>>>>>>>>exe

>>51628121
Madshi is completely butthurt about mpv, he outright claimed that the mpv devs stole his algorithms (somehow). When people suggested he open source his code, he began rambling about how his work will be stolen.

What a sad guy, afraid that if he lets others modify and fork his code he could be deemed irrelevant.

>>51628140
Yeah it actually runs better on Windows than Linux because of the graphics card driver support.

There's no fancy guide for laymen to follow so they feel like an advanced hacker.

Is lanczossharp still the best scaler?

>>51628121
>Copy paste premade configs and tweak it to desired

where?

>>51628163
Then explain why mpv works better on my Linux than it does on my Windows?
They're both installed on the same machine, with Linux using open source drivers.
Explain that wintard.

>>51628121
>install KCP
>takes 5 seconds

>install mpv
>idk but I never installed it and don't want to because with lacks nnedi3

>>51628121
mpv takes weeks; it's so good you need to tweak the config file, but of course you need to make it first

not to mention you need to change the config file for different kind of files

>>51629723
mpv has nnedi3 dude...

>>51629723
>install KCP
Have to go to haruhichan download some NEETs shady .exe file. Greeted with a windows 95 clusterfuck UI, options nestled within options. Consumes needless amounts of system resources.

>install mpv
Open a terminal window, type "brew install mpv". 20 seconds later all video filetypes are already associated, superior quality, minimal resources. Changing settings is done with a simple text file and requires minimal English knowledge.

>>51628121
>scared of mpv
I'm not. VLC fits my needs.
For the moment, I don't see any reason to switch to mpv. If there is any, I'll consider using mpv.

Oh and I didn't see any way to create/manage playlists under mpv.

>>51628157
Source?

I don't blame him for not open sourcing. For one, it's written in delphi. You have to dick around with build instructions and might be tempted to write things in a more accessible way. You need to deal with forks forming everywhere. People want to merge changes into your codebase.

The feeling sucks. I don't blame him whatsoever, different people like working in different ways. The only time I would ever provide source for anything I made, is if it would substantially benefit others and there was some function reason for it (eg a compression algorithm), or if I quite literally, never planned on touching it again. Open source workflows are a hassle, and surprisingly I work more poorly when I feel like someone is looking over my shoulder. Not the opposite.

>>51628121
>>MPC-HC full install + LAV Filters + madVR + xyVS/xySub-filter + ReClock
Serious question. Why is all of this necessary or even important?

File: IMG_0359.jpg (2 MB, 2800x2100) Image search: [Google]

2 MB, 2800x2100

>>51630182
Same reason we have audiophiles. Placebo.

>>51628121
People still use reclock?

>>51629794
>mpv takes weeks
>it takes weeks to put vo=opengl-hq in a config file. WEEKS!

>>51630182
It's part falsehood, part exaggeration.

LAV are more less ffmpeg based decoders. MPC-HC comes with them bundled as the default, and has for quite some time. You don't need to install them yourself.

madVR is a renderer that's more configurable and better quality than EVR custom presenter, etc.

xy subfilter is a subtitle renderer.

ReClock... I'm not entirely sure how it does what it does. But it more or less stretches audio to better deal with frame timings. Hence "re"clock. It makes timings more consistent and tighter. I've never used it, and never saw a need to. Maybe I never saw a need because I never tried it, don't know.

I don't know why people make this more complicated than it really is. I built newer versions of mpv every once in a while. Don't see much difference, and kind of prefer madVR / mpc-hc. If I was running Linux, I'd obviously use mpv.

>>51630237
Thanks. Is all of this necessary because the stock player is deficient somehow? Like, are there rendering bugs or inqccuracies?

I've been using MPC-HC for a while now and have never run into any issues with it.

>>51628121
Why the FUCK does interpolation require video-sync now?

>>51630262
None of it's really necessary if you just want to play files. The internal subtitle renderer usually works well enough, and the bicubic resampling used by EVR-CP is okay.

The main benefit of xy-subfilter is better quality subtitles, and also, it looks ahead and prevents hitches if more complex instructions are ahead and it has to load some file, font, figure out an animation, whatever. They're also somewhat sharper and scale more cleanly, depending on how you want them (works better with madVR features as well).

madVR has better colorspace conversion. Obviously your monitor works with red green and blue primaries, and those colors are probably 8 bits each. Most videos are in YCbCrm with the chroma planes (Cb and Cr, Y usually refers to luminance / brightness) subsampled, and therefore must be converted to RGB. Doing so introduces quantization error when you go to reduce the result to whatever integer it's closest to. madVR deals with this better, in my mind. You might remember years ago playing back videos and seeing massive bandung everywhere, that was caused by just rounding those values and not dithering.

The chroma planes are subsampled, so they need to be upscaled to the same resolution as the video. Having madVR just means being able to choose more finely what algorithm is used for what, and when. mpv allows more or less the same. I've stuck with "jinc" (non-sigmoidized ewa lanczos) for years now. It usually looks the best to me. mpv allows you to use sigmoidization though and configure it. etc.

Don't know anything about reclock. I'm one of those "purists" everyone hates. If something doesn't need to be altered, it should be left alone, even if the consciousness is not directly aware that alteration.

>>51630356
Damn. Thanks for the spoonfeed, anon. I appreciate it.

>>51628157
>When people suggested he open source his code, he began rambling about how his work will be stolen.
He actually said that he mainly wants people to use his renderer and not others', and that's the reason he keep it closed. Which could be an egoistic reason, but well, it's his code, so he can do it whatever he wants with it, we are not entitled to it.

>>51630154
>Source?
http://forum.kodi.tv/showthread.php?tid=223175&page=153
Almost everything he said is half truths, bullshit and FUD, so I really doubt he's not butthurt like he says.

>>51630295
Because by doing so they removed the hack they used to time interpolation to vsyncs, which is a major code's simplification.
It sucks, because I can't use interpolation anymore, but from a thecnical standpoint is understandable.

>>51631399
>Almost everything he said is half truths, bullshit and FUD
Such as? It all seems fairly solid to me.

You said it yourself, it's his code and his project. Don't like it, don't use it. If anything he seems to feel as though his project and motivations are being misrepresented. The more people using his renderer, the more notoriety, the more suggestions and bug reports. I'm slowly writing a compression algorithm (system is a more accurate term), I don't really know what I'll do when it's finally done. So this all really catches my interest.

I can't say much about what I think the driving factors behind his psychology are, beyond in the most generic senses. The deeper you try to go, the more apt you are to only end up mostly describing yourself.

I want you to know I understood nearly nothing of that lingo but I would install whatever you d recommend.
So, is pure mpc-hc enough? How much difference does madvr make?

>>51631789
Quite a bit if you have the hardware.

http://wiki.mikejung.biz/MadVR_Image_Doubling_720p
http://wiki.mikejung.biz/MadVR_Image_Upscaling
http://wiki.mikejung.biz/MadVR_Chroma_Upscaling_1080p_Image_Quality (IMO the most noticeable improvement).

You may notice the difference when you start using better scalers if you don't right now. These are just representative screenshots. However I would still suggest mpv because it comes with all of this stuff stock and has as good, if not better (as admitted by MADshi himself) implementations of the scaling algorithms. With MPC-HC you would need xyVS filters and madVR to get the same quality as mpv.

>>51631590
I wasn't referring to that part.

He said that mpv copied his algorithms, which is false. jinc-based scaler were implemented in imagemagick ages before they were in MadVR, and he didn't develop them. Same for NNEDI and super-xbr, whose authors are credited in their respective source. Or maybe he thinks that being the first to implement them in a video player gives him some sort of paternity or last word on those ideas, but then it's frankly ridiculous.

He said that they discovered a bug with his code and they didn't report it to him. Half truth: haasn discovered a slight problem with a default in MadVR's jinc (which mpv itself has), that was easy to solve with mpv because the radius can be fine-tuned, while in MadVR you're stuck with a few presets.

>>51631590
>>51631995

He also seems quite butthurt by the fact that they (the developer team) supposedly started to advertise mpv as superior to MadVR on "various forums"* online. A search on the archive will show you that the only dev that used to come here, haasn, always said that mpv was on par with MadVR or "superior" on specific features, always presenting proofs or facts to back up what he was saying. And he actually ammitted that MadVR's antiringing and dithering algo are sightly better. Of course madshi didn't say any of this, nor he showed anything to back up his statements.

Then he said that MadVR has better quality anyway without stating where or why or presenting any sort of proofs.

*and by this he quite clearly means /g/. mpv is otherwise discussed almost only in linux's forums, where MadVR is not even an option. And while here mpv's fanboys tend to be quite obnoxious, MadVR's partisans are pretty much the same, so he doesn't really have too much to recriminate

>>51628121
I recently went back to mpc-hc, I don't really dislike mpv and use it on linux but at the same time I don't see much point in using it on Windows. mpc-hc is really solid overall and I fucking hate those on screen controls in mpv. I'm not going to use FLOSS player for the sake of using it on propriety botnet OS.

>>51631590
>It all seems fairly solid to me.
He's claiming there are stolen algorithms, which is bullshit. Almost all of the algorithms that mpv/madVR use were things before madshi or wm4 were born. He did not invent lanczos, nnedi3, super-xbr, jinc, bicubic, spline, mitchell, or any other scaling algorithm in madVR/mpv. He took them and put them into his program, just like mpv did. Maybe he's mad because a certain implemenation was "copied" and improved, but how do wm4/haasn know if they don't even have the source code for madVR?

>>51631995
I'm aware of all that. I've watched madshi's posting and inferred probable thought style, and none of it really adds up. I'd like to know what he actually means by what he said, because it is an absurd claim to make if you take it as it looks. Which is a bit out of character and suggests there might be more to the context.

I understand his desire to want to widdle down settings into the most useful states and leave them as a few preset, but I personally dislike him gradually removing more fine grained control. Which has been the pattern for a while now.

I'd have to dig and spend more time than I'm really willing to. At face value the situation doesn't make much sense.

>>51632057
>I fucking hate those on screen controls in mpv

script-opts=osc-layout=bottombar

>>51632073
>but how do wm4/haasn know if they don't even have the source code for madVR?
Reverse engineering is a thing, and it isn't too difficult if you're able to watch some of what the hardware is actually doing between two points. Don't know why they'd bother though.

>>51632022
Seems you think it makes as little sense as I do. He has no obvious reason to say or do most of what he did, even if you infer a severely butthurt psychological status.

File: 1353696142708.jpg (141 KB, 906x1024) Image search: [Google]

141 KB, 906x1024

>>51628157
All mpv devs do is wait for his release notes to copy his ideas. After he's gone through the process of thinking up new ideas, sifting through them for suitability, and finding a workable implementation, mpv comes along, plucks it out, and calls the idea "theirs."

If madVR didn't exist to copy, mpv would just be WMP with a different skin.
He has no reason to open the source. Meme player fags don't give him any credit, so why should he make their stealing of his ideas as easy as copy and pasting?

mpv has innovated literally nothing. It takes MPC-HC (poorly) and madVR and makes it cross-platform. That's it.
I don't get the 24/7 shilling. What's the fucking point? MPC-HC, LAV filters, and madVR are still in active development. They're still giving us new features. Why shill mpv trying to go for some static "win." "Oh look, due to our incessant rhetoric, now fewer people are using MPC-HC and madVR, making it so those developers have less incentive to keep going with their projects. THIS MEANS WE WON!"
You won what? What does the media player ecosystem gain by moving to a player whose devs have literally no ideas of their own?

Let's face it, the only reason you Linuxfags shill mpv is because you hate being reminded that all you have is a second-rate clone. "If MPC-HC and madVR go away, that means we'll have the best!"
Grow up.

>>51632665
go to bed madshi

>>51628121
>he doesn't use ffplay

stay pleb normie :^)

>brand loyalty on fucking freeware
You have a brain disease. Shut the hell up and use what plays YOUR videos the best. There is no single solution; mpv is only one of my media players.

>>51628121
MPC-HC doesn't need that shit in ages. It's literally just install and play.

>>51632665
This has been my general observation as well. Everything mpv has, madVR had first.

Who spurred what ideas? What made what, do what, when, and why? I don't know. It could be that mpv is just developing and we're observing the process of coming to some level of feature parity with madVR, but nonetheless. mpv's main strength is the granularity of its configuration potential. Beyond that it hasn't brought much of anything novel, and I'd like to see that change honestly. Its fruit dithering is technically inferior in theory, even if it's perceptually near transparent relative to error diffusion.

Every mpv thread I've been in, and there have been a few, I've constructed many comparisons (which to put it bluntly, is a lot of fucking around to do properly. madVR's overlay / fullscreen exclusive. Chronic pain dulling the mind etc) and the most I've seen is just people generally talking shit about madVR. Nothing is absolute. Nothing is standalone. The whole goddamn thing is always relative to madVR, it's always a comparison. Even people sharing configurations is centered around it. You can't deny these tendencies, and I don't know if madshi browses /g/ (can't imagine why an actual developer would), but I wouldn't blame him for calling it out. It really reflects poorly on mpv if anything, and that's a real shame as well.

Fuck /g/, in general.

Eh, call me a pleb if you like but I can't be assed to do a lot of config tweaking for a fairly minor (in my opinion) difference in appearance in video playback. I've used mpc-hc + madvr + whatever other shit is recommended in windows and I've used both mplayer and mpv in linux; they all do just fine in my opinion. I never even consciously noticed or looked for banding in my videos before I saw it talked about on /g/ and even now some minor banding doesn't bother me all that much. This shit just doesn't matter to me as much as it does some other people I guess.

Hell, the only reason I stopped using VLC player on windows and linux was because it gets shit on by everyone on /g/ so much....

>>51633644
The more you know, the more you notice when things are wrong and can be better.

For example, think most people notice when source content has been upscaled? Think they notice and focus on pulldown choppiness during pans? Think they notice interlaced credits, or interlacing in general? Think they notice color correction? Think they notice compression artifacts?

jpg artifacts on line art bothers the hell out of me. I don't like it one bit. You know why? Because I know. I know it doesn't have to be there. I know it can be better. And I know how to make it better.

I know what the right way is, and I know what I see is not it. And that's when it takes root. Once you catch it, it cannot ever be truly uncaught.

>>51632665
>>51633471
I remember CoreAVC praised like MadVR back in the day. We all know the outcome

>>51633797
You're comparing a decoder to a renderer though. madVR isn't new, it's been around quite a while. Formats come and go, but a truly solid rander is apt to stick around until it's abandoned, or something better comes about.

>>51633823
>rander
renderer*
Fuck, what's wrong with my brain today.

>>51633823
So? You are comparing a full fledged player to a renderer.

>Formats come and go
So do developers, madshi won't be around forever. MPC-HC, ffmpeg, lav filters, mplayer/mplayer/mpv are all open source and they will be here to stay in some form. They are improved upon by hundreds of different people.

>>51630101
People always complain about the side features being bad, but VLC's fatal flaw is that it plays video badly. mpv with opengl-hq just looks better. VLC's video is also horribly artifacted when seeking.

HI guys I get on mpv.

Invalid video timestamp: 15.340000 -> 15.240000

And the video stops when this thing is printed, and continues. I tried vlc and it doesn't stutter.

>>51633910
THis is what I am getting with vlc.

 core warning: picture is too late to be displayed (missing 100 ms)

>>51633889
>So? You are comparing a full fledged player to a renderer.
No. I'm comparing the rendering components of a full fledged player to another renderer.

>madshi won't be around forever
No one will, and they don't have to be. Madshi has worked on his projects for years, and hasn't expressed any inclination to stop. Even if he just disappears one day, oh well. That's that. And that's all that has to be. It's not like you or I have never had a device or resource that stopped being usable before. What happens? You move to a replacement. It's nothing special, even open source is not permanent.

You mention a strength of open source software:
>They are improved upon by hundreds of different people.
Most open source projects max out at ~15 core devs, and I really do absolutely mean max out. The Linux kernel and a few GNU projects are the only real exceptions, maybe counting ffmpeg. Most other projects hover around 5-10. mpv is no exception. It might have a few small bits contributed by people who only very seldom get involved here and there, but I don't think that meaningfully constitutes being worked on by "hundreds of people." I'm not anti free software / open source / whatever, but the statement you just made is a common fallacy. It simply isn't supported by any real data.

And yet, even then, they haven't caught up to what only one person has done when left to work on things as they please. mpv always seems to be playing catch up on the renderer side.

Let's just treat them as their own things.

>>51633910
I had the same issue, I filed a bug on github but wm4 closed it. Could you post the details on there? He might pay more attention to it if more people are having problems.
https://github.com/mpv-player/mpv/issues/2551

>>51633905
That's because its decoders are flawed and had been for years upon years. It's not an active project by any means. Defaults settings often produced improper colors. There was no dithering. Resampling algorithms were limited and often achieved poor results, etc.

This was the case up until vlc 3.0. When did that happen, earlier this year? It took that long to have basic features? I always hate to see large websites suggest vlc. There are far better alternatives, and they're even easier to use. Why continue to suggest something that's obviously wanting. If it's not doin' nothin', abandon the goddamn thing and move on, few things stay true forever. Hell, for all we know Dirac was right and not even G is a constant.

>>51634046
>No. I'm comparing the rendering components of a full fledged player to another renderer.
Not obvious from your text " Everything mpv has"

>hasn't expressed any inclination to stop
For now

>Most open source projects max out at ~15 core devs
They come and go. I am talking about all the different people ever contributed not the "regulars".

>. It simply isn't supported by any real data.
Check the commit logs of ffmpeg and mplayer/mpv from the very beginning

>mpv always seems to be playing catch up on the renderer side.
"always"? No mplayer/mpv developer gave a shit on rendering side before haasn. He did his thing and that's it. madshi seems to be offended to wm4 but he has ridiculed some of haasn work placebo.

>>51634046
>mpv always seems to be playing catch up on the renderer side.
Not him, but mpv caught up once they implemented jinc and there's plenty of advantages mpv has over madVR (more configuration and finetuning, etc.)

>>51634052
Ok I commented.

>>51634220
>Not obvious from your text " Everything mpv has"
I figured it was implicit by the context of this thread. Imagine if we were all telepathic.

>For now
Nothing is permanent. And it doesn't even matter. Maybe he open sources it when he stops, maybe not. Utilities to optimize deflate stream block headers have had developers drop out of existence (seemingly), website goes down, etc. And within a year someone has either reverse engineered or rewrote their work, or there was simply a better alternative waiting to surface.

Man, that's just how it works. If you get mired down in your perception of the ideal case, you're often led to lose sight of the whole. I am very prone to this myself.

>They come and go. I am talking about all the different people ever contributed not the "regulars".
The regulars are the people do all the real work of any project. Most of the important parts of the programs framework are maintained and expanded by them. Like I said, it doesn't really matter if hundreds of people CAN contribute, because they rarely do so substantially. It's mostly bug fixes that happen that way.

>Check the commit logs of ffmpeg and mplayer/mpv from the very beginning
But 3 projects among many. I already said they're the exceptions.

>"always"? No mplayer/mpv developer gave a shit on rendering side before haasn.
Yeah, I'll give ya that. Can't really be considered catch up if no one actively gives a damn about getting anywhere. I think you know what I meant.

>>51634394
>I figured it was implicit by the context of this thread.
People sometimes also mean MPC-HC + madvr when they just say madvr in these threads.

>The regulars are the people do all the real work of any project
Many times the little bug fixes can make a day and night difference wrt performance and usability, you can't disregard them. It's the difference between the project works meh and it works great, it can be difference the project works in one os version and not in other.

>But 3 projects among many. I already said they're the exceptions.
madshi is also an exception on the freeware longetivity side. But seriously video rendering stack is dominated by open source. madvr won't have much in the long run.

>I think you know what I meant.
Good thing that in open source you are not at the will of a single developer.

>>51633725
Yeah, I mean bad source material is something I try to avoid, and choppiness on pans makes my hair stand on end. At a certain point though, I just say, "that's good enough" and go on with my life.

It's the same way with everything though whether you're talking about speakers, engine / suspension performance, furniture / room painting, smoking meats (or just cooking in general), etc. There are always things you can do differently or other products that you could use or even minor differences in technique that will incrementally improve just about anything but at some point diminishing returns kicks in and you begin getting less and less back from the effort/time/money that you put into it. I've been bitten by the perfectionist bug before -- tweaking and tweaking dirtbike engines and suspension to get it just perfect when I'm not even a professional motocross racer, re-texturing and repainting rooms multiple times until the paintjob looks near perfect from a technical standpoint (no overrun anywhere 2 different colors/surfaces meet, wall/ceiling texture is as uniform as possible) when the only thing most people notice is what color you put on the walls, I've lapped CPUs and heatsinks to a mirror finish and tried out multiple different thermal compounds to only achieve a difference of ~3C-4C under 100% load, I've lapped rifle bores and glass bedded barrels and scopes and honed trigger groups and painstakingly worked up optimal loads in ammunition just to achieve a .5"-.25" tighter grouping when I'm not even a competition shooter, on several occasions I've trashed entire projects and started over from scratch because of one minor cosmetic blemish or error...

I've spent 100's of hours and 1000's of dollars in search of perfection in different things over the years and have just become disillusioned with it because it just really isn't worth it in the end most of the time.

>>51634220
>advantages mpv has over madVR (more configuration and finetuning, etc.)
I mentioned this above. The issue is, they're only theoretical advantages. Have you seen any comparisons where this finetuning was able to produce a superior result? Or is it only on the basis of the capacity to configure something if you so desire (which is something I'm certainly fond of and indicated above).

The only advantage it has as far as ewa lanczos goes is the option of using sigmoidization and specifying an arbitrary radius. madVR doesn't use or even offer sigmoidized "jinc", as madshi apparently thought it produced poorer results more often than not. As well as manually specifying taps.

I don't really care much about NNEDI3 because I think it looks like plastic garbage. It's overly expensive and non-viable for anything other than field interpolation, as far as I'm concerned. Its proficiency in connecting diagonals and avoiding aliasing is outweighed by how artificial its output looks even at 256 neurons.

mpv is fine af.

>>51628121
windows babbys cannot comprehend config files and command line

>>51634616
>Have you seen any comparisons where this finetuning was able to produce a superior result?
Sure with and without icc-profile. I don't really know how madVR handles color management these days, but I remember it being a huge pain in the butt.

>ewa lanczos
There's different presets though (soft or sharp).

>>51628121
>mpv
>Total setup time: 5 minutes

Riiight.... That's why we have these constant config topics, cause thousands of people are going through the 5 minute journey to video nirvana.

And that's why we never see MadVR config topics, all those people have starved to death clicking through the ui.

>>51630356
yep, thanks from me, too.

>>51634751
i didn't even configure mpv desu. too lazy and never thought it was necessary.

>>51634616
>NNEDI3 because I think it looks like plastic garbage
nnedi3 is mostly a meme. Slow as balls for upscaling to anything over 1080p and everytime opencl drivers borked doom9 threads was filled with cries.

The real use for nnedi3 is its deinterlacer, encoders use it frequently with avisynth

>>51634559
>Many times the little bug fixes can make a day and night difference wrt performance and usability, you can't disregard them.
It's certainly cumulative. Sometimes people do come along with far more efficient ways of doing a given operation as well. I understand the value of distributed cognitive labor.

>madvr won't have much in the long run.
That's unlikely, given that madvr has been the main innovator on the rendering side. This can't really be argued, I mean colorspace conversion didn't even have dithering in video before madvr came along.

>Good thing that in open source you are not at the will of a single developer.
I'm never at the will of a single developer. If something isn't the way I want it to be, I work on an alternative until I've created something better. I don't care if something is open source or not. People making solid software that I don't see as at all wanting is what matter for my use case. It's all that saves me (and my time) from myself.

>>51628121
Don't use mpv:
Massive RAM usage.
Waste of CPU time.
Broke everything every minor release.
Will crash from every single sneeze.
Developer incompetent jackass.
OS unportable.

>>51634751
That's because most people using MPC-HC and the whole shebang just download KCP or CCCP. That's the equivalent of just using someone else's config file in mpv.

>>51634814
Yeah. That's all I ever used it for. Field interpolation for deinterlacing, or for recovering orphaned fields when ivtc'ing something. Never liked it for upsampling.

I still use vlc. Heres why:
It works right away.
The defaults are sane.
It has a lot of nice features.

mpv:
Works right away.
UI is ugly as fuck and the default hotkeys are really bad.
has only a subset of the features so I have to use multiple applications (so much for "single exe")

>>51630154
>Delphi
Is this for real? HAHA

>>51634881
Yes.

Why I use mpv:

It just werks
GPU Acceleration
Built-in youtube-dl

I haven't touched the config(maybe for youtube-dl).

>>51634872
>I still use vlc
How you deal with farted sound?
And wrong color?

>>51634872
>UI is ugly as fuck
What's wrong with it?

I've been using MPC-HC + madvr for years now, it's really not that hard to figure out.
No need for extra subtitle filters unless you're into anime.

Out of interest, I played around with mpv and after a couple of days, I decided to stay with mainly for 2 reasons:
- mpv does not support hardware decoding for 10 bit HEVC, which LAV does (provided you got the hardware, of course, e.g. a gtx 960)
-more importantly, I couldn't get rid of slight horizontal tearing in mpv for the same level of PQ that I get with MPC-HC.

>>51634927
RISC OS is ok?

I love mpv, even use it on my Windows desktop with an tricked out to shit config. But I think the effects of certain settings may have changed with updates or something, but suddenly my mpv install will either stutter during video playback or start to drop frames. Aside from that, I'm still trying to figure out how to get 5.1 audio to downmix properly to 2.0. I mean there's no sound missing, but sometimes it's just absurdly quiet compared to an audio track that was 2.0 to begin with. I'll post my config, my system is an i5-4430, GTX 960, and 8 GB of RAM.

# Video settings
vo=opengl:scale=ewa_lanczossharp:dscale=mitchell:cscale=ewa_lanczos:interpolation:swapinterval=1:tscale=oversample:scale-antiring=0.5:scaler-resizes-only:sigmoid-upscaling:fancy-downscaling:pbo:dither-depth=auto:temporal-dither:waitvsync:dwmflush=windowed:blend-subtitles=yes:target-prim=bt.709:fbo-format=rgba32f:deband
framedrop=no
hwdec=no
video-sync=display-resample

# Audio settings
ao=wasapi
softvol=yes
softvol-max=200
audio-pitch-correction=yes
ad-lavc-downmix=yes
audio-channels=2
alang=en, eng, jp, jpn

# Subtitle settings
ass-force-style=Kerning=yes
sub-auto=fuzzy
demuxer-mkv-subtitle-preroll
slang=en, eng

# OSC settings
script-opts=osc-layout=box,osc-seekbarstyle=bar

# OSD settings

# Screenshot settings
screenshot-format=png
screenshot-png-compression=6

# Window settings
keep-open=yes
autofit-larger=90%x90%

# Miscellaneous settings
save-position-on-quit
cache=auto

>>51634938
there is a lot better sound control on vlc than mpv.
I don't know about color I don't watch a lot of movies or cartoons with it so it never bothered me.
>>51634959
The buttons are huge and you don't get as many options as you get with vlc.

>>51635112
Learn keybinds

>try mpv
>doesn't even show proper subtitles in mov files

And this is why I still use VLC.
Hipster video players of the week get their share of autist hype, but at the end of the day they still can't beat VLC.

>>51629723
>install mpv
>idk but I never installed it and don't want to because with lacks nnedi3
welcome to this year, anon.

>>51628121
>MPC-HC full install + LAV Filters + madVR + xyVS/xySub-filter + ReClock

LAV has been embedded in MPC-HC for awhile now, the subfilter and ReClock are optional and rarely actually needed.

>All that shit has to start up every time, even if you even just want to watch a 10 second webm

Wtf? Watching webm on a media player? Do you also use DSD to listen to inception horns?

>You must configure each with a GUI using check boxes, menu trees and sliders

Oh no, check boxes, menu trees and sliders, oh my! The ui foundations of every OS used by civilization!

>Total setup time: All day

You mastubate too much.

>Now madshi's proprietary madVR was the shit back in the day, but it's over. Open source wins again.

"It's old dilapidated shit"

>If he wants madVR to stay relevant, he should probably just release his algorithms and further the glory of mpv.

"... That would improve my superior software?"

topkek

>>51635112
>there is a lot better sound control on vlc than mpv.
???
Wat? I talk about serious sound distortions (1000+Kb/s 48Khz sound like 64Kb/s 5-10Khz).
>I don't know about color I don't watch a lot of movies or cartoons with it so it never bothered me.
.... and do you not want know how original look like?
Reason to use player which can't just play file without serious distortions?

>>51628121
>bind alt-m to open the prmary selection with mpv
>shit's supposed to open fast, wth
>bind win-m to open with vlc
>nearly 20% faster and I get proper options like deciding whether I want the video to loop or not

>>51631927
>Using different frames every time
What the fuck is the point?

>>51629690
Not him but OpenGL support is trash on Windows, it's why they're implementing ANGLE support.