MPC-HC (users discussions/info) (to support go to github) [Archive]

hajj_3

22nd December 2024, 12:16

Here are some ideas that i think would make MPC-HC's user interface look better:

Current version (v2.3.9):

https://i.ibb.co/Qv75y1D/mpc-hc2.png

My UI ideas:

https://i.imgur.com/usdNVJA.png

https://i.ibb.co/hKzZy2q/mpc-hc3.png

https://i.ibb.co/jwTGPVC/mpc-hc9.png

https://i.ibb.co/vsNG7QP/mpc-hc7.png

https://i.ibb.co/KrFg5tP/mpc-hc19.png

https://i.ibb.co/mv6824h/mpc-hc14.png

https://i.ibb.co/j5qwspF/mpc-hc12.png

1. Moving the video time elapsed from the status bar into the seek bar like in MPC-BE would be an optional setting. The reason for this is that i'd like to disable the status bar so that the video size can be larger but i still need to see the time elapsed. Alternatively the time elapsed could automatically move to the seek bar if you disable the status bar without requiring a setting.
2. I moved the volume level percentage to the right of the volume bar and changed the volume level from grey to white or the same blue colour as the seek bar position colour. The grey volume level is difficult to see.
3. putting the playback icons in the centre and putting the next icons to the left and right side of the combined play/pause button looks better imo. I have combined the play and pause button as there is no need for 2. i made the play/pause blue too when it has been clicked.
4. I would like to see a blue square to appear around the icons when you hover your cursor over them like the blue square around the pause button in the screenshot.
5. I would like the grey colour around the pause/play alert in the top left of the program to be blue and for the text to be white instead of pale grey.
6. I would like the height of the menu bar at the top to be increased as it is very small and increase the font size for it, the same goes for the filename and logo.

Unfortunately i don't have the programming knowledge to implement these changes.

clsid

22nd December 2024, 13:46

New thread to info about new versions (https://github.com/clsid2/mpc-hc/releases) or users discussions.

For bugs or feature request use github (https://github.com/clsid2/mpc-hc/)

Github is for bug reports only!

Feature requests will be deleted immediately.

I do not have the time and interest in implementing other peoples requests. Half the requests are for functionality that is already available in the player, but people are too lazy/stupid to even look through the options. The other half are just plain stupid ideas, impossible, or a shitload of work that nobody wants to do.

I AM NOT IMPLEMENTING ANY FEATURE REQUESTS !!!

Thanks to those wankers who kept submitting requests, there is no longer proper support for MPC-HC anymore here on the forum.

Sunspark

22nd December 2024, 19:36

> shitload of work that nobody wants to do

You have said in the past that you don't want pull requests either.

People do not write because they want to give you a hard time, they write because they aren't aware that they aren't allowed to participate.

You absolutely need to work on your social skills. Torvalds took time off to do it, and you need to do it too.

Grimsdyke

22nd December 2024, 19:58

Madshi has left also now years ago but his MadVR thread here is still very active with lots of useful informations and discussions almost every single day !!
So I really don't understand this fuzz. So clsid please leave or stop the insults.

clsid

22nd December 2024, 21:47

You have said in the past that you don't want pull requests either.No, I never said that. I only denied PRs for specific changes that I don't want.
In fact, when excluding translation PRs, my fork of MPC-HC has a HUNDRED times more PRs than MPC-BE.

I don't need to adjust my social skills and never will do either. I am not here to make friends. I just tell people straight to their face what the reality is. It is your problem that you can't handle that reality.

You guys can discuss in this topic for years. You can make thousands of feature requests. But they will go ignored.

MadVR thread is a giant chaotic mess. Plus that is now about USAGE and not about new features, since it is obvious that they aren't coming.

VoodooFX

22nd December 2024, 23:58

It's almost Christmas but an important feature is missing, a Santa hat in a tray, like VLC:

https://i.imgur.com/ZdFo7jR.png

clsid

23rd December 2024, 00:19

MPC-HC has two much better x-mas eggs (https://imgur.com/a/HI6GeCO).

varekai

23rd December 2024, 08:52

MPC-HC has two much better x-mas eggs (https://imgur.com/a/HI6GeCO).
That's so ridiculously silly it makes you cringe, not an iota of finesse, very appropriate coming from you...

Perenista

26th December 2024, 23:55

There's something going on that I can't explain...

I did a reencode of that video:

https://pastebin.com/NTQmzvUT

Which is fine, BTW. All subtitles displayed OK.

This is the new file, created by HANDBRAKE, using a "5000" bitrate, same FPS as source, as H.265 (also MKV).

https://pastebin.com/5Usja10a

Also, all tracks from file #1 were imported to the smaller one, using MKVToolnix. The exception for the DTS-HD MA one, downconverted to DTS-ES by TSMuxer.

Then, when I open the smaller file with MPC-HC, all I see are bugs:

- The subtitles are stretched.
- It seems the player has some lag when going fullscreen. At least with MADVR (OK for MPC Video Renderer).

What exactly happened with these subtitles? Simply importing SRT files never messed with them whenever I used MKVToolnix, regardless of the targeted file resolution. Was this file broken for displaying with MPC-HC only?

Or is it broken according to these new specs?

P.S. One more thing: when I did the reencode, I only handled the MKV with the video inside it. I created a copy removing all other tracks from it. Reinserting the audio/subtitles from the bigger file was done after the reencoding was finished.

huhn

27th December 2024, 00:38

Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 2.35:1

Width : 3 840 pixels
Height : 2 160 pixels
Display aspect ratio : 16:9

sar=255 / sar-width / : / sar-height=360:277

Perenista

27th December 2024, 01:22

Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 2.35:1

Width : 3 840 pixels
Height : 2 160 pixels
Display aspect ratio : 16:9

sar=255 / sar-width / : / sar-height=360:277If this was reencoded incorrectly why VLC displays perfectly, with no stretched subtitles at all? I tried using MKVToolnix and inform 1920x818 in the "display width / height" and didn't help, same if we use 16/9 or 2.35 in the "set aspect ratio". Don't tell me I am going to have to reencode this whole thing in Handbrake again...

All other 1080p files I reencoded were fine, except this 4K one.

P.S. There's no lag anymore, it was probably a memory problem after I did this delayed reencode (restarting Windows fixed this). As for this option:
https://forum.doom9.org/showthread.php?p=1985942#post1985942

"Apply aspect ratio compensation for anamorphic videos"

Turning it off fixes the issue I was having with these subtitles. It was always on by default.

huhn

27th December 2024, 08:29

Perenista

27th December 2024, 17:14

that doesn't fix that fact that the encode is odd. VLC may not care about disapplying streams correctly.

resolution is 1/4 but has a different AR for what ever reason and none square pixel.

the subtitle just follow the spec and stretch because the video says it is stretched that option is default for a good reason.I figured out what was the problem with the reencode. "Anamorphic" needs to be turned off, completely (in HANDBRAKE settings).

If it's on (automatic), the "pixel aspect" will not be 1:1. Instead, it can be something like 45:34, and below we will see the video will be probably created at 1920x1080 with the 2.35:1 aspect ratio (original res. was 3840x2160, and this is not taken from a 4K-UHD disc, it's a regular video), but on the left side it will appear the "display size" is 2541. Scaled size to become 1920x1080.

This is how it was been reencoded, creating the stretched subtitles:

https://i.postimg.cc/ZTxkBBSq/BEFORE.jpg

And this is how it appears after I changed:

https://i.postimg.cc/zJrZGw4M/AFTER.jpg

Also, it's consistent the resolution was reduced to 1920x832, with other files I saw on the internet. I was telling the reencode to be 1920x1080 no matter what, but this caused the anomaly in the 1st place.

I checked the new reencoded file with subtitles, and they were displayed correctly. Even with the "Apply aspect ratio compensation for anamorphic videos" in MPC-HC settings, ENABLED again. So this is the solution.

(And there was nothing wrong with MPC-HC, it was just a default setting from HANDBRAKE creating this problem).

hajj_3

30th December 2024, 18:24

Another thing i'd like changing is for "save image..." to be renamed "save video frame image..." as it is unclear.

DvineLord

31st December 2024, 13:56

Sorry for what might be an annoying first post. I have been dealing with a "Recent Files" bug that has been present in all versions past "2.1.5". If I clear "Recent Files" then play a file it should start to populate right away, after "2.1.5" I have to exit and reopen the player before it starts to populate again. I am running "Windows 7 64-bit".

I decided to signup on Doom9 instead of Github because I might actually post here more often although I visit Github daily. I just feel old and tired of signing up for things. I didn't realize the clsid MPC-HC thread was closed until after I registered here.

After trying to post earlier it told me I had to wait 5 days.

tebasuna51

1st January 2025, 10:43

@DvineLord, welcome to the Doom9 forum.

Yes, seems a little bug but, like you know how solve it, we can't do anything more.

This thread is not for bug reports/feature request and clsid can read, or not, this post and can solve, or not, the problem in next versions, is their choice.

hajj_3

8th February 2025, 22:39

v2.4.0 has merged the play/pause buttons in to a single button :)

DvineLord

20th February 2025, 08:20

... I have been dealing with a "Recent Files" bug that has been present in all versions past "2.1.5". If I clear "Recent Files" then play a file it should start to populate right away, after "2.1.5" I have to exit and reopen the player before it starts to populate again. I am running "Windows 7 64-bit". ...

I see lots of code changes likely related to the "Recent Files" bug. https://github.com/clsid2/mpc-hc/compare/2.1.5...2.1.6
Can someone else confirm the bug and possibly create an "Issue" on github!?

v0lt

12th August 2025, 17:02

MPC-HC 2.5.2 released (https://github.com/clsid2/mpc-hc/releases). :)

bob88

16th September 2025, 15:19

Hi,
Windows HDR turns ON on its own and I cant stop it.
This started happening ever since I got a new monitor.
It does not happen with VLC and it does not happen when I uninstall the nvidia driver and use win driers.
Nvidia control panel has a convert SDR to HDR when Win HDR is on setting but that is off and grayed out.
And Windows HDR is off.
Yet when I play a video with MPC-HC Win HDR turns on and when I turn it off in windows settings it does blink but does not affect the situation.
So basically I am unable to play video with MPC-HC on this monitor because Win HDR goes on automatically.
To make things worse I have a movie projector which has not had any such issue before and now Win HDR turns on there as well. And I have to reinstall the nvidia driver and never play a video on the monitor to ensure that the projector works normally.
Any advice would be appreciated. Thanks.

DvineLord

20th September 2025, 11:21

So at first I didn't want to see any UI changes made to MPC-HC, but I actually like the new "ToolBar" and "ToolBar Layout" options in "2.5.3"
I would love for the "Recent Files" issue I posted earlier to be resolved without me having to make a GitHub account and create a "New issue". =)

JarrettH

24th September 2025, 04:29

I increased my toolbar size to about 28, added the audio and subtitle tracks (thank you, a massive convenience!), and for sure removed the increase/decrease rate because I can't think of why I would use that and have only accidentally clicked it :D The RRR Artline toolbar is a nice addition for when I get tired of the built-in one.

sam0t

12th October 2025, 12:41

Unsure how to label this issue/matter, as it's not a bug, but though to share just so that other not so tech savvy MPC-HC users can possibly save quite a bit of frustration and time.

Couple weeks ago noticed that video files with DTS-HD MA audio tracks (all other DD audio tracks worked ok), I'd start to get a lot of dropped frames with skipping video and garbled audio, when jumping back or forward with the video. Long story short, after a lot of fruitless MadVR related setting tweaking with AI, I stumbled across a 5y old issue post at MPC-HC (clsid2) Github, where it was suggested that one needs to have the DTS selected at Internal Filters settings under Transform Filters settings to resolve the issue. Add to that disabling DTS-HD bit streaming at LAV Filter's Video section (since my soundbar doesn't support DTS-HD MA audio tracks), and videos with DTS-HD MA audio track play again perfectly (albeit as DTS 5.1 of course) without dropped frames. Unsure how or why DTS was disabled with the Transform Filters section of MPC-HC settings in the first place (maybe these are the default settings for MPC-HC after installation?), but just glad I came across the old post about it, since AI did not bring up this setting at all.

clsid

12th October 2025, 20:14

In case of problems always do a settings reset:
Options > Miscellaneous > Reset

Little_Dragon

19th October 2025, 18:54

This is probably a foolish question, but I am admittedly a fool, and have no one else to ask. Is it possible to achieve the video equivalent of gapless playback in MPC-HC? I'm working on some looping videos, and have a number of short clips that I'd like to play back-to-back to see how well they transition, but the window goes blank for a split-second whenever the next video in the sequence starts.

I'm hoping there's some sort of obscure pre-buffer setting I can tweak, but suspect that it's probably an inherent limitation of the renderer, and I'll just have to live with it.

clsid

19th October 2025, 23:04

With independent separate files, impossible, even for audio files.

If videos are all same format and encoding, then Matroska segment linking could be used to turn them into a single virtual file from POV of player. But you need to go ask elsewhere for help on that.

Little_Dragon

20th October 2025, 03:55

I can just append them in VirtualDub. Previewing them in MPC-HC would have been more convenient, but I knew it was a longshot. Thank you for replying.

DvineLord

2nd November 2025, 11:56

clsid, thank you so much for fixing the recents issue in 2.5.3.10 it means ALOT to me!!! I didn't realize it was fixed until 2.5.5, I had to go back and test earlier versions to figure out when it got fixed. It was broken from 2.1.6 to 2.5.3 of the stable releases. Thanks again!!!

CruNcher

29th November 2025, 13:51

Works Perfect outside of DWM most fail

very stable from execution to destruction

no sync and resync failed yet

with this stability you can easily detect wrong syncs

clsid for testing purposes it would be cool if you could define a skip time for the playlist so lets say X seconds playback before it switches to the next

stability no matter the framerate is just awesome in this build

i see or hear no adaption failure every lock is as perfect as it can be and if it fails for your eyes you can say something in the source is wrong

and just you know firefox is running in the background with a lot of tabs and multimedia things another hardcore stability test level.

this build from a core stability view is excellent (core stability means video/audio/sync only)

Don't change your compilation environment to fast now clsid

The ball hitting the ground and and the audio feedback perfectly sync the racket hits perfectly sync

https://iili.io/fnS3eXs.jpg
https://iili.io/fnSU1JS.jpg

Seems sync as well (annotation guessed/bat hit perception/ramble sounds and refuge hand smashing perception/the golfball hit perception/voice and music latency perception perfect)

https://iili.io/fnUJKvt.jpg
https://iili.io/fnUXlP1.jpg
https://iili.io/fnUgFHB.jpg
https://iili.io/fngYsZg.jpg
https://iili.io/fngG4fe.jpg
https://iili.io/fng88a2.jpg

Voice/Wind (environment) audio all latency perception perfect

https://iili.io/fngmjJ1.jpg
https://iili.io/fnrnvXp.jpg

Artificial Audio Effects (Boom Explosion sounds by audiovisual perception plausible,voice latency ok, by audiovisual perception full playback suspected)

https://iili.io/fnr51Ie.jpg
https://iili.io/fnrjB2t.jpg
https://iili.io/fnr81Ul.jpg

Strange problem (probably in source identified) Transport Stream Recording with massive problems also possible IO transfer problem.

0 Playback Full unplugged live voice shouting into the microphone

https://iili.io/fn4umfR.jpg
https://iili.io/fn4wc42.jpg

0 Playback Full unplugged live voice shouting into the microphone

https://iili.io/fnPGimJ.jpg

My latvian/finish/czech/hungarian ??? is not so good but it's sync whatever the language

https://iili.io/fns9Fyl.jpg
https://iili.io/fnsxt9I.jpg

These down spikes are interesting they do not seem to come from the system it wants to sync to 59 actually it does for a split second , it makes the impression that could be keyframes or micro record failure in the transport stream.

https://iili.io/fnLd82n.jpg
https://iili.io/fnLUsxR.jpg

CruNcher

29th November 2025, 21:11

This stuff is crazy very old streams from the beginning sometimes hard to understand what what for errors i see here Recording System/Tape Failures/Encoder/Decoder Error/Bitstream Data Corruption ?

Still some streams to go and understand how they damaged or not.

But MPC-HC should be everyone's first choice to test with when its about stability and than test others if you need features

The combination of MPC-HC with Lav Decoder internal 0.80.0.9 in this build is pretty awesome, from a core stability point for normal playback use cases.

https://iili.io/fntzMOl.jpg

Another weirdo you hopefully don't see these days anymore someone tried some Hollywood kinda simulation but outside of Film Spec ?

How that Car Stutters on its last moves the whole thing stutters

How perverted you can be, apart from the content context ?

They experimented with VC-1 and did some really crazy things, could be a mistake in using the Encoder.

But its not just 1 time.

https://iili.io/fnmZXyJ.jpg

This is how you do digital film (speed)

https://iili.io/foHGHtR.jpg
https://iili.io/foHugKx.jpg
https://iili.io/foHr9nI.jpg

Sunspark

29th November 2025, 21:29

Calm down bro. You're too excited nobody will understand what you're trying to communicate.

Slow down, just say you're happy.

Klaus1189

30th November 2025, 11:24

Can somebody explain what is fixed/better now? I didn't understand anything :confused:

CruNcher

30th November 2025, 12:49

Can somebody explain what is fixed/better now? I didn't understand anything :confused:

Look in the changelog, what i did here was a general assessment of the development state mainly on the core point of overall playback stability in a complex system playback environment.

The combination integration of the ffmpeg/lav core with the ontop player gui integration and its overall stability together specifically on Windows NT as the Host OS and specifically on Nvidias GPU Hardware (Ampere/VP11) in co function with AMD Ryzen (Zen 2) CPU and specifically outside of DWM.

Calm down bro. You're too excited nobody will understand what you're trying to communicate.

Slow down, just say you're happy.

Yeah that's the correct translation thank you.

Klaus1189

30th November 2025, 13:14

Can you post link to latest changelog? I can only find changelog of 2.5.5 stable release.

For me MPC-HC was stable for years, what changed?

huhn

30th November 2025, 13:36

https://github.com/clsid2/mpc-hc/releases

Klaus1189

30th November 2025, 15:51

OK, where do I see the changelog for this Development Build? Somehow I can not find it there.

lvqcl

30th November 2025, 15:59

https://github.com/clsid2/mpc-hc/commits/develop/

Sunspark

30th November 2025, 16:04

Can you post link to latest changelog? I can only find changelog of 2.5.5 stable release.

On the 2.5.5 releases page, you will see it say something like 32 commits to develop since this release. If you click on that, you will see the individual changes to the code since the 2.5.5 tag.

CruNcher

30th November 2025, 16:17

Graph and Counter run Sync as soon as the spike leaves the visual area the counter adapts the framerate back

gonna isolate all streams with the spike

https://iili.io/fok63Yu.jpg
https://iili.io/fovQfMx.jpg

spike comes from other processes (probability) 0%

https://iili.io/fo8S1VI.jpg
https://iili.io/fogSFku.jpg
https://iili.io/for8RCQ.jpg

Same Stream Capture after a transmission brake and change and here they are again ?

https://iili.io/forZccG.jpg

Looks like LG/MSICT/Ateme made some tests on the transmission side, yep they tested something here.

Hmm do we know each other ?

https://iili.io/fotES6B.jpg

Klaus1189

30th November 2025, 16:59

Yeah, I can see the commits, but what is exactly different? You said I should see the changelog. Where is it?
As I already said, it was good before for me, so what is exactly better now? I don't get your hype. I want be be hyped, too.

Sunspark

30th November 2025, 17:47

Does it make a difference when you use different renderers? mpcvr+lav configured for d3d11, madvr+lav configured for d3d9, etc?

clsid

30th November 2025, 19:41

Klaus1189

30th November 2025, 20:02

Thanks for confirming.

CruNcher

30th November 2025, 22:44

He is "testing" an old obsolete and unmaintained renderer that hasn't been modified in YEARS...

What you can be hyped about is that MPC-HC has been rock solid and stable for YEARS...

Anyone that doesn't get good sync has a broken Windows install.

obsolete maybe stable surely

solid for years you would attest that for every system configuration im not sure if i would do that

what about the crashing when the energy saving comes into play and the device doesn't respond fast enough and the respond timer neither while a change on the playlist happens and MPC-HC goes by by completely

calling something rock solid on Windows alone i think many did that also before the chinese took their systems apart.

Who not has a broken Windows Install ?

I was just trying MPC-BE and MPC Video Renderer together and i was like ehhhhh what all the judder test fail at some cyclepoint maybe the longest stability was 4 cycles before it broke ?

i hope MPC-HC and MPCVR will do better

I think Elon also said that before he put the driver into Betatester mode but nicer than the Alphatester mode before.

Anyway MPC-HC does good work in finding something in that "Obsolete configuration"

The Spikes are coming from the Keyframes that's clear now but those are even just the sane ones

ATEME Titan KFE 3.5.1 <- it was going wild pretty fast replaced by
ATEME Titan KFE 3.6.1

i wonder if those transmission was even supposed to be 60 fps int and not 59.940 i mean broadcast in 60 that's weird

First transmission part was ok 2nd got high wire and they really did the broadcast in 60 fps int

There was this Broadcast in 60 FPS int and LG made several demo versions out of it in 59.940 for different Device Demos

_________________________________________________________________________________________________________

Process with highest pagefault count: firefox.exe

Total number of hard pagefaults 204
Hard pagefault count of hardest hit process: 102
Number of processes hit: 12

CruNcher

2nd December 2025, 13:49

I think the problems in MPCVR are a overall performance issues you can get 23.976 stable but higher framerates it becomes more and more problematic

but your options are mostly judder/skips or broken sync

at 30 and 60 fps and the difference it becomes very obvious

Disable the statistics and it already improves a lot also a clear sign and we now improve the stability

Hispasat test 23.976

Potplayer + Potplayer (Framework) + MPC Video Decoder = 1 Frames skipped (No Statistics active)
MPC-BE + MPC (Framework) + MPC Video Decoder = 17 Frames skipped (No Statistics active)
MPC-HC + LAV (Framework) + MPC Video Decoder = 120 Frames skipped (No Statistics active)
MPC-HC + LAV (Framework) + EVR FullSync = 111 Frames skipped (OSD Level 2 Statistics (minimal counter) with Realtime Graph was active all the time)

Framework Components (Parser/Audio + Video Decoder) Calculate different impacts differently mostly scene changes not really impacting the viewing result lot of the differences in the Framework go back to IDR handling.

CruNcher

2nd December 2025, 21:19

Yeah especially that the old parts in MPC-BE outside of MPC Video Decoder fail completely by now in terms of stability, you could as well remove the old bloat completely

flossy_cake

9th March 2026, 12:02

I know we're not supposed to make feature suggestions, but since making feature requests on clsid's github is a bannable offense, where else am I supposed to make one?

Anyway this feature suggestion will knock your socks off so please hear me out.

Imagine this: you've got MPC-HC volume bound to a hotkey and you're currently at, say, 75% volume. You want a little more volume so you increase it to 100%. But you still want more volume because there is a lot of dynamic range in the source and you'd like some normalization/compression via MPC-HC's built in Normalizer feature.

Imagine that the volume key now lets you go beyond 100% and when you do that it activates the Normalizer. So now when you go +1 click of volume beyond 100% it's the same as Normalizer enabled to a value of 110% or whatever 1 click increases it to. One more click after that could be 120% normalized.

Kind of a cool way to conveniently enable/disable the built-in Normalizer wouldn't you say?

Another strategy might be for MPC-HC to read the normalizer setting from file path tags like "C:\My Quiet Video [normalize=200%].mkv" and MPC-HC sees that tag in the path and sets it just when playing that file or any files in the folder if the tag is found anywhere in the path string.

Any good?

:thanks:

lvqcl

9th March 2026, 13:45

I know we're not supposed to make feature suggestions, but since making feature requests on clsid's github is a bannable offense, where else am I supposed to make one?

Nowhere, that's the point. Re-read post #3 in this thread.

clsid

9th March 2026, 14:24

Nowhere. Making requests is pointless because nobody is interested in implementing them. How difficult is that to understand?
And even if someone else would implement crappy functionality, I would reject it, when I an not interested in maintaining it.

There is a hotkey to toggle normalize.
There are hotkeys for volume boost.

flossy_cake

9th March 2026, 15:17

Making requests is pointless because nobody is interested in implementing them.

That's okay, I wasn't requiring anyone to be interested in them. You're free to ignore feature requests. You're under no obligation to do anything.

And even if someone else would implement crappy functionality, I would reject it...There are [already] hotkeys for volume boost.

I see, so the features in MPC-HC's Normalizer page and its associated hotkeys are "crappy functionality"? In that case it would be in your interests to remove this "crappy functionality" from your application!

Unfortunately I can't tell if the volume boost feature (which I'm assuming is a gain multiplier?) is applied BEFORE or AFTER the Normalizer's processing - and this is critical for audio quality in regards to clipping - I'm assuming you care about this kind of audio quality stuff?

Now, if the volume boost is applied AFTER normalizer processing, then there will be clipping, and that would be crappy functionality indeed.

On the other hand, if it's applied BEFORE normalizer, then that would be just the crappy functionality I wanted, in which case thanks!

But I'm not sure it's being applied before, because otherwise there would be no difference between Normalizer's volume boost % and the secondary volume boost slider % (the one with hotkeys). I hope I'm wrong.

:thanks:

flossy_cake

9th March 2026, 15:27

AI reckons boost gets applied to the output of Normalize - doh!

In that case OMG PLEASE CAN YOU GIVE US A HOTKEY FOR ADJUSTING THE NORMALIZE % UP/DOWN IT WOULD BE SUCH LITTLE WORK TO ADD 2 EXTRA ENTRIES IN THE HOTKEY ARRAY COME ON PLEASE YOU HAVE TO DO THIS IT WOULD BE SUCH A BIG QUALITY OF LIFE IMPROVEMENT FOR OUR MEDIA LIBRARIES WHICH WE HAVE SPENT YEARS CURATING

I know you will definitely implement this feature! Thanks in advance!

flossy_cake

9th March 2026, 15:38

So just to explain why toggling Normalize on/off isn't enough... when LAV downmixes surround to 2 channel this naturally causes clipping due to summation of audio channels which can cause them to go above 0db which is an inherent mathematical problem with downmixing.

To solve this without applying a constant and large gain reduction, LAV offers "clipping protection" feature which only reduces gain when clipping is detected in the downmix. Unfortunately it doesn't raise the volume back up to 100% when clipping stops, so you're stuck with quieter than usual dialogue after a loud gunshot caused the clipping protection to kick in.

So I switched to using MPC-HC's Normalize set to 100% combined with "regain volume" which does the same thing LAV is doing except the "regain volume" raises it back up to 100% afterwards so you don't get quiet dialogue after a loud gunshot like you do with LAV's solution.

So it is superior to LAV's clipping protection and I want to use it on everything all the time.

So that's why I need Normalize 100% to be enabled by default, and so enabling/disabling Normalize via hotkey wouldn't be of any use because it always has to be enabled for me.

If you could add a couple of hotkeys to adjust the Normalize % that would allow us to hit those hotkeys to make the adjustment for quiet sources.

And if by some chance you are willing to do this please could you implement it in the same way as the Audio Switcher's delay hotkey is implemented:

status bar at bottom of GUI window tells the user what it was adjusted to (just like audio delay hotkeys do)
the setting gets reset to default at the next MPC-HC instance (just like audio delay hotkeys do - or maybe this one should stick? opinions?)

:thanks::thanks::thanks::thanks::thanks:

flossy_cake

9th March 2026, 15:55

Unfortunately I can't tell if the volume boost feature (which I'm assuming is a gain multiplier?) is applied BEFORE or AFTER the Normalizer's processing - and this is critical for audio quality in regards to clipping

Tested this now with Avisynth ColorBars() which outputs a sine wave audio signal and can hear it clipping when raising Audio Switcher's Boost, so AI was right about that.

So yeah, please give us those 2 hotkeys clsid! Pretty please!

Sunspark

9th March 2026, 16:11

I think you need to start looking into that Equalizer APO mixer. It has hotkeys for preamp with configurable steps of gain.

flossy_cake

9th March 2026, 16:49

I think you need to start looking into that Equalizer APO mixer. It has hotkeys for preamp with configurable steps of gain.

Thanks but I run a very tight ship and would only use something like that as a last resort.

I'm not a person who likes to install a lot of helper apps that are all hooking into various things in the OS.

The only exception to that is Autohotkey cause I spent years working with it and had discussions with the developer Lexikos and know exactly what it is and is absolutely essential and fundamental to the way I use Windows.

MPC-HC already has the built in audio processing that I want, I just need a hotkey for one of the features that's all.

Anyway it's up to clsid now whether he wants to add those hotkeys to the list.

If he doesn't want to, I stand by my original statement that he is under no obligation to do it.

And obligations and virtues are two different things - it's virtuous for me to help an old lady reach something on the top shelf at the supermarket, but I'm not obligated to do this.

hello_hello

12th March 2026, 17:01

So I switched to using MPC-HC's Normalize set to 100% combined with "regain volume" which does the same thing LAV is doing except the "regain volume" raises it back up to 100% afterwards so you don't get quiet dialogue after a loud gunshot like you do with LAV's solution.

Doesn't "regain volume" after a loud peak increase the volume again fairly slowly? Slow enough to hear it's slowly getting louder again? I haven't used it for a long time but that's how I remember it.

It's a pity MPC-HC can't tap into ffmpeg's filters so there'd be a choice of compression methods. MPC-BE lets you use ffmpeg's compand filter, which is way better than the built in normalizing if it's configured correctly, but for some reason it's the only one.
SMPlayer/MPV lets you tap into most of ffmpeg's video and audio filters, even the VaourSynth filter if MPV is built with VapourSynth support.
It's unfortunate that MPC-HC will never be able to do it, but maybe there's some sort of DirectShow audio compression filter?

flossy_cake

16th March 2026, 08:33

Doesn't "regain volume" after a loud peak increase the volume again fairly slowly? Slow enough to hear it's slowly getting louder again? I haven't used it for a long time but that's how I remember it.

I've only tested on that one breaking bad scene with the gunshot that clips at +5db (https://forum.doom9.org/showpost.php?p=2028811&postcount=25480/) and the decay back to full volume wasn't audible. If you have that show in your archive it's the final scene of S03E07

flossy_cake

21st March 2026, 19:26

Doesn't "regain volume" after a loud peak increase the volume again fairly slowly? Slow enough to hear it's slowly getting louder again? I haven't used it for a long time but that's how I remember it.

Just heard a source that pumped pretty badly with it set to 200 or 400. So I guess that's a no-go and I'll just have to leave it at 100 and stick to Dolby DRC profiles applied via avisynth.

If clsid is interested in putting Dolby DRC profiles (the ones like "film standard fast", "movie light slow" , "speech", etc.) into MPC at anytime in future, please PM me and I'll hook you up with all the db curves and attack/decay times (courtesy of tebasuna).

I presume MPC already has some parameters in the code for the normalizer so maybe it's just a matter of plugging in these Dolby values. I'm probably over simplifying.

Sunspark

21st March 2026, 20:42

flossy_cake

22nd March 2026, 10:58

What is my TV doing differently compared to the PC? When I was running the sound out of the 3.5mm to speakers, I needed to use the soundcard's compression in shared mode to make dialogue louder, etc. but now that that machine is connected to the old TV and I'm sending the pcm audio to it over hdmi using exclusive mode (even turned on AC-3 DRC for good measure though almost nothing these days is ac-3) and the volume is fine. I don't need to use the compression anymore. Is the TV compressing it? (I have no special settings turned on), or does it scale things differently?

If the TV is compressing it there is usually an option for that on the TV. Or maybe the TV has that turned on all the time to better suit it's built in speakers.

Another possibility is the 3.5mm audio cable had some attenuation . If I connect my headset through a "line level" cable for speakers vs a "headphones level" cable I get dramatically different volume.

Is LAV doing the down mixing in both setups? If not, then it could be due to different downmixing , eg. LAV vs windows audio driver.

Sunspark

22nd March 2026, 17:35

Yes, LAV is doing the down-mixing in both scenarios. I just find it curious. The 3.5mm output on the NUC is probably headphone and the 3.5mm on the speaker is possibly line (there is no labelling) but it was designed originally to be used with 1990s computers so while that might imply headphones, sound cards back then also had line..

I do have a USB-C to 3.5mm dongle and the volume is the same with the dongle on an entirely different system.

I do have stable sound turned off on the TV. It's not a big deal, I just find it curious that there is a difference.

I'm currently thinking that the issue is headphone vs line level, a dongle won't fix it, and that it would need to be standalone usb dac that can do line-out.

huhn

22nd March 2026, 18:35

a dac power conversion can have a different volume all on it own. TV at least new ones usually do headphone. the last i at least tested shortly is LG CX is is headphone and doing a crossfeed you don't apply a crossfeed to speaker...

Sunspark

26th April 2026, 21:28

Tip for those with old systems: If you're working with a 10-bit file with .ass subtitles, put a check in use libass in the options for subtitles default styles. It'll double the fps from 9 to 20+. It does seem like this path is more performant than the default one.

mzso

28th April 2026, 01:39

Hello!
Why don't any renderers have decent SW fallbacks? Right now it seems my vid card refuses to work. But every single renderer that even works in this case result in screen tearing on the video.
Is there such a thing as a SW video renderer? (That's intended to be such)

At least I could play FHD video as long as the display is set* to FHD. It only requires some chroma upscaling.

*"set" is a strong word since the cursed windows generic drive doesn't allow changing resolutions. So I get FHD if the videocard is already disabled, or UHD if windows tries to use it, but then disables it with code: 43.

huhn

28th April 2026, 06:30

cause tearing is a hardware issue and the software side can be as good as it wanted to be it will never fix it.

mzso

28th April 2026, 10:33

cause tearing is a hardware issue and the software side can be as good as it wanted to be it will never fix it.

Ah, disappointing. I guess it could only be improved in the fallback driver, if that even can do anything.

Sunspark

28th April 2026, 13:25

He's saying to you that you need to understand and fix why you have tearing. This is a Windows app, there is no tearing unless you turned off the Windows compositor.

mzso

29th April 2026, 09:49

He's saying to you that you need to understand and fix why you have tearing. This is a Windows app, there is no tearing unless you turned off the Windows compositor.

Right now I don't have a working GPU*, so only "Microsoft Basic Display Adapter" provides video, with no HW acceleration. But I could watch 1080p on CPU power only. But the tearing is bothersome.

So as I understand I can't do anything to improve tearing.

*Got disabled with this message: "Windows has stopped this device because it has reported problems. (Code 43)"

clsid

29th April 2026, 10:57

The generic driver probably doesn't do proper VSync.

Use Display Driver Uninstaller to remove your driver and re-install it fresh. Pick an older version with known good stability.

Sunspark

29th April 2026, 14:23

Another option is to match your display's refresh rate to exactly the fps of the video, likely 23.976.

DDU as clsid recommended does a good job of cleaning out driver residue. Run it in safe mode as the instructions recommends for a thorough job.

huhn

30th April 2026, 18:17

not how vsync works. if you have to real GPU driver you are just wasting your time with broken hardware.

hajj_3

3rd May 2026, 08:52

Dav2d decoder has been released: https://code.videolan.org/videolan/dav2d

This decodes the upcoming AV2 video codec. Hopefully support will be added for it.

tormento

10th May 2026, 11:30

Does MPC-HC support Film grain synthesis (FGS)?

If positive, AV1 only or HEVC too?

DvineLord

8th June 2026, 19:35

So I tried MPC-HC.2.7.2.x64 and with a loaded playlist I see a horizontal scrollbar, if i maximize/unmaxmize it goes away as it shouldn't be present in the first place. It didn't appear in MPC-HC.2.7.1.x64 and earlier builds. I traced the beginning of the occurrence somewhere in-between MPC-HC.2.7.1.31.x64 (not present) and MPC-HC.2.7.1.36.x64 (present).

I assume the changes here might of caused it!?
https://github.com/clsid2/mpc-hc/commit/419dadc92018eae349e4c475a19d59e1fa83e90b

GeoffreyA

10th June 2026, 13:43

I finally moved to 11 LTSC from 10 Pro. When opening a video for the first time on boot, it takes quite a few seconds to play after MPC-HC opens. Never once experienced this on 10. Any ideas, anyone?

clsid

10th June 2026, 14:04

If delay happens just once, it might be related to font loading. Windows bug.

Test file without subtitles, test once with null video renderer, and test once with null audio renderer. That should narrow down possible causes.

You can use Process Monitor to see what happens during the delay.

GeoffreyA

10th June 2026, 17:09

Thanks, clsid. I'll try to narrow this down soon.

GeoffreyA

11th June 2026, 20:05

I think Defender had a hand in it, not disabling itself properly with AVG installed. After disabling Defender, the first opening of a video sped up, but there was still a slight delay, present even with no subtitles and null video and audio renderers.

clsid

12th June 2026, 11:42

AVG itself probably also adds some delay on first load of stuff.
I dunno if LTSC has Smart App Control, but that also checks files before loading.

GeoffreyA

13th June 2026, 14:10

Disabling Smart App Control did the trick. Now, first opening is at 10's level.

Z2697

13th June 2026, 16:16

LOL typical Microslop feature...
BTW How can you betray us by upgrading to 11??? I'm infuriated

GeoffreyA

13th June 2026, 17:47

LOL typical Microslop feature...
BTW How can you betray us by upgrading to 11??? I'm infuriated

Shocking, I know! And me, one of the biggest haters of 11 on the planet :-)

Well, recently, I've been testing 11 in VMs, both LTSC and consumer (insider), and I thought, "This thing doesn't look half bad any more." So, the seed was planted in my head. After backing up, I installed 11 LTSC—alarmingly, it was fast and streamlined. However, I took it out and put on 10 LTSC: also fast but I wasn't too happy. Then, strangely, 10 22H2 felt a bit sluggish. Placebo effect? In the end, I formatted and installed 11 LTSC (activated as IoT), and it has been running well, apart from edge cases: Defender and Webview2 running in the background (fixed), Explorer's toolbar drawing slowly, etc. I'm starting to appreciate the UI, which feels simple and polished, and Settings is more coherent than 10's version. I think the curved corners got me. Maybe I was replaced by a doppelganger :)

https://i.imgur.com/xUKb6e7.jpeg

oniiz86

15th June 2026, 06:59

DvineLord

17th June 2026, 06:31

@DvineLord Have you tried to see if this issue has been fixed with the latest MPC-HC.2.7.2.16.x64 build release? Maybe @clsid has already addressed this regression.

I have been checking the commits on github and didn't see any relevant ones, I also just updated to MPC-HC.2.7.3.x64 and the issue is still present.

oniiz86

17th June 2026, 08:37

I have been checking the comments on github and didn't see any relevant ones, I also just updated to MPC-HC.2.7.3.x64 and the issue is still present.
@DvineLord Oh okay, perhaps you can open this issue over github to receive a more proactive response, I figured @clsid would have addressed your concerns over here.

JarrettH

20th June 2026, 03:45

Is there a way I can inverse telecine in MPC? Just want to watch some DVDs and playback correctly at 23.97. Doesn't look like anything in LAV will do this either. Thanks. Running NVIDIA if it matters.

huhn

21st June 2026, 10:19

madVr has realtime IVTC.

JarrettH

21st June 2026, 15:56

Ohhh, I forgot about that option! Ok that will do! :devil:

peter0815x

12th July 2026, 21:28

hello

i have a small question
(its not a feature request, just a question)

i use mpc hc also for audio files
and with mpc audio render and wasapi exclusiv audio playback is identical to foobar2k
is there a way to modify the information that is shown in the status bar?

with active information
mpc hc shows titel, author and copyright.

and this is shown under properties:
Audio: FLAC (framed) 88200Hz stereo 4233kbps [A: flac, 88200 Hz, stereo, s24]
is there a way that this info can also be shown in status bar ?
for eg. flac, 88200 Hz, stereo, s24

is there a way to modify the status bar without changing source code?

thx peter

tebasuna51

13th July 2026, 09:26

1) View -> Options -> Player -> User Interface -> Status Bar Elements -> Audio Info

That show a limited info, not all info showed when push Properties button obtained with:

2) View -> Options -> Player -> Toolbar Layout -> Add Properties button

Modify this seems a feature request to the source code.

peter0815x

13th July 2026, 16:59

thx for the info with the toolbar

yep seems i have to edit the source code
google ai even told me more or less what to do
to get the info from MediaInfo.dll and include it into the status bar

only a few lines of code (and google ai said it can do it for me....)
but i would have to do this with every new mpc hc version.
guess not worth it and the properties button will work :-)

DvineLord

24th July 2026, 13:34

huhn

24th July 2026, 14:14

yes looks normal.

could do an d3d9 -> d3d11 interop instead.

v0lt

25th July 2026, 04:51

Using MPC-HC.2.7.4.x64 on Windows 7, is this normal behavior when playing H264!?
MPC Video Renderer D3D11 + LAV DXVA2 (native) = Software Decoding
MPC Video Renderer "No" D3D11 + LAV DXVA2 (native) = Hardware Decoding
MPC Video Renderer D3D11 + LAV DXVA2 (copy-back) = Hardware Decoding
MPC Video Renderer "No" D3D11 + LAV DXVA2 (copy-back) = Hardware Decoding
Don't use D3D11 in Windows 7 because it doesn't support video decoding (https://learn.microsoft.com/en-us/windows/win32/api/d3d11/nn-d3d11-id3d11videodecoder) and processing (https://learn.microsoft.com/en-us/windows/win32/api/d3d11/nn-d3d11-id3d11videoprocessor).
Don't use CopyBack decoding because you'll lose performance.
Use D3D9 and DXVA2 for maximum performance in Windows 7.

Schwartz

26th July 2026, 19:06

Do use CopyBack to fix driver issues. Direct is not always stable.

v0lt

27th July 2026, 04:15

Do use CopyBack to fix driver issues. Direct is not always stable.
I haven't had any decoding issues that were resolved by CopyBack mode. There were a few rare files that only played correctly with NVDEC (CUVID) or a software decoder.

Disabling DXVA2 Video Processor (settings NV12, P010, etc. in MPC VR) might be sufficient.

clsid

27th July 2026, 13:11

The only actual real-life stability issue with Native mode that anyone should care about is the combination of D3D11 Native and MadVR. That can hang/deadlock after (many) seeks due to a bug in MadVR where it doesn't free up surfaces for the HW decoder to use.

JarrettH

28th July 2026, 06:27

Shocking, I know! And me, one of the biggest haters of 11 on the planet :-)

Well, recently, I've been testing 11 in VMs, both LTSC and consumer (insider), and I thought, "This thing doesn't look half bad any more." So, the seed was planted in my head. After backing up, I installed 11 LTSC—alarmingly, it was fast and streamlined. However, I took it out and put on 10 LTSC: also fast but I wasn't too happy. Then, strangely, 10 22H2 felt a bit sluggish. Placebo effect? In the end, I formatted and installed 11 LTSC (activated as IoT), and it has been running well, apart from edge cases: Defender and Webview2 running in the background (fixed), Explorer's toolbar drawing slowly, etc. I'm starting to appreciate the UI, which feels simple and polished, and Settings is more coherent than 10's version. I think the curved corners got me. Maybe I was replaced by a doppelganger :)

https://i.imgur.com/xUKb6e7.jpeg

Had to look up this artist of your wallpaper - wow!:thanks:

GeoffreyA

28th July 2026, 11:06

Had to look up this artist of your wallpaper - wow!:thanks:

None other than the great John Constable, a British painter of the 19th century known for his tranquil landscape paintings. The Hay Wain and The Cornfield are probably his most iconic works.

Sunspark

29th July 2026, 17:11

If using IVTC you do need copy-back, but if you aren't playing DVD media with that option then you wouldn't need that.

For the rest of it, hasn't it been said for years now that you need copy-back with madvr if you want proper rendering? Specifically that renderer. Seems strange to me to see you all saying to just use native now. I agree native is fine for other renderers like evr-cp.

My concern and where I am coming from with this, I believe I do see very subtle differences in blacks or colour in the different decoding methods. People disagree that it exists but why can't there be a difference, the driver plays a role, it's a mathematical formula and calculations happen at different stages in the process depending on what method is used?

PS. There is one additional hardware difference as well, if you're using a discrete video card then yes I agree the data needs to travel across the PCI bus, but if you're using an IGPU, then it's different because the data doesn't leave the CPU because the GPU is integrated and both share the same RAM, it doesn't travel to a different set of RAM on a different bus. This is my understanding of it.

huhn

29th July 2026, 18:08

if they are there make a screenshoot.

nevcairiel

29th July 2026, 23:22

- Any native with madVR has some feature limitations, I believe black bar detection doesn't work. Not sure if others.
- DXVA2 Native with madVR has some issues because DXVA2 does not offer direct shader access to the pixel data (its really only designed to be used with a DXVA2 video processor), so madVR has to do some workarounds to get the data which are not always fully pixel exact.
- D3D11 Native with madVR has some issues because madVR is essentially a D3D9 renderer (its that old), and while it mostly works, there is some issues and oversights (also, I believe DXVA Deinterlacing doesnt work)

Essentially, pick your poison. Personally I used copy-back when I was still using madVR on the daily.
A more modern renderer D3D11 Native would be perfectly fine. Well, as long as it also wouldn't do a CPU-driven black bar detection algorithm, but instead do it on the GPU.

clsid

30th July 2026, 00:49

The discussion started with a question regarding MPCVR. And with that renderer, native mode is fine for most use cases.

In case of MadVR, copyback mode is indeed preferred, for several reasons as outlined above.

Since v2.7.4 MPC-HC will actually prompt asking to change HW decoding to copyback mode when you change to MadVR in the settings.

huhn

30th July 2026, 03:19

v0lt

30th July 2026, 04:30

dxva processor and d3d11 processor have so many bugs just move on...
Often, users themselves incorrectly configure the driver when they select the full range for the video and force it.
Because they don't understand the difference between video and display parameters.

Here are the default video driver settings (https://i128.fastpic.org/big/2026/0730/dc/c5490b00b947449ee8a39b7468f444dc.png). There are two important points.
1. The player controls the color. There's no need to force anything in the driver.
2. The color range is set to limited (16-235). This setting is grayed out, but it will be applied if the player doesn't report the color range.

huhn

30th July 2026, 11:03

and that fixes broken scaler and misplaced chroma somehow?

range issue where last part of the AMD side i'm not sure if it even has any range settings.

Sunspark

30th July 2026, 17:29

v0lt, hope this question is considered on-topic. Where mpcvr is concerned, if one is using shaders for scaling instead of the card's dxva2/d3d11 video processor, is it considered ideal to turn off native decoding for the shaders and texture size? My thought process here is that shaders are a copy-back operation, especially if in D3D9, but every renderer is different. mpcvr does accept copy-back input, but is it better in that usage scenario or it makes no difference if you're using native decoding and shaders?

One reason I had been using shaders is because it worked correctly for video levels and the vp was making blacks grey, other renderers were not doing that. Fixing it in driver settings, made the other renderers break, so I only used shaders with mpcvr.

huhn

30th July 2026, 17:32

no use of a shader is a copyback operation.

pure shader are created to be run on the GPU.

clsid

30th July 2026, 17:37

@sunspark
I assume you are using AMD with its buggy DXVA2 processing?
You should switch to D3D11, unless you run Win7 or ancient hardware.

v0lt

31st July 2026, 04:07

Where mpcvr is concerned, if one is using shaders for scaling instead of the card's dxva2/d3d11 video processor, is it considered ideal to turn off native decoding for the shaders and texture size? My thought process here is that shaders are a copy-back operation, especially if in D3D9, but every renderer is different. mpcvr does accept copy-back input, but is it better in that usage scenario or it makes no difference if you're using native decoding and shaders?
If your hardware decoding is working correctly, I see no reason to disable it.
If you're not happy with the DXVA2/D3D11 Video Precessor, disable its formats and use shaders for color conversion. CopyBack isn't needed for this.

But keep in mind that without the DXVA2/D3D11 Video Precessor, you may get poor deinterlacing.

hajj_3

31st July 2026, 06:51

has anyone reporting the AMD issues to AMD to try and get them to fix them?

Schwartz

31st July 2026, 10:07

AMD bug reporting was a joke 8 years ago when I tried it and I'm sure nothing has changed. It's a ghost town, sadly. NVIDIA not much better.

It's reassuring that all the stuff works with MPCVR. I find it hard to give up the excellent scalers and doublers available in madVR. SuperRes is a very blunt tool by comparison.

Sunspark

5th August 2026, 20:55

@sunspark
I assume you are using AMD with its buggy DXVA2 processing?
You should switch to D3D11, unless you run Win7 or ancient hardware.

Actually no, the machine I am making this observation on is Intel Broadwell.

These days it spends most of its time in Win 7 so that exclusive mode can be enjoyed more fully.

I should test to see if the issue exists also on the Win 10 partition.

Sunspark

5th August 2026, 21:05

I had an interesting observation to make the other day vis-a-vis the two renderers.. sometimes a file is tagged wrongly.. a movie I was watching, the madvr HUD was reporting it as 47.952.. which doesn't really make sense (it has happened once before too that I noticed) since afaik no streaming service sends media out at that rate, and it almost certainly was 23.976. Manually forcing the desktop to 23.976 Hz, it played fine. Loading the same file in mpcvr, it didn't report 47.952, instead it reported 24.000 and was trying to make the frame time correction go up to 24 instead of staying at 23.976.

DvineLord

10th August 2026, 04:57

v0lt, so i recently ticked dynamic range in driver settings and set it to full (0-255). are you saying the player will change it to full if needed and just keep driver settings at defaults!? i'm using very old hardware : amd radeon hd 5670, phenom ii x2 555, amd 780g chipset, 4gb ram under win7 64bit with catalyst-15.7.1 drivers. i dont use any shaders.

clsid & adipose, thank you for fixing the horizontal scrollbar playlist issue!!! looks like y'all made alot of changes to recents and options, so far pretty good. =)