Media Player Classic - Home Cinema (MPC-HC) - v1.7.13 [Archive] - Page 7

Armada

17th March 2013, 21:53

Does version 1.6.6 correct as planned the subtitle language advanced selection provided by lav / haali splitter broken after mpc-hc 1.6.6.6500 ?
Yes, extra options regarding that have been added to Subtitles > Misc, but the default should work.

chano22000

19th March 2013, 11:16

Yes, extra options regarding that have been added to Subtitles > Misc, but the default should work.

I have just updated with the latest build (1.6.7.6967) from 1.6.6.6500 (the latest before the bug) and made a quick check.

The advanced subtitles selection as handled by haali or lav splitter seems to work again, but the audio selection menu is now greyed out (just above the subtitle selection menu available in full screen mode) as if there is only one audio track even though several are available for selection. Shortcuts still work to get next/previous audio track.

Apparently a GUI issue.

Thanks again for your fantastic collaborative work.

clsid

19th March 2013, 16:02

Embedded audio tracks can be selected in the Navigate menu. It has always been that way. Showing those tracks in the Play menu is still on the ToDo list.

vivan

19th March 2013, 18:02

Feature request: support for "wildcard character" in Autoload paths (Subtitles -> Misc).
It's useful if you have several folders with subtitles. E.g. I have different translations in folders with name like "Subs [somebody]" and want to add folder "Subs *" to autoload path once and for all.

Armada

19th March 2013, 20:59

jkauff

20th March 2013, 03:44

Search didn't turn up anything, has there been any information posted about the pixel shaders listed in the right-click menu? I'd like to know what each one does, and when they might be appropriate to use.

EDIT: I realize these shaders are enabled by madVR, but if I understand the issue correctly the choice of what shaders to offer is up to the player developers.

Dodgexander

21st March 2013, 03:43

Check the pixel shader pack thread in the same sub forum as this one.

You can add to madvr even.

Sent from my Blade S using Tapatalk 2

pankov

21st March 2013, 11:59

Check the pixel shader pack thread in the same sub forum as this one.

You can add to madvr even.

Sent from my Blade S using Tapatalk 2

Would you mind sharing a link?
I did a quick look and didn't see it. Are you sure it's in this subforum?

LigH

21st March 2013, 12:44

You did not try entering "pixel shader pack" into the forum search, did you?

Video pixel shader pack (http://forum.doom9.org/showthread.php?t=157634&highlight=pixel+shader+pack)

jkauff

21st March 2013, 17:19

The information is in that thread, but it's outdated now that madVR has incorporated some of the routines. It would be good to know why the MPC-HC devs chose the particular 22 shaders they did (some seem redundant with madVR settings), along with a short explanation of when it might be appropriate to use them. Some, like Emboss, seem to be included just because they exist rather than being truly useful.

clsid

21st March 2013, 18:10

Shaders are not just for use with madVR, but also for other (less advanced) renderers.

wanezhiling

22nd March 2013, 02:30

http://forum.doom9.org/showpost.php?p=1620759&postcount=14526
jitter in dxva mode.
work fine in sw mode (MPC MPEG Video Decoder).

latest build 6978

pankov

22nd March 2013, 09:32

You did not try entering "pixel shader pack" into the forum search, did you?

Video pixel shader pack (http://forum.doom9.org/showthread.php?t=157634&highlight=pixel+shader+pack)

You are absolutely correct - I didn't try it because I was browsing from the phone and didn't have the time.

Now that I'm on the PC I did see it using either your link or the search function but I find it very strange that even now I can't seem to find it while simply browsing this subforum. I see a list of all topics and it's not there !?!?!? It says "Showing threads 1 to 23 of 23" and the one for the shaders pack is missing !?!?
Is it only me or there is a bug in the forum software?

romulous

22nd March 2013, 09:49

Is it only me or there is a bug in the forum software?

Down the bottom of the forum (just under where it shows 'showing threads 1 to 23 of 23'), change the 'from the' field from 'last month' to 'last 2 months'. Then, check the second page of threads - the relevant thread is the first thread on the second page :) As the last reply in that thread is over a month ago, it does not show in the thread list by default (as 'from the' defaults to 1 month).

For anyone wondering why the sudden interest in the pixel shaders, Blight is looking to add support for custom pixel shaders to Zoom Player (for use with madVR), and has asked for comment from the community on the design, and what pixel shaders to include etc

judelaw

29th March 2013, 12:44

monitor (pri) + TV (ext)
mpc-hc on TV, play
Win+P - disable monitor
player hangs, but not crash (if return monitor (win+P -enable monitor), playback restores already on the monitor)
EVR - good, EVR CP - bad, VMR* - not good, SYNC renderer, madVR - good

it seems the problem appeared after r1050 (1048?) (4 years ago) :) 1043 - good

chano22000

29th March 2013, 19:12

Hi,

I updated from MPC-HC.1.6.7.6967 to the latest available built (MPC-HC.1.6.7.7010).

When in D3D fullscreen (Windows XP - EVR/CP) and moving the mouse down, the navigation bar appears and is clickable but no cursor appears anymore which now makes the feature unusable.

I suspect this bug might be related to the recent "Replace "On top > Never" by "On top > Default" committed March 24th, 2013.

Best regards.

Dodgexander

30th March 2013, 05:41

Down the bottom of the forum (just under where it shows 'showing threads 1 to 23 of 23'), change the 'from the' field from 'last month' to 'last 2 months'. Then, check the second page of threads - the relevant thread is the first thread on the second page :) As the last reply in that thread is over a month ago, it does not show in the thread list by default (as 'from the' defaults to 1 month).

For anyone wondering why the sudden interest in the pixel shaders, Blight is looking to add support for custom pixel shaders to Zoom Player (for use with madVR), and has asked for comment from the community on the design, and what pixel shaders to include etc

But you can already add custom pixel shaders in mpc-hc

Sent from my Blade S using Tapatalk 2

romulous

30th March 2013, 06:13

But you can already add custom pixel shaders in mpc-hc

As I mentioned in my post, we are talking about Zoom Player, not MPC.

nikusor665

31st March 2013, 08:53

I'm having a problem with mpc-hc

It plays all the audio tracks of a movie file at once , I cant select only one of them under the Filters menu, like I did with earlier versions of mpc-hc.

I even tried installing the older versions, but the problem persists.

What can I do ?

Skibicki

31st March 2013, 13:25

nikusor665, enable audio switcher in internal filters

nikusor665

31st March 2013, 16:26

Ok that seems to do the trick.

However, I'm using and have used (an older version of) ffdshow for its wonderful normalize feature and the audio switcher has its own not so good normalize feature and some other stuff that I do not enable of course, but running both internal audio switcher and ffdshow is that theoretically ok ? Cause if I dont enable audio switcher, ffdshow works fine on its own where there's only one audio track.

jmartinr

31st March 2013, 16:50

wanezhiling

31st March 2013, 17:14

Did anyone notice latest mpc-hc encounter corrupt image in dxva mode on H.264 ts files, especially AMD.

vBm

31st March 2013, 17:21

Did anyone notice latest mpc-hc encounter corrupt image in dxva mode on H.264 ts files, especially AMD.

Yes, and that commit (https://github.com/mpc-hc/mpc-hc/commit/6d51fc694ab79ea51e530ba7eeab7ea3a548e9a8) has been reverted for now.

It was v0lt's commit with help of Aleksoid1978 for fixing playback on Intel HD.

Dodgexander

31st March 2013, 20:40

As I mentioned in my post, we are talking about Zoom Player, not MPC.

Read the thread title? I know that's what a mod would say ;D

---

Since updating to the latest version the mpc-hc.exe process often(but not always) doesn't end when I close the program. Anyone else with this problem?

Sent from my Blade S using Tapatalk 2

JanWillem32

1st April 2013, 00:45

Since updating to the latest version the mpc-hc.exe process often(but not always) doesn't end when I close the program. Anyone else with this problem?Is this only when the internal subtitle rendererer was used? The subtitle renderer shares a badly designed thread lock with the main thread. When closing it can stall a very long time. Especially in debug builds it can take over a minute (without actually faulting). Though this particular lock has been in there for a long time, thread locking issues may have been aggravated by recent revisions.

romulous

1st April 2013, 01:52

Read the thread title? I know that's what a mod would say ;D

Everyone who has posted about the shaders here knows this is a MPC-HC topic. The reason why the question was posted here was that the MPC-HC shaders are ones that Blight is considering using in Zoom - so, the obvious place to ask 'what do these particular shaders do' (which is essentially what was asked), is here in the MPC-HC topic. So, not O/T at all.

truexfan81

1st April 2013, 04:25

Read the thread title? I know that's what a mod would say ;D

---

Since updating to the latest version the mpc-hc.exe process often(but not always) doesn't end when I close the program. Anyone else with this problem?

Sent from my Blade S using Tapatalk 2

Is this only when the internal subtitle rendererer was used? The subtitle renderer shares a badly designed thread lock with the main thread. When closing it can stall a very long time. Especially in debug builds it can take over a minute (without actually faulting). Though this particular lock has been in there for a long time, thread locking issues may have been aggravated by recent revisions.

i have this problem also, and i do not use subtitles, mine does it every 3 or 4 times i exit

Dodgexander

1st April 2013, 05:41

No, afraid I don't use subtitles.
This happened only since the latest update on two pc's.

LigH

1st April 2013, 18:03

http://gcc.gnu.org/gcc-4.8/

GCC 4.8.0 released.

WARNING

Changelog: (http://gcc.gnu.org/gcc-4.8/changes.html)
GCC now uses a more aggressive analysis to derive an upper bound for the number of iterations of loops using constraints imposed by language standards. This may cause non-conforming programs to no longer work as expected, such as SPEC CPU 2006 464.h264ref and 416.gamess. A new option, -fno-aggressive-loop-optimizations, was added to disable this aggressive analysis. In some loops that have known constant number of iterations, but undefined behavior is known to occur in the loop before reaching or during the last iteration, GCC will warn about the undefined behavior in the loop instead of deriving lower upper bound of the number of iterations for the loop. The warning can be disabled with -Wno-aggressive-loop-optimizations.

:o A compiler which fails compiling the SPEC benchmark suite correctly without specific switches was released. :stupid:

nevcairiel

1st April 2013, 18:07

:o A compiler which fails compiling the SPEC benchmark suite correctly without specific switches was released. :stupid:

A benchmark which relys on undefined/non-spec conform behaviour in a compiler :stupid:

LigH

1st April 2013, 18:22

Not only ... there may be more software which expects a compiler to generate binaries working as expected, no matter if optimization is enabled or not. ;)

nevcairiel

1st April 2013, 18:28

Not only ... there may be more software which expects a compiler to generate binaries working as expected, no matter if optimization is enabled or not. ;)

If you write your code so that someone reading the code with a C spec at hand can't tell you what is the expected result, what is the compiler supposed to do?

The point is, the only code thats broken is code which is itself broken, because it abuses parts of C which are basically undefined, so the result is undefined.
Read the GCC 4.8 release notes, it links to pages that explain why stuff breaks, and why the code is illegal.

And in any case, its not going to get chaged back in GCC, so might as well test and if such occurances occur, the undefined code should be fixed, not the compiler.

I didn't switch to 4.8 myself yet because i didn't have time to test everything, and the benefit is usually very small, but this wouldn't be a reason not to.

corporalgator

1st April 2013, 21:51

Dodgexander

2nd April 2013, 21:11

I'm also having a problem with the program not closing after I hit close since the last update. No subtitles used. It happens about every other time for me. I have it to close automatically or I just hit close, which should dump me back into XBMC, but instead I get stuck on a black screen, and I have to bring up task manager to kill the process.

Switching applications doesn't work since the app has been closed, but not the process.

Exactly the same as me, I first noticed it because reclock comes up each time saying that the audio is locked by another application (the previous process still running).

Sent from my Blade S using Tapatalk 2

clsid

3rd April 2013, 00:23

Test old builds to narrow down the range in which the problem started. That will be helpful for the devs.
http://xhmikosr.1f0.de/mpc-hc/old/

vood007

3rd April 2013, 20:12

I just found out that FXAA works with MPCHC directly when adding a profile for MPC in NV Control Panel and turning on FXAA. Works surprisingly good for reducing flickering from over sharpening. Try it... Maybe some On/Off options can be build into MPC directly?

JanWillem32

3rd April 2013, 21:49

FXAA is a 3D-perspective attuned pixel shading method to approximate higher levels of supersampling ('natural') anti-aliasing on output to the presentation buffers. It most certainly works on a 2D view of a single texture (the video output form), but such a heavy process is in the case of a video renderer better spent on the natural form of the input video. Meaning; filtering the images from the video in their generally Y'CbCr, chroma down-sampled, unresized state (and maybe a mild, but specific filter to compensate for the various resizer flaws after resizing). With the more decent video renderers the quality issues seen with consumer-grade video sources are not caused at the output stage of the video renderer where FXAA can be activated. Pretty much only the low quality of the video source matters.

turbojet

4th April 2013, 06:20

vood007: Thanks for sharing the tip. FXAA helps clean up the edges nicely, without much blur, especially SD sources on HD displays even with Jinc resizing. Another scenario it helps me is downscaling sources, due to lack of a Jinc like resizer. BTW if anyone is interested in sharpening, LumaSharpen (http://dropcanvas.com/l46x9) which doesn't get much attention on this board, does a really good job sharpening everything but high contrast edges by default. It might be deserving of adding to default mpc shaders.

vood007

4th April 2013, 13:57

vood007: Thanks for sharing the tip. FXAA helps clean up the edges nicely, without much blur, especially SD sources.....

Yes Lumasharpen and FXAA is a nice combo and a perfect finisher for my ffdshow/avisynth filter chain and best thing is it runs on GPU. For better control of GPU clocks i use Nvidia Inspectors (http://blog.orbmu2k.de/) Multi Display Power Saver and added MPC to its video profile. This way the card does not use full 3D clocks while watching videos.

toniash

4th April 2013, 15:15

Yes Lumasharpen and FXAA is a nice combo and a perfect finisher for my ffdshow/avisynth filter chain and best thing is it runs on GPU. For better control of GPU clocks i use Nvidia Inspectors (http://blog.orbmu2k.de/) Multi Display Power Saver and added MPC to its video profile. This way the card does not use full 3D clocks while watching videos.

How can I use FXAA on AMD/ATI?

JanWillem32

4th April 2013, 15:48

By implementing your own pixel shader stages. From what I've seen from the source code of an FXAA project, the regular stages for pre-resize and post-resize pixel shaders won't suffice, as secondary texture buffers are used (and it depends on a few extra environment variables). It's reasonably easy to create such stages when you have access to a renderer's internal rendering path.

Dodgexander

5th April 2013, 03:16

Test old builds to narrow down the range in which the problem started. That will be helpful for the devs.
http://xhmikosr.1f0.de/mpc-hc/old/

I have been trying to repeat the problem for a while, but the problem is it happens so intermittently its almost impossible to repeat.

GTPVHD

5th April 2013, 10:07

corporalgator

5th April 2013, 17:53

I downloaded the latest nightly and the problem with mpc not exiting correctly seems to have been fixed. I'll report back if it really isn't.

Armada

5th April 2013, 18:51

http://blogs.msdn.com/b/somasegar/archive/2013/04/04/visual-studio-2012-update-2-now-available.aspx

http://www.microsoft.com/en-us/download/details.aspx?id=38188

http://support.microsoft.com/kb/2797912

We were hoping that the compiler bug (https://connect.microsoft.com/VisualStudio/feedback/details/776147/bug-in-vs2012-x64-optimizing-compiler) we reported would be fixed in that release. It seems they fixed a problem very similar to it, but we're still able to reproduce it. Don't worry about compiling MPC-HC under VS2012 though, we worked around this bug as soon as we reported it.

corporalgator

6th April 2013, 02:58

Nevermind, latest nightly still has no exit problem. I'll go back and see if I can find out where it started.

betaking

6th April 2013, 05:11

S.Chinese update
http://www.sendspace.com/file/vaoq67

Dodgexander

6th April 2013, 15:26

Nevermind, latest nightly still has no exit problem. I'll go back and see if I can find out where it started.

Thanks for this, I am trying the penultimate stable build at the moment and so far no exit problems. But it doesn't happen every time(or when you expect it too) which is a pain.

Sent from my Blade S using Tapatalk 2