View Full Version : Media Player Classic - Home Cinema (MPC-HC) - v1.7.13
JarrettH
24th April 2016, 20:46
Couple things:
1) No audio in WASAPI exclusive mode in nightly build (March 14, 1.7.10.130), also the latest stable
2) In the Internal Audio Renderer status window it shows 44100 -> 48000...why? This WASAPI output does not look bitperfect. I have a feeling exclusive mode is needed for this to work properly with the device (USB DAC).
Thanks
kasper93
25th April 2016, 00:13
1) Works fine for me. Do you use USB DAC or something like that? We will need more info to diagnose the issue. Would be great if you could create ticket on https://trac.mpc-hc.org/timeline
2) Yes WASAPI output will match you hardware compatibilities/settings with our high quality processing. No one said it is bit perfect nor it should be. It will be bit perfect if it matches your audio device settings.
JarrettH
25th April 2016, 00:43
Yep, it's this one...
http://www.hirestech.com/product/?pid=107
re: 1) Sorry, my stream might not have had audio at the time, anyway, I can play in exclusive mode, but the 200 ms buffer (in the status window) is too high. I get pops/clicks during playback. For USB DACs, you need something like 25 - 50 ms for playback (pretty much confirmed everywhere as a solution to pops/clicks for USB DACs).
Some examples of lowering the buffer to fix pops/clicks...
https://yabb.jriver.com/interact/index.php?topic=55681.msg378794#msg378794
http://www.audiocircle.com/index.php?topic=119290.0
https://hydrogenaud.io/index.php/topic,92299.0.html
Could MPC devs allow the buffer to be changed? :thanks:
chros
25th April 2016, 11:06
2) In the Internal Audio Renderer status window it shows 44100 -> 48000...why? This WASAPI output does not look bitperfect.
2) Yes WASAPI output will match you hardware compatibilities/settings with our high quality processing. No one said it is bit perfect nor it should be. It will be bit perfect if it matches your audio device settings.
I believe it's fixed in the current master branch of sanear but it didn't get integrated into MPC-HC yet.
Btw, I'd really like to get this feature: "Surrender exclusive-mode device after inactivity" :)
https://github.com/alexmarsev/sanear/commit/af6d327bc3682977488d8c8f6f193592d4a927ed
PS: thanks guys for the continuous development!
rock
25th April 2016, 11:49
Here I found some issue that might be considered to be fixed
- it forgot the downloaded subtitles after click refresh, the checkbox appear again from the downloaded subtitles.
- supported language tooltip from opensubtitles is too long.
- I have local sub .ass, tried save subtitles function. It always suggest to save it as .srt. Would be good if it tried to save with original extension.
Just in case you missed it. Actually now I'm not sure about the first one, maybe we can just forget it. As for third request, I just found a provider that have other than .srt subtitle (.ass and .sub).
vBm
25th April 2016, 12:49
Just in case you missed it. Actually now I'm not sure about the first one, maybe we can just forget it. As for third request, I just found a provider that have other than .srt subtitle (.ass and .sub).
Majority of devs don't read doom9, and even when they do, they tend to forget about reports/feature requests if they don't start working on it right away.
Best bet is to open ticket at our trac. (https://trac.mpc-hc.org/)
rock
25th April 2016, 16:16
Majority of devs don't read doom9, and even when they do, they tend to forget about reports/feature requests if they don't start working on it right away.
Best bet is to open ticket at our trac. (https://trac.mpc-hc.org/)
That's fine. I hope I could do that again as until now I can't create ticket or make post due to my ip flagged as spam. Will try again some time later. Thanks. :)
Megalith
2nd May 2016, 01:38
Where are the settings for MPC-HC stored? I accidentally installed the x64 version and chose to get rid of the configuration file during the uninstall process, thinking it was linked to that specific version, but it wiped out the settings for the x86 version as well.
fvisagie
2nd May 2016, 06:19
Where are the settings for MPC-HC stored?
Searching for "MPC-HC" in Regedit brings one to HKEY_CURRENT_USER\Software\MPC-HC.
GCRaistlin
2nd May 2016, 10:56
+ Option to set audio delay values for each display mode
Where can it be found?
Where can it be found?
Option->Playback->Fullscreen
GCRaistlin
2nd May 2016, 12:13
rock, thanks. Though I don't understand why it is there:
Isn't the audio delay needed in non-fullscreen mode?
Why is the audio delay bound to the display mode? I use madVR/ReClock combination to change display refresh rate depending on the source FPS, so I don't need/want to change it by MPC-HC.
MPC-HC doesn't really recognize display device names and shows them all as "Plug and Play Monitor" (madVR shows them correctly). I have 3 display devices connected to my video card, only 2 of them can be active simultaneously ('cause my video card is AMD and all display devices are connected legacy), 2 devices that are being swapped require different audio delays - I have no idea how to set up this configuration properly by the current means.
I believe this feature was partially inspired by my ticket (https://trac.mpc-hc.org/ticket/5773), though unfortunately it doesn't seem to be really usable now.
Megalith
4th May 2016, 00:50
Searching for "MPC-HC" in Regedit brings one to HKEY_CURRENT_USER\Software\MPC-HC.
Thanks. I should have thought to check there, aside from the AppData and Program Files folder.
So, I'm currently using MPC-HC in conjunction with MKVToolNix to make clips out of MKV files, which can only cut on keyframes. Is there any way to ensure that you land on a keyframe in MPC-HC aside from clicking on the seek bar? I feel like using the mouse is inaccurate and existing keyframes that are close to one another are missed. If only there was an option to change frame step (Shift+Left/Right Arrow) to keyframe step...
Snowknight26
4th May 2016, 01:55
Thanks. I should have thought to check there, aside from the AppData and Program Files folder.
So, I'm currently using MPC-HC in conjunction with MKVToolNix to make clips out of MKV files, which can only cut on keyframes. Is there any way to ensure that you land on a keyframe in MPC-HC aside from clicking on the seek bar? I feel like using the mouse is inaccurate and existing keyframes that are close to one another are missed. If only there was an option to change frame step (Shift+Left/Right Arrow) to keyframe step...
Options -> Tweaks, enable Fast seek.
Megalith
4th May 2016, 02:31
Never mind, I'm an idiot. For some reason, I thought you could only frame step using the arrow keys. I even mentioned Shift+Left/Right Arrow in my last post, which *is* keyframe step...
Warner306
5th May 2016, 00:10
rock, thanks. Though I don't understand why it is there:
Isn't the audio delay needed in non-fullscreen mode?
Why is the audio delay bound to the display mode? I use madVR/ReClock combination to change display refresh rate depending on the source FPS, so I don't need/want to change it by MPC-HC.
MPC-HC doesn't really recognize display device names and shows them all as "Plug and Play Monitor" (madVR shows them correctly). I have 3 display devices connected to my video card, only 2 of them can be active simultaneously ('cause my video card is AMD and all display devices are connected legacy), 2 devices that are being swapped require different audio delays - I have no idea how to set up this configuration properly by the current means.
I believe this feature was partially inspired by my ticket (https://trac.mpc-hc.org/ticket/5773), though unfortunately it doesn't seem to be really usable now.
I also support this feature as it's also needed in my setup. But I would prefer it if it could be used in conjunction with madVR's display mode switching. I don't want MPC-HC to handle display modes.
Edit: This is better a feature request for madVR.
kasper93
5th May 2016, 20:20
Where are the settings for MPC-HC stored? I accidentally installed the x64 version and chose to get rid of the configuration file during the uninstall process, thinking it was linked to that specific version, but it wiped out the settings for the x86 version as well.
Settings are stored in registry (in already mentioned key) and are shared all MPC-HC instances you might run. You can however enable storing settings in .ini file, this way each instance will have own file with settings.
rock, thanks. Though I don't understand why it is there:
Isn't the audio delay needed in non-fullscreen mode?
Why is the audio delay bound to the display mode? I use madVR/ReClock combination to change display refresh rate depending on the source FPS, so I don't need/want to change it by MPC-HC.
MPC-HC doesn't really recognize display device names and shows them all as "Plug and Play Monitor" (madVR shows them correctly). I have 3 display devices connected to my video card, only 2 of them can be active simultaneously ('cause my video card is AMD and all display devices are connected legacy), 2 devices that are being swapped require different audio delays - I have no idea how to set up this configuration properly by the current means.
I believe this feature was partially inspired by my ticket (https://trac.mpc-hc.org/ticket/5773), though unfortunately it doesn't seem to be really usable now.
Audio delay can be set globally in audio switcher and has always been. It is completely independent from display modes. New feature lets everyone to override default audio delay per display mode. We have received requests for this feature because some TVs have different input lag depending of post processing and refresh rate. About display modes switcher in MPC-HC it could use some changes which may come soon.
I also support this feature as it's also needed in my setup. But I would prefer it if it could be used in conjunction with madVR's display mode switching. I don't want MPC-HC to handle display modes.
Edit: This is better a feature request for madVR.
Unfortunately if MPC-HC doesn't handle mode switching it can't really base audio delay on the mode. While this could be done by some monitoring on current mode, it would be confusing what's going on. Now audio delay is set when mode is changed and I think this is easy to understand for users. If madVR is used to control display mode it might also ask player/filters for appropriate audio delay. But this is little bit not sure if video renderer should do that.
GCRaistlin
5th May 2016, 21:13
New feature lets everyone to override default audio delay per display mode. We have received requests for this feature because some TVs have different input lag depending of post processing and refresh rate.
As I wrote in my ticket, some devices (projectors) have permanent input lag regardless of refresh rate. That's why I asked for the possibility to set not overriding but additional audio delay (that should be summarized with the current global audio delay) that is being set simply per device.
Warner306
5th May 2016, 23:16
As I wrote in my ticket, some devices (projectors) have permanent input lag regardless of refresh rate. That's why I asked for the possibility to set not overriding but additional audio delay (that should be summarized with the current global audio delay) that is being set simply per device.
Can't you do that with LAV Audio (set a global audio delay)?
GCRaistlin
5th May 2016, 23:45
Can't you do that with LAV Audio (set a global audio delay)?
What do I need setting global audio delay with LAV Audio for? I suppose you don't get what I'm talking about.
I have 3 display devices connected to my video card: 2 "general" displays and 1 projector. General displays don't introduce the input lag; the projector does. I watch movies sometimes on the projector, sometimes on the main monitor, sometimes on the second monitor. Now I'm forced to manually change the global audio delay depending on what device I'm watching the video currently on. Of course sometimes I forget doing that; then I'm forced to stop the playback, go to the console (if I'm watching a movie on the projector), exit fullscreen, change the audio delay... or at least to change the audio delay using my IR remote control... anyway it's not good while you're watching a serious movie!
What is more is that sometimes a movie's audio track requires the specific audio delay itself. So I need to sum up the delays and set the required delay. It's not a big deal - but not to forget to clear this delay after watching IS.
In short all this is a real PITA right now.
mbordas
8th May 2016, 19:52
Is there any way to toggle on/off the use of ac3filter? I'd like to enable it for stereo sources, but leave it off for surround sound. The only way I've found it to check prefer/block under external filters, but then I have to restart mpc-hc. Is there a better solution?
Piyoko
8th May 2016, 22:25
Hello, is it possible to somehow load an image and overlay that on the video as part of a pixel shader?
The uniformity of my TV is not perfect; the left side is tinted very slightly red, in the middle there's a band that's tinted very slightly green, plus the usual faint clouding etc. and it's kind of distracting when watching black and white movies.
So I had this idea of taking a photo of the TV displaying a solid grey image in order to create a sort of difference map from perfect, uniform grey, then load this image in a shader and use it to modify the corresponding pixel values to offset the variations of the panel. Similar idea to the LCD angle correction shader but more complex than a simple ramp.
Just wanted to check if anyone knew off-hand if this would at all be possible or not. Thanks.
fvisagie
9th May 2016, 05:24
Hello, is it possible to somehow load an image and overlay that on the video as part of a pixel shader?
The overlay can be done in Avisynth, and I understand Avisynth can be invoked from MPC-HC (I haven't done that myself).
AviSynth may be able to overlay an image in different modes, but not as fast pixel shader.
MPC-HC supports pixel shaders as rendering effects, but I can't tell you how to write one which applies a texture over a video. But it sounds possible to me, a kind of RGB offset texture, maybe in lower resolution and scaled up to the video resolution.
jmartinr
9th May 2016, 08:58
The overlay can be done in Avisynth, and I understand Avisynth can be invoked from MPC-HC (I haven't done that myself).
You can use ffdshow for that.
jmartinr
9th May 2016, 09:04
Come to think of it, you can also overlay a png directly with ffdshow. That's probably the most simple solution.
Piyoko
10th May 2016, 00:33
I was really hoping to avoid adding more stuff to the playback chain but I might try ffdshow, thanks for the tip.
Had a look at some HLSL tutorials and from what I gather it doesn't seem like it's possible to load another texture in the shader code, rather it has to be passed as an input (like the video texture).
HelmedHorror
13th May 2016, 04:37
Hey, I'm having trouble getting MPC-HC to work with a single mkv file with multiple video streams within it. For example, I have many movies with deleted scenes or alternate endings in a separate video stream.
When I change the video track in MPC-HC it does successfully change the video stream. However, the problem is the track duration and seekbar do not change accordingly. This is a huge problem that ends up crashing MPC-HC, as I'll explain below.
So, for example, let's say Video Track 1 (default video track) is 02:00:00 and Video Track 2 is 00:06:00. When I change the video track in MPC-HC to Video Track 2, the track duration (seen at the bottom of MPC-HC) remains at 02:00:00. So when I click in the middle of the seekbar, it tries to seek to an hour within Video Track 2 (which is only 6 minutes long). MPC-HC then immediately crashes. It ought to seek to 3 minutes (since that's half the duration of Video Track 2) but it doesn't because it still thinks the video track is two hours long.
This is true for every multi-video-track mkv file I've ever attempted to play in MPC-HC, and that's a lot of them.
I'd really hate to have to demux them all into separate mkv files because of this...
nevcairiel
13th May 2016, 08:38
Multiple streams in containers like MKV are expected to have the same (or at least a similar) duration. What you are trying to do is not what having multiple streams in one container is supposed to be used for.
If you want them all in one file, there are other ways with MKV, using ordered chapters and editions, which would allow MPC-HC to play these properly and allow you to switch from the GUI what to play.
I would not expect your current files to improve, and I would honestly be surprised if any player handles that situation any different. The file duration is largely independent of which track you currently selected.
GCRaistlin
13th May 2016, 08:46
nevcairiel
I suppose crashing is something that has to be fixed anyway, isn't it?
HelmedHorror
13th May 2016, 19:32
Multiple streams in containers like MKV are expected to have the same (or at least a similar) duration. What you are trying to do is not what having multiple streams in one container is supposed to be used for.
If you want them all in one file, there are other ways with MKV, using ordered chapters and editions, which would allow MPC-HC to play these properly and allow you to switch from the GUI what to play.
I would not expect your current files to improve, and I would honestly be surprised if any player handles that situation any different. The file duration is largely independent of which track you currently selected.
Interesting, well thanks for clearing that up! I'm glad I now know. I had spent hours trying to figure out what I was doing wrong...
Megalith
15th May 2016, 18:25
So what's the verdict on the Internal Audio Renderer and WASAPI output? I have been using it, but tickets like these are giving me pause.
https://trac.mpc-hc.org/ticket/5687
Does the default renderer actually do a better job of doing things like downmixing?
Stereodude
15th May 2016, 18:47
So what's the verdict on the Internal Audio Renderer and WASAPI output? I have been using it, but tickets like these are giving me pause.
It's totally useless as far as I'm concerned. It doesn't do the one thing most everyone expects of a WASAPI renderer, which is override the Windows speakers settings.
JarrettH
15th May 2016, 18:53
To add to that (I won't be registering for the bug tracker, sorry, it's not convenient): a selectable buffer for WASAPI exclusive mode would be useful (10, 25, 50 ms, etc). 200 ms is too high for USB DACs (pops, clicks are a common problem)
Megalith
15th May 2016, 19:16
It's totally useless as far as I'm concerned. It doesn't do the one thing most everyone expects of a WASAPI renderer, which is override the Windows speakers settings.
Really? I thought it did, since ticking "exclusive" mode prevents anything else from playing on my system.
Stereodude
15th May 2016, 23:16
Really? I thought it did, since ticking "exclusive" mode prevents anything else from playing on my system.
That's not what I was talking about. I'm talking about it playing 2 channel as 2 channel, 5.1 as 5.1, and 7.1 as 7.1 over HDMI regardless of how Windows speaker configuration is set. This is how Foobar2k's WASAPI output plugin does it and is what MPC-HC should be doing, but if you have Windows speaker configuration set to stereo MPC-HC will downmix to stereo when using the new internal renderer in exclusive mode. :scared:
chros
16th May 2016, 13:51
That's not what I was talking about. I'm talking about it playing 2 channel as 2 channel, 5.1 as 5.1, and 7.1 as 7.1 over HDMI regardless of how Windows speaker configuration is set. This is how Foobar2k's WASAPI output plugin does it and is what MPC-HC should be doing, but if you have Windows speaker configuration set to stereo MPC-HC will downmix to stereo when using the new internal renderer in exclusive mode. :scared:
Yep, that's an issue with my setup as well. Using exclusive mode and setting 2ch in windows speakers setup working fine until I play multichannel audio (pcm, flac), then I have to manually configure the system setting again (and back later on).
I raised an issue on github about making a new release.
Until then, can somebody compile a new build from current master that we can try as an external filter? @clsid, maybe? :)
Thanks!
chros
17th May 2016, 20:23
Yep, that's an issue with my setup as well. Using exclusive mode and setting 2ch in windows speakers setup working fine until I play multichannel audio (pcm, flac), then I have to manually configure the system setting again (and back later on).
I raised an issue on github about making a new release.
Until then, can somebody compile a new build from current master that we can try as an external filter? @clsid, maybe? :)
Thanks!
I've just built both x32 and x64 from f7612c7 @master (latest commit) on Windows 8.1 with Visual Studio 2013 Update 5.
It seems to be working fine with a quick test (both x32 and x64 versions), use it at your own risk since it's not an official release.
sanear-bin-master-f7612c7-20160517.zip (https://www.dropbox.com/s/7be96xtdumbd9l9/sanear-bin-master-f7612c7-20160517.zip?dl=0)
sanear-pdb-master-f7612c7-20160517.zip (https://www.dropbox.com/s/lo1lvftwpolbhnm/sanear-pdb-master-f7612c7-20160517.zip?dl=0)
You only need the bin package, register the filters, then select Sanear Audio Renderer at Output page in MPC-HC/BE instead of Internal Audio renderer. A tray icon will be present during playback where you can set properties.
And Wasapi Exclusive finally works as it should: no channel and sample rate conversion anymore. :) Also you can pause, make a skype call, then unpause. :)
Post your feedback of its stability (mention version and mode (exclusive or shared)) then maybe our beloved devs will integrate this current version if it's working well.
Post your feedback of its stability (mention version and mode (exclusive or shared)) then maybe our beloved devs will integrate this current version if it's working well.
As alex said (https://github.com/alexmarsev/sanear/commit/f7612c7f71316f357b00e43f5a497d1d0289ffe6#commitcomment-16589367), he wanted to finish "reclock thingie" before releasing next version. Also he said that he wouldn't want to integrate non 'stable' versions as "IAR".
dbcooper
18th May 2016, 04:22
As alex said (https://github.com/alexmarsev/sanear/commit/f7612c7f71316f357b00e43f5a497d1d0289ffe6#commitcomment-16589367), he wanted to finish "reclock thingie" before releasing next version. Also he said that he wouldn't want to integrate non 'stable' versions as "IAR".
Does this mean that MPC-HC is going to get a reclock alternative?
lych_necross
18th May 2016, 06:44
As alex said (https://github.com/alexmarsev/sanear/commit/f7612c7f71316f357b00e43f5a497d1d0289ffe6#commitcomment-16589367), he wanted to finish "reclock thingie" before releasing next version. Also he said that he wouldn't want to integrate non 'stable' versions as "IAR".
FramesToTime() helper for MPC? Very cool!! :D
I hope this corrects the PAL speedup problem on some movies.
chros
18th May 2016, 12:38
As alex said (https://github.com/alexmarsev/sanear/commit/f7612c7f71316f357b00e43f5a497d1d0289ffe6#commitcomment-16589367), he wanted to finish "reclock thingie" before releasing next version. Also he said that he wouldn't want to integrate non 'stable' versions as "IAR".
OK, understood, no worries. If somebody doesn't like the current internal one that can use the external one until then (if it's working fine).
To add to that (I won't be registering for the bug tracker, sorry, it's not convenient): a selectable buffer for WASAPI exclusive mode would be useful (10, 25, 50 ms, etc). 200 ms is too high for USB DACs (pops, clicks are a common problem)
If you like I can build one with modified buffer length in the meantime (right now the min is 60 : https://github.com/alexmarsev/sanear/blob/f7612c7f71316f357b00e43f5a497d1d0289ffe6/src/Interfaces.h#L15) that you can try out.
But, you should make an issue on github otherwise this feature will be never implemented. https://github.com/alexmarsev/sanear
JarrettH
18th May 2016, 17:23
Okay, I'll create an account. :p I already did find the links online to support it in an earlier post. Thanks for the offer, chros :sly:
https://github.com/alexmarsev/sanear/issues/13
Megalith
19th May 2016, 04:03
Would anyone be kind enough to explain how LAV Audio's mixing option affects reference levels? I've got my stereo monitor pair calibrated to 85 dB for movies. Since there is a significant difference (increase) in SPL when mixing (stereo) is enabled, which setting would actually give me reference level at 0 dB (105 dB peaks) on my preamp's dial? Enabled, or disabled?
I did the calibration by playing back pink noise in MPC-HC with the mixing turned off. But, AFAIK, none of that should matter since it's just a mono .wav file.
Nvidia graphics card, H.264 1920x1088, Microsoft DTV-DVD Video Decoder, EVR-CP + Half/Full Floating Point Processing, Microsoft DTV-DVD Video Decoder = Player crash.
Why is this happening? Driver bug?
Test files (https://yadi.sk/d/hnn00O1xrzVuM).
Bug is missing with LAV Video Deocder, because there is used a hack.
if (width == 1920 && height == 1088) {
height = 1080;
}
tahaa7
25th May 2016, 22:22
What is the difference between selecting "2: Speakers <device name>", "7: DirectSound: Speakers <device name>", and "11: Internal Audio Renderer" in Options -> Playback -> Output -> Audio Renderer?
filler56789
25th May 2016, 23:55
Bug #3881 (https://trac.mpc-hc.org/ticket/3881) is back — using build 1.7.10.207. MPC-BE is working as it should.
EDIT — To be more precise: when MPC-HC uses EVR, only LAV Video is detected as "H/W", but when using EVR-CP, all the other DXVA-capable decoders are also perceived as "H/W".
JarrettH
26th May 2016, 00:57
I think #2 and #7 are basically the same thing.
#11 is WASAPI audio output, which is better
Does not read the names from this variant playlist. In my country it's not rare variant such a playlist.
kasper93
28th May 2016, 22:57
Nvidia graphics card, H.264 1920x1088, Microsoft DTV-DVD Video Decoder, EVR-CP + Half/Full Floating Point Processing, Microsoft DTV-DVD Video Decoder = Player crash.
Why is this happening? Driver bug?
Test files (https://yadi.sk/d/hnn00O1xrzVuM).
Bug is missing with LAV Video Deocder, because there is used a hack.
if (width == 1920 && height == 1088) {
height = 1080;
}
I can't really tell. I don't have Nvidia GPU to test it. But it is likely we set something wrong for this case.
What is the difference between selecting "2: Speakers <device name>", "7: DirectSound: Speakers <device name>", and "11: Internal Audio Renderer" in Options -> Playback -> Output -> Audio Renderer?
"2" and "7" is the same and uses DirectSound audio renderer. "11" uses our internal audio renderer which uses WASAPI, has additional processing like stereo crossfeed and audio bitstreaming. See also Options -> Audio Renderer
Bug #3881 (https://trac.mpc-hc.org/ticket/3881) is back — using build 1.7.10.207. MPC-BE is working as it should.
EDIT — To be more precise: when MPC-HC uses EVR, only LAV Video is detected as "H/W", but when using EVR-CP, all the other DXVA-capable decoders are also perceived as "H/W".
Works fine for me. And I don't recall any changes that could break it. Are you sure it was working in previous builds?
Does not read the names from this variant playlist. In my country it's not rare variant such a playlist.
I see I can take a look at this, I think our playlist was designed to show filenames primary.
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.