estd is turned off. On my machine with Radeon card, I think this was the essential step that made it work for me. I got "Using hardware decoding (vaapi-copy)." when combined

Re: mpv and estd and --vo=xv

2021-07-15 Thread RVP
On Thu, 15 Jul 2021, Rhialto wrote: On my 10 years old laptop, which has Intel graphics, which were accellerated until (I think) I installed NetBSD 9, I found that the best way to run is with the intel (not modesetting) driver, and have LIBGL_ALWAYS_SOFTWARE=1. That is because the intel driver

Re: mpv and estd and --vo=xv

2021-07-15 Thread RVP
On Thu, 15 Jul 2021, nia wrote: Well, Mesa is part of native Xorg, (/usr/X11R7/lib/libGL.so.3), etc. You can use Mesa from pkgsrc as part of modular Xorg or standalone, but the standard binary packages aren't built for that use case. OK, but, MesaLib still shouldn't need 2 API versions of

Re: mpv and estd and --vo=xv

2021-07-15 Thread Rhialto
On Wed 14 Jul 2021 at 21:38:51 +, RVP wrote: > Getting back to the original topic, I've pulled together all the > things I needed to get mpv video playback on NetBSD to be (almost) > on par with Linux/FreeBSD. See if these help you (assuming the OS > ver. is 9.99.86 and the card is Intel): On

Re: mpv and estd and --vo=xv

2021-07-15 Thread RVP
On Wed, 14 Jul 2021, RVP wrote: On Wed, 14 Jul 2021, nia wrote: On Wed, Jul 14, 2021 at 09:38:51PM +, RVP wrote: 2. Create symlinks missing on 9.99.8x (only if using binary packages): (cd /usr/lib; ln -s libstdc++.so.9 libstdc++.so.7; cd /usr/X11R7/lib; ln -s libglapi.so.1

Re: mpv and estd and --vo=xv

2021-07-15 Thread nia
On Thu, Jul 15, 2021 at 07:05:33AM +, RVP wrote: > On Thu, 15 Jul 2021, nia wrote: > > > Are you mixing native and pkgsrc xorg? :) > > > > Nope. Not a modular in sight... > > -RVP Well, Mesa is part of native Xorg, (/usr/X11R7/lib/libGL.so.3), etc. You can use Mesa from pkgsrc as part of

Re: mpv and estd and --vo=xv

2021-07-15 Thread RVP
On Thu, 15 Jul 2021, nia wrote: Are you mixing native and pkgsrc xorg? :) Nope. Not a modular in sight... -RVP

Re: mpv and estd and --vo=xv

2021-07-15 Thread nia
On Thu, Jul 15, 2021 at 06:05:46AM +, RVP wrote: > On Wed, 14 Jul 2021, RVP wrote: > > > On Wed, 14 Jul 2021, nia wrote: > > > > > On Wed, Jul 14, 2021 at 09:38:51PM +, RVP wrote: > > > > 2. Create symlinks missing on 9.99.8x (only if using binary packages): > > > > > > > >(cd

Re: mpv and estd and --vo=xv

2021-07-14 Thread RVP
On Wed, 14 Jul 2021, nia wrote: On Wed, Jul 14, 2021 at 09:38:51PM +, RVP wrote: 2. Create symlinks missing on 9.99.8x (only if using binary packages): (cd /usr/lib; ln -s libstdc++.so.9 libstdc++.so.7; cd /usr/X11R7/lib; ln -s libglapi.so.1 libglapi.so.0) FYI, likely this will

Re: mpv and estd and --vo=xv

2021-07-14 Thread RVP
On Wed, 14 Jul 2021, Rhialto wrote: On the other hand, from there I found this description (I selected my cpu, https://ark.intel.com/content/www/us/en/ark/products/134896/intel-core-i5-9600k-processor-9m-cache-up-to-4-60-ghz.html and as explanation for Intel Turbo Boost Technology it has

Re: mpv and estd and --vo=xv

2021-07-14 Thread RVP
On Wed, 14 Jul 2021, Michael van Elst wrote: r...@sdf.org (RVP) writes: But, maybe, MHz xxx1 != Turbo Boost mode. I'm very confused now :) It's Turbo Boost mode, which mostly says that the CPU runs at xxx0 MHz but may increase that as long as the thermal limits aren't reached. You use it as

Re: mpv and estd and --vo=xv

2021-07-14 Thread nia
On Wed, Jul 14, 2021 at 09:38:51PM +, RVP wrote: > 2. Create symlinks missing on 9.99.8x (only if using binary packages): > >(cd /usr/lib; ln -s libstdc++.so.9 libstdc++.so.7; > cd /usr/X11R7/lib; ln -s libglapi.so.1 libglapi.so.0) FYI, likely this will break things in mysterious

Re: mpv and estd and --vo=xv

2021-07-14 Thread Rhialto
On Tue 13 Jul 2021 at 23:15:43 +, RVP wrote: > On Tue, 13 Jul 2021, Mike Pumford wrote: > > > But boost only scales up from the top clock speed to above the top speed > > it cant scale down or at least it can't on older generations of > > processor. > > > > You're right about this: Turbo

Re: mpv and estd and --vo=xv

2021-07-14 Thread Rhialto
On Tue 13 Jul 2021 at 23:15:43 +, RVP wrote: > https://ark.intel.com/content/www/us/en/ark/search/featurefilter.html?productType=873 > Then select the "blab blah 3.0 Freq." filter and set a Min. freq. I was hoping that that page would offer a "Speed Shift" filter, to find out if my xxx1

Re: mpv and estd and --vo=xv

2021-07-13 Thread Michael van Elst
r...@sdf.org (RVP) writes: >But, maybe, MHz xxx1 != Turbo Boost mode. I'm very confused now :) It's Turbo Boost mode, which mostly says that the CPU runs at xxx0 MHz but may increase that as long as the thermal limits aren't reached. You use it as an additional ACPI p-state with an unspecified

Re: mpv and estd and --vo=xv

2021-07-13 Thread Michael van Elst
r...@sdf.org (RVP) writes: >It has Enhanced SpeedStep though. So it looks like I have to >re-enable estd--except that the latest estd detects my Intel CPU >as a (ARM?) Rockchip: >--- >$ estd -f >Supported frequencies (Rockchip Mode): That says that estd uses sysctl machdep.cpu.frequency.* to

Re: mpv and estd and --vo=xv

2021-07-13 Thread RVP
On Tue, 13 Jul 2021, Mike Pumford wrote: But boost only scales up from the top clock speed to above the top speed it cant scale down or at least it can't on older generations of processor. You're right about this: Turbo Boost seems to go from the Processor Base Frequency to Max Turbo

Re: mpv and estd and --vo=xv

2021-07-13 Thread RVP
On Tue, 13 Jul 2021, Mike Pumford wrote: But boost only scales up from the top clock speed to above the top speed it cant scale down or at least it can't on older generations of processor. Intel says (same Turbo Boost 2.0 page): Maximum turbo frequency indicates the highest possible

Re: mpv and estd and --vo=xv

2021-07-13 Thread Mike Pumford
On 13/07/2021 22:41, RVP wrote:     When the processor is operating below these limits and the user's     workload demands additional performance, the processor frequency will     dynamically increase until the upper limit of frequency is reached. But boost only scales up from the top

Re: mpv and estd and --vo=xv

2021-07-13 Thread RVP
On Tue, 13 Jul 2021, Mike Pumford wrote: Ah thats good info that I DIDN'T know. Sadly all my BSD hardware is from the generations before the full 'auto adjust' was added. I think that's intel core gen 6 and mine is gen5. So in my case I really do need estd :( On older hardware the xxx1

Re: mpv and estd and --vo=xv

2021-07-13 Thread Mike Pumford
On 11/07/2021 20:59, RVP wrote: On Sun, 11 Jul 2021, Rhialto wrote: I also use estd to dymamically throttle down the cpu freqency when the system is not so busy. So most of the time it is set to 800 MHz, the lowest possible value. From what nia@ tells me (and this is also in the guide:

Re: mpv and estd and --vo=xv

2021-07-11 Thread RVP
On Sun, 11 Jul 2021, RVP wrote: GPU-acceleration does not necessarily imply HW-assisted video decoding. This latter depends heavily on the decoding capabilities of one's CPU and the video being played. Some CPUs don't do H.265. I forgot to mention this: it also depends on the decoding

Re: mpv and estd and --vo=xv

2021-07-11 Thread RVP
On Sun, 11 Jul 2021, Rhialto wrote: It doesn't tell me that it does, and I've never seen any indication of it, so this is indeed as I expected. If you have FreeBSD (or Linux) installed, you can use the same `I' key to see what HW-assisted video decoding looks like. Keep `top' running, then

Re: mpv and estd and --vo=xv

2021-07-11 Thread RVP
On Sun, 11 Jul 2021, Rhialto wrote: If I play a much "harder" video, say H.265 at 1920x1080, the default output from mpv gets way behind and starts to drop lots of frames. The default output is "sdl", which I think uses MesaLib/GL. Which is claimed to be accellerated: (output from glxinfo)

Re: mpv and estd and --vo=xv

2021-07-11 Thread RVP
On Sun, 11 Jul 2021, Rhialto wrote: I also use estd to dymamically throttle down the cpu freqency when the system is not so busy. So most of the time it is set to 800 MHz, the lowest possible value. From what nia@ tells me (and this is also in the guide: section 11.1.4.), you shouldn't

Re: mpv and estd and --vo=xv

2021-07-11 Thread Rhialto
On Sun 11 Jul 2021 at 20:13:08 +, RVP wrote: > GPU-acceleration does not necessarily imply HW-assisted video decoding. I'm not really expecting that. With XVideo, as I understand it, the cpu has to do the decoding. XVideo just scales and puts the image in a video overlay. But still, this

Re: mpv and estd and --vo=xv

2021-07-11 Thread Rhialto
On Sun 11 Jul 2021 at 14:09:45 -, Michael van Elst wrote: > In the best case, the CPU is still decoding the input stream and > filling the display pipeline. If the CPU is too slow, or the decoder > latency cannot be anticipated, the decoded results will come too > late and frames get dropped.

Re: mpv and estd and --vo=xv

2021-07-11 Thread Mike Pumford
On 11/07/2021 14:13, Rhialto wrote: I keep having weird things with graphics performance. I got myself a new box ("Intel(R) Core(TM) i5-9600K CPU @ 3.70GHz") with 6 cores. I got a "new" old Radeon HD-5450 because at least that chipset is supported for dri/drm. I also use estd to dymamically

Re: mpv and estd and --vo=xv

2021-07-11 Thread Michael van Elst
rhia...@falu.nl (Rhialto) writes: >How can this be? If it's GPU-accellerated, this should be independent of >the CPU frequency, I would think? In the best case, the CPU is still decoding the input stream and filling the display pipeline. If the CPU is too slow, or the decoder latency cannot be

mpv and estd and --vo=xv

2021-07-11 Thread Rhialto
I keep having weird things with graphics performance. I got myself a new box ("Intel(R) Core(TM) i5-9600K CPU @ 3.70GHz") with 6 cores. I got a "new" old Radeon HD-5450 because at least that chipset is supported for dri/drm. I also use estd to dymamically throttle down the cpu freqency when the