Media Player Classic - Home Cinema (MPC-HC) - v1.7.13 [Archive] - Page 62

e-t172

21st March 2017, 15:21

I'm experiencing some extreme audio distortion/clipping/crackling during loud explosion scenes in the Star Wars films. VLC plays it with no audio problems at all. I've spent over an hour trying to fix it. I've changed virtually every audio setting there is in MPC-HC trying to identify the culprit. No other media I play has this problem. Ever. And I watch a lot of stuff. The films (original trilogy) are just the raw Blu-ray, with DTS-HD MA.

Resetting all settings to default does not resolve the issue. I've tinkered with every conceivable audio setting in MPC-HC. Yes, that includes everything in the audio decoder settings, splitter settings, and audio renderer settings. When I say every setting, I mean it.

I've also tinkered with all of VLC's audio settings to see if there is a way to introduce the problem into VLC's playback of the film, which might give an inkling as to what the problem is. No luck.

I've searched MPC-HC's bug tracker for various relevant keywords but couldn't find anything.

I'm open to any and all suggestions (although please keep in mind the facts I've given you from my troubleshooting attempts. I don't want to hear "Maybe it's your headphones", since it works just fine in VLC,) and I'll gladly give any pertinent information that anyone thinks might help diagnose this. It's driving me crazy.

Smells like high frequency garbage in the LFE channel. Try the following:

Enable stereo downmixing in the LAV Audio "mixing" tab. Check "Normalize matrix" to rule out clipping in the next steps.
In the mixing options, set the LFE level to zero. Then check if the issue is still there.
In the mixing options, set the LFE level to something high (like 1.5). Then check if the issue is still there.

If the issue disappears with the LFE level set to zero, and reappears as soon as LFE is brought back in, it means the LFE channel in the content you're playing is screwed up. This would not be the first time. (http://forum.doom9.org/showthread.php?p=1714617#post1714617) One reason VLC might not have this problem is because maybe it discards LFE, or maybe it applies a low-pass filter to it (a similar suggestion was made regarding LAV Audio, but I don't think nevcairiel ever got around to implementing it).

nevcairiel

21st March 2017, 15:24

VLC may not support lossless DTS-HD MA decoding, which would give you quite a different result.
An extra special part is that Star Wars uses 6.1 audio, not 5.1 or 7.1, that can cause weirdness in some cases.

huhn

21st March 2017, 16:07

reminds me of this: https://forum.doom9.org/showpost.php?p=1746666&postcount=19938

HelmedHorror

21st March 2017, 21:02

Smells like high frequency garbage in the LFE channel. Try the following:

Enable stereo downmixing in the LAV Audio "mixing" tab. Check "Normalize matrix" to rule out clipping in the next steps.
In the mixing options, set the LFE level to zero. Then check if the issue is still there.
In the mixing options, set the LFE level to something high (like 1.5). Then check if the issue is still there.

If the issue disappears with the LFE level set to zero, and reappears as soon as LFE is brought back in, it means the LFE channel in the content you're playing is screwed up. This would not be the first time. (http://forum.doom9.org/showthread.php?p=1714617#post1714617) One reason VLC might not have this problem is because maybe it discards LFE, or maybe it applies a low-pass filter to it (a similar suggestion was made regarding LAV Audio, but I don't think nevcairiel ever got around to implementing it).

VLC may not support lossless DTS-HD MA decoding, which would give you quite a different result.
An extra special part is that Star Wars uses 6.1 audio, not 5.1 or 7.1, that can cause weirdness in some cases.

To clarify, by "Enable stereo downmixing", you mean to set Output Speaker Configuration to Stereo, correct?

When I do that and set LFE to 0.00, the problem is fixed. When I raise the LFE slider, the problem comes back.

But somehow I doubt the film was mastered with such glaring LFE problems, would you agree? So where along the chain of audio processing is this problem occurring do you think? nevcariel mentioned the 6.1 audio in Star Wars, and I think he might be on to something (the Star Wars original trilogy are the only movies I have with DTS S24 6.1.) I just don't know enough about the lav filters or MPC-HC to know what to try next...

But if Star Wars is basically the only thing I can expect to have this problem with, that isn't so bad I suppose...

nevcairiel

21st March 2017, 21:37

But somehow I doubt the film was mastered with such glaring LFE problems, would you agree?

You would be surprised, such problems happen quite often because LFE is typically played through subwoofers that simply cannot play any higher frequencies - so noise in higher frequencies in those channels is not detected.

The proper way to deal with this is to lowpass the LFE before downmixing, and filter out the higher frequencies, unfortunately LAV does not support such a functionality (yet).

HelmedHorror

21st March 2017, 21:45

You would be surprised, such problems happen quite often because LFE is typically played through subwoofers that simply cannot play any higher frequencies - so noise in higher frequencies in those channels is not detected.

The proper way to deal with this is to lowpass the LFE before downmixing, and filter out the higher frequencies, unfortunately LAV does not support such a functionality (yet).

Interesting... so, why don't other movies have this problem? I have tons of DTS 7.1 movies which, by definition, all have LFE tracks. Did the audio technicians who mastered these other movies' audio simply cut out the higher frequencies in the LFE track, unlike the guys behind the Star Wars audio mastering? If that's the case, then the problem isn't Star Wars's unique 6.1 audio itself, correct? And that any audio of any surround configuration would have the same problem if they didn't take out the high frequencies in the LFE track?

nevcairiel

21st March 2017, 21:56

It definitely sound like a mastering mistake in the LFE channel. It happens occasionally, but the majority of discs are fine, of course.

And yes, such a problem is independent of the channel count, as long as it contains a LFE that is. If you use stereo downmixing, the 6.1 layout wouldn't matter either way, the LAV downmixer can handle that fine.

e-t172

21st March 2017, 23:13

Interesting... so, why don't other movies have this problem?

Because a lot needs to go wrong for that to happen:

There was audible noise that got into the LFE channel in the first place.
It wasn't removed by any lowpass filter anywhere in the production chain.
It wasn't removed by the encoder (the encoders for some formats, such as the lossy versions of DD/DTS, would have removed it, AFAIK).
It wasn't caught by the monitoring system of the sound engineer that mastered the audio track.

That's quite a contrived scenario, which is why it only happens rarely. But it can absolutely happen if enough things go wrong simultaneously. Then you're left with a disc that has garbage in the LFE channel. Meanwhile, 99% of the audience won't hear any problem because the LFE channel goes through a lowpass filter on their system (either an actual electronic lowpass filter, or the subwoofer itself), thus masking the issue.

Tier777

22nd March 2017, 19:55

Latest stable just reintroduced this old bug

https://trac.mpc-hc.org/ticket/1567

I tried old nightlies to pinpoint when the bug got introduced and found that
MPC-HC.1.7.10.40.x64.VS2015 doesn't have this issue and
MPC-HC.1.7.10.56.x64.VS2015 does.

Which means the bug was introduced on Jan 17th 2016.

I'm not a dev so I wouldn't know which commit is responsible for this behavior. Could anyone take a look at this?

https://github.com/mpc-hc/mpc-hc/commits/develop?after=f6d4865e4e35f2cab226fc08f4a324daf24f7683+244

JEEB

27th March 2017, 22:58

The about screen for MPC-HC should contain the exact hash. That should be useful to pin-point things :) .

clsid

28th March 2017, 14:05

Try disabling the "Global Media Keys" option in MPC-HC. I don't seen any obvious changes in the code between .40 and .56 related to key handling.

Tier777

28th March 2017, 18:00

I did, but no change. That's what makes it so weird.
Same bug in MPC-BE btw

rico001

4th April 2017, 06:11

I have Miami Vice set on Blu-Ray from Fabulous Films. Unfortunately they added too much contrast so i wanted to restore proper colors.

In MPC-HC Right click->"Renderer Settings" -> "Output Range" change from 0-255 to 16-235 option has no effect at all. No matter which setting is selected, My graphic card is ATI Radeon X1600 Series.

Thank You for help!

raymondjpg

15th April 2017, 03:09

Using latest versions of madVR 89.19 and Emby for WMC (Windows 7 Professional). With MPC-HC 1.7.10 after refresh rate switching by madVR (windowed overlay mode) then video hangs and does not play. Video will play after closing then re-starting MPC.

Refresh rate switching when using MPC-HC 1.7.9 works with video playing fine after refresh rate switching.

Has anyone else seen this, better still have a solution for it? I have also posted this in the madVR forum, but I suspect it is more of an issue with the latest version of MPC-HC than madVR
Exact same issue with MPC-HC 1.7.11, both 32 bit and 64 bit, and latest version of madVR. There has been no response to my post about this in the madVR forum, but I still suspect it is more of an issue with the latest version of MPC-HC than madVR. Does anyone have a solution for this?

Garringo

15th April 2017, 14:19

I would like to use MPC-HC to take screenshots (from Blu-Rays) that look as close to the source as possible. And I'm stupid. :(

I have taken the same shot with several different players and there is a noticeable difference between them in regards with the color red.

In the following image

http://imgur.com/a/sjIqF

1. Mplayer (bt.601 as I can't seem to be able to take them in 709)
2. MPC-HC with ffdshow with both high quality rgb conversion & dithering off.
3. MPC-HC ffdshow with both hq rgb conversion and dithering on
4. mpv
5. virtualdub (bt.601)

Does the higher quality rgb conversion in MPC-HC (and whatever the mpv and virtualdub use) make the red letters look better than what is actually on the disc? So the first two would be closer to the source? Or is the non-high quality conversion so shitty it makes it worse? :confused:

oddball

22nd April 2017, 11:19

nsnhd

22nd April 2017, 13:35

Currently the only way to play HDR content correctly without converting to SDR is to use Win10's Film & TV player. I am hoping at least one decent player/renderer fixes this. I suspect it requires output via DX12 which renderers like MadVR don't use AFAIK.

Is your Film & TV able to play 4k 10-bit hevc clips ? Mine was able to do that before creators update, now on CU it reports error 0xc00d6d6e, though it plays VP9 HDR clips fine.

GMJCZP

22nd April 2017, 16:22

Thank you for having this wonderful player.

I would like to make two observations:

- When playing mp3 files with front and back cover, the program gives priority to show the back cover, when it should display the front cover by default.

- If possible, in options you could choose which cover to display by default.

Thanks once again.

GMJCZP

24th April 2017, 00:56

I was analyzing this case better and the problem with MPCHC is that it shows the last saved cover, it is not if it is front or back cover. The program, I repeat, should give you preference to display the front cover, regardless of the order allocated in memory, such as WMP.

khanmein

27th April 2017, 10:05

Should I disable all the Internal Filters? (Source & Transform)

Thanks.

Blackwalker

27th April 2017, 13:11

hi all, ok my english is not good :(
i use always mpc-hc with MadVR and Lav filters.

My HTPC is a i5 750 (socket 1156), 6GB DDR3, ssd 120gb, nvidia 1050

why when i play 4k hdr video/movie , the movie snaps or i
if I try to send it back it hangs?

is a CPU/motherboard problem or software, suggestions?
thx :)

LigH

27th April 2017, 15:47

Blackwalker

27th April 2017, 15:59

It may be a problem with longer GOPs and the seek mode (fast vs. accurate). Try to switch this option: Fast seeking will seek to the closest GOP start, where decoding can start immediately, but it may be not exactly where you wanted to seek to; accurate seeking will decode from the previous GOP start to the exact seek position, which may take some time if your seek position is far away from a GOP start. And decoding HEVC is slow (assuming that 4K movies are encoded in HEVC).

hi and thx, i'll check it.
Anyway, i dont have to change CPU, Motherboard and Ram, correct? my 1050'd handle the 4k hdr movie!!

yes theh 4K movies are encoded in HEVC.

Sometimes when the movie/clip in 4k start ,
the video begins with obvious slowdowns and when i try to seek another position, the player die, lol :(

amayra

28th April 2017, 12:20

hi and thx, i'll check it.
Anyway, i dont have to change CPU, Motherboard and Ram, correct? my 1050'd handle the 4k hdr movie!!

this is definitely software problem how about refresh your htpc "reset default settings"
and you use latest version and driver check it with this ?
MPC-HC.1.7.11.13
madVR v0.91.9
LAVFilters 0.69.0-43
NVIDIA 381.89

Blackwalker

28th April 2017, 15:13

this is definitely software problem how about refresh your htpc "reset default settings"
and you use latest version and driver check it with this ?
MPC-HC.1.7.11.13
madVR v0.91.9
LAVFilters 0.69.0-43
NVIDIA 381.89

ho well, thats a great idea, damn me i didnt do it before lol
i'll try when at home and i'll let you know!
:)
thx to confirm thats a software problem!!! :goodpost:

Blackwalker

28th April 2017, 15:19

FarQueue

5th May 2017, 05:20

I'm having an issue with MediaInfo integration in MPC-HC. The analysis is taking a long time (over a minute sometimes) on certain files. The whole program locks up if you try and do anything while MediaInfo is loading. I'm not having this problem with other players that use MediaInfo (MPC-BE or PotPlayer) or MediaInfo itself.

The problem seems to have been introduced in 1.7.10.296 and remains in the current stable and nightly versions.

Perenista

7th May 2017, 03:30

I am having an unusual problem here... been using in Windows 10 64 bit an USB dongle 4.0 and the AirPods. They work fine, but if I use MPC-HC or even VLC and select another audio track other than the main one (example: a dubbing instead of the English track), the audio is muted and even the YouTube stream is muted. And if I switch back to the first track now it becomes muted.

The only fix is to close the player and reopen, however I'll continue to be unable to listen to the alternate track, the dubbing, or audio commentary for example. And the file checked is a Matroska lossless from Blu-ray.

Do you have any idea what is causing this and how to fix it?

P.S. I paired the AirPods with the USB dongle, and repaired just to make sure. Still happening.

P.P.S. I made an important discovery: if Bluetooth is disabled in Windows settings, then I open the video and switch to the dubbing/alternate track and enable Bluetooth and the AirPods again, the track will be heard. But if there's a manual switch from me to another track from the same file, the bug occurs AGAIN and it's all muted until I close and reopen the player. That proves to me that it's the mere switch that causes all this. Do you guys have any idea why?

For more info about the dongle:
https://forums.macrumors.com/threads/airpods-new-owners-views-problems-and-questions.2022574/page-51#post-24549191

***********
Update (June 5):

The bug was only fixed after I updated Windows 10-64 bit (Creator's Update). It was not the dongle's fault!

It's working as it should now! And it was affecting all players before, not just MPC.

JarrettH

7th May 2017, 03:41

Will there be a new nightly?

Btw, the "call for help" is attached to the bottom of the 1.7.11 news post, which you have to click into. Surely nobody will see it.

Damien147

7th May 2017, 17:04

Internal Audio Renderer is Wasapi?

sneaker_ger

7th May 2017, 17:17

Yes, it uses wasapi.

Damien147

7th May 2017, 22:21

Ok,thank you.

chros

10th May 2017, 19:09

Ok,thank you.
Internal is buggy, you have to use a newer build of it (sanear* (http://tmod.nmm-hd.org/Misc/)) as an external filter.

mbordas

14th May 2017, 18:46

Internal is buggy, you have to use a newer build of it (sanear* (http://tmod.nmm-hd.org/Misc/)) as an external filter.

I thought this was updated at some point, but can't find a version number in the changelog? I've been using the internal renderer since it came out so I could get off reclock and upgrade to windows x64, and haven't noticed anything particularly buggy. But I also don't know if I would have been better off staying with reclock.

blaster00

15th May 2017, 03:42

Internal is buggy, you have to use a newer build of it (sanear* (http://tmod.nmm-hd.org/Misc/)) as an external filter.

sanear-bin-master-f7612c7-20160517.zip (https://github.com/alexmarsev/sanear/issues/12#issuecomment-219823720) is made by you, right?
The internal and sanear both have slight sound come out after mute in mpc-hc, very funny. Sanear nightly crashes mpc after pause on my XMOS based usb dac, but works fine with my motherboard buildin sound chip. Sanear 0.3 works fine with both.

chros

17th May 2017, 11:54

I thought this was updated at some point
No, it wasn't. From the linked Github issue:
"I went through the commit logs, these are the major changes:
- Ignore System Channel Mixer option
- Guided reclock
- Use FFTW library with rubberband
- Add DspTempo/DspTempo2 selection
- Surrender exclusive-mode device after inactivity (Now you can pause, make a skype call, then unpause.)

The biggest change is that Wasapi Exclusive finally works as it should: no channel and sample rate conversion anymore."

sanear-bin-master-f7612c7-20160517.zip (https://github.com/alexmarsev/sanear/issues/12#issuecomment-219823720) is made by you, right?
Only compilation :)

The internal and sanear both have slight sound come out after mute in mpc-hc, very funny. Sanear nightly crashes mpc after pause on my XMOS based usb dac, but works fine with my motherboard buildin sound chip. Sanear 0.3 works fine with both.
Interesting, not for me (internal Intel audio with HDMI).
You can give MPC-BE a try, it has its own WASAPI implementation.

blaster00

17th May 2017, 13:32

Interesting, not for me (internal Intel audio with HDMI).
You can give MPC-BE a try, it has its own WASAPI implementation.

Used audio render from MPC-BE for a while, but tired to update from two projects. Sanear is acceptable for now. Hopes Sanear get the reclock thing working soon. ReClock from slysoft won't support x64, what a pitty.

chros

18th May 2017, 11:36

Used audio render from MPC-BE for a while, but tired to update from two projects.
:) I hope you're kidding, it's the simplest thing and you don't have to do it every day.
Sanear is acceptable for now. Hopes Sanear get the reclock thing working soon. ReClock from slysoft won't support x64, what a pitty.
I don't think so, since nothing happened in the last year.

Yanak

24th May 2017, 12:40

Hello,
I'm not sure if MPC-HC is even concerned by this but better ask instead of being worried :

http://blog.checkpoint.com/2017/05/23/hacked-in-translation/
https://torrentfreak.com/malicious-subtitles-threaten-kodi-vlc-and-popcorn-time-users-researchers-warn-170523/

Thank you.

clsid

24th May 2017, 14:52

No, MPC-HC doesn't allow overwriting arbitrary files when extracting files from a downloaded ZIP file. It only extracts files with a valid subtitle file extension.

Yanak

24th May 2017, 20:26

Hello and thanks for the reply,

You mention zip file extraction, i don't know exactly how this thing is supposed to work to attack but saw no mention of extracted files from compressed archives on the various articles i saw today on this subject, unless i missed it .

The only thing i saw is on the second link i posted where there is a link going to github supposed to bring a fix on a program :
https://github.com/butterproject/butter-desktop/commit/0ab1938a69f4277ea9a36b08b8b1b06cd987759c

I'm not a coder but i don't see any mention of zip files, only .srt .

Again i don't know how this thing is supposed to work nor is a coder but problem seems to come directly from the .srt file itself where some malicious code could be implemented so i don't know.

Thank you.

LigH

24th May 2017, 21:01

One possible attack vector mentioned in a blog reporting about it:

Some providers offer compressed files in archives. Players would extract subtitle files from these archives in such cases. But if they blindly unpack the whole archive, without sanity checks, they might be vulnerable to malicious structures like filenames containing patterns like "../", which provokes extracting and overwriting files in surprising and even dangerous locations by traversing to the root of the file system and back into system directories from there.

There may be other attack vectors, like linked files in advanced subtitle formats.

clsid

24th May 2017, 22:53

The vulnerability is in how the player handles archives. There is no issue with the subtitles files themselves. Those are just text and images, nothing scriptable/executable.

Yanak

24th May 2017, 22:58

Thanks for the answers guys, just wanted to be sure.

Thanks a lot.

LigH

29th May 2017, 10:55

There seem to be different attack vectors (https://news.ycombinator.com/item?id=14408859) depending on the different playback systems and their individual vulnerabilities:

>> From what I understood, one is an XSS (popcorn-time), one is a heap-based buffer overflow (VLC), and one is a zip-transveral (Kodi). {traversal}
^
Looks like PopcornTime was rendering subtitle text as HTML, inside their app (html/js-based), creating an XSS vector...

MPC-HC is probably not affected by issues provoked by directory traversal via malicious file names in ZIP archives; that can't exclude a possible, yet unknown weakness in any other method to display subtitles it uses, though. Attack vectors can be quite exotic. I remember a vulnerability in the TrueType renderer which allowed code execution in kernel space, because fast text rendering was once a benchmark bottleneck.

For now, MPC-HC is not affected by either of the attacks this article mentioned.

sneaker_ger

29th May 2017, 15:30

You can be pretty sure MPC-HC has holes as well. Every software has, especially if it's old and messy. They don't get found unless some security engineers start searching for them (like Google is doing for ffmpeg).

cork_OS

1st June 2017, 17:35

Hello.
Is it possible in MPC-HC to set playback rate to 1.25 (like youtube does) and zoom video to 400%?
Thanks.

sneaker_ger

1st June 2017, 17:49

In "Options"->"Playback" you can set speed step to 25% and then use the speedup button. Zoom can be done using hotkeys but I don't know if factors >3 are possible.

marcurling

1st June 2017, 22:18

Pals, I'm desperate.

I'm running last stable x64 versions of MPC-HC & MPC-BE and they now crash when using any File menu feature (while a file tries to launch for file related functions) and this with any type of file (video, audio, image).

They work when booting Win7 in Safe Mode but with no audio and with the included error.png

VLC, ffplay/ffmpeg and WMP still works fine.

This is a dump of MPC-HC last crash
https://drdump.com/UploadedReport.aspx?DumpID=15242385

The problem occurred more than 1 month ago but I can't find what could be the cause.
I'm running Win7 64bit.
I'll give you any info you'll need to help me solve this : MPC players are really important to me.

Thanks in advance, marcurling.