[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k] [s4s] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/a/


File: logo-256x256.png (31.59 KB, 256x256)
31.59 KB
31.59 KB PNG
Is MPC's internal subtitle renderer not as shitty as it used to be nowadays?

I was getting a ton of dropped frames on certain files (like the very beginning of KLK 14 for example). I thought it was madVR at first, but turns out xySubFilter was killing me.

I know MPC's has had a ton of work done on it recently, but is it up to snuff with xy-VSFilter or xySubFilter?
>>
>>100723426
You can make xysubfilter faster by reducing subpixel position and rendering to original video size.
>>
ISR still has positioning problems. xysubfilter is the best available right now
>>
Nope, still slower than xysubfilter / xy-vsfilter.
>>
Apparently some work has been done to the ISR, at least in CCCP's build of MPC-HC. I doubt it's faster than xy-VSFilter or even XySubFilter, but I haven't done any testing so I can't really say. If XySubFilter isn't fast enough for you then try the good old xy-VSFilter.
>>
>>100723426

MPC-HC's VSFilter & ISR are no longer dead-slow, but on my Intel i5, xy-VSFilter & XySubFilter are still 30-100% faster, depending on the script.
>>
>>100723426
MPC-HC devs is improving it because they don't like the xy* projects, so if you use the latest nightlies or CCCP betas, they should be able to handle most typesetting decently. xy* will still be better, though.
>>
Oh look, its the A/V club.
>>
Why don't mkv files work with my MPC anymore?

I hear the audio, but the video is all black. Other video types work fine.
>>
>>100728288
Because you're doing it wrong.

Install latest CCCP, use the included MPC-HC.
>>
File: CPUkawaiso.png (22.29 KB, 767x472)
22.29 KB
22.29 KB PNG
>>100728288
If you're using madVR, probably you fucked up the settings. just reset it and you'll be fine.

anyway, if you're using commie-subs, did you maxed-up your CPU and GPU queue size just to play the episode two of Nisekoi?

pic related.
>>
Is there any modern guide for MPC-HC? BishojoProject is gone now.
>>
https://github.com/mpc-hc/mpc-hc/pull/126

80% faster now

Of course cached solutions (xySubfilter) are faster for karaoke effects and such
>>
>>100728761
http://underwater.nyaatorrents.org/?nav=playback
>>
File: fuckingcommie.png (836.13 KB, 1288x741)
836.13 KB
836.13 KB PNG
>that frame drop
>that delayed frames
>best guess?

my fucking toaster is on fire while i'm watching this release. fucking 10bit
>>
CCCP has XySubFilter now?
>>
>>100728991
No.
>>
>>100728991
they're using xy-VSFilter because it's stable than XySubFilter
>>
>>100729162
it seems xySubfilter development will be back next month
>>
How is KCP in comparison to CCCP? I know they set everything up for you right out of the package but is that a good thing or a bad thing?
>>
>>100729278
We'll see about that.

The best thing to do would probably be writing a decent font manager for libass, now that there's this guy speeding it up. It already "supports" most of the quirks of VSFilter, so older releases should look fine enough.

Or just make a new format altogether.
>>
>>100729463
CCCP also configures everything for you, but does not include madVR and xySubFilter.

KCP will probably work for most people, but CCCP does more checks and stuff, and is more likely to both work and not fuck up your system otherwise.
>>
A new version of madVR will be released before the weekend. New features :
> debanding
> NNEDI3 upscale
> upscaling profiles (different for MHD/HD 24/30/50/60/... fps)
>>
File: 01.jpg (318.40 KB, 900x1250)
318.40 KB
318.40 KB JPG
>>100729601
Strictly in terms of picture quality, KCP would be the better choice then?
>>
>>100729914
Yeah, as long as it uses madVR.
>>
>>100729868
what does this mean?
>>
>>100729914
as long Jinc(8 taps, AR/LL) & Catmull-Rom(AR/LL)
>>
File: madVrTeaser.png (25.34 KB, 776x487)
25.34 KB
25.34 KB PNG
>>100729868
Already using those test builds with debanding. Works pretty good, but of course not even that is able to save the worst of CR/Funi encodes.

NNEDI3 will be great as long as it's not to demanding for me. Profiles also seems promising.

All in all, I'm hyped.

>>100730203
madVR should look better anyway in terms of 10-bit dithering.
>>
>>100723615
>rendering to original video size.
What's the point of using it instead of xy-vsfilter then?
>>
>>100728827
I'm using xy-subfilter. The whole episode was fine for me, except for this part at the start. It works just fine at 720p, but when I try it at 1080 or 1440, the video just stops completely.
>>
>>100730292
Worked for me with xySubFilter/madVR after I increased the CPU queue size to 64 in madVR settings.
>>
>>100730409
Oh, I didn't know about that. It works when I have it set to 128, but the memory usage shoots up to 1.5GB.
>>
>>100730566
Heh, yeah. 64 worked nicely for me in fullscreen, so I'll just keep it at that.

Typesetting like that isn't really common anyway.
>>
>>100728827

>commie

ew man. ew.
>>
I wonder if that new madVR profiling will be able to take file names in to account. If so, since you can apparently profile debanding strength, you can do something like
if (nameContains("HorribleSubs") "Debanding strong"

Because I don't see how else you'll be able to distinguish HS stream rips with already filtered fansubs.
>>
>>100731225
Well, CR rips can be identified from the framerate since it's always either 23.810 or 30.303 instead of 23.976.
>>
>>100731462
Ah yeah, that's true. But Funi rips are still 23.976.

Why is CR rips tagged as 23.810, anyway? The video is still actually 23.976.
>>
>>100731615
>>100731462
The dumper that HS isn't very accurate, for some reason.
>>
>>100731716
Right. So they use something different for Funi rips, I guess. I don't see why it would be hard to fix, but I guess it's not really a big deal.
>>
>>100731716
Well, to be fair, I think they posted that it's essentially just a Perl script running 24/7.
>>
>>100731955
That's their leecher bot (automatically download the video and the encrypted *.ass in specific time and mkvmerge)
>>
EVR-CP + xySubfilter = fastest subtitle rendering
>>
>>100732265
Is XySubFilter really faster than xy-VSFilter?
>>
>>100732378
it's better optimized as they started anew
>>
>>100732458
Not really, it's still VSFilter based.
>>
This seem like the appropriate place.

Anyone know, how i can get more vivid colors with madvr?

Pic is comparison of my default renderer and madvr in mpc.
>>
>>100732538
code is better organized. That's why they will only update xySub from now onfilter
>>
>>100732265
>>100732378
xy-VSFilter should probably be faster, at least if you take scaling into account. I haven't really tested though, but xy-VSFilter handles Commie's Nisekoi 02 much better than xySubFilter at 1080p.
>>
>>100732569
Your GPU driver is probably interfering with picture in EVR screenshot, oversaturating it.
>>
>>100732649
xySubFilter uses much more memory, and because of VSFilters bad memory management, that can lead to problems. If you use a build from the xySubFilters git head, it will even crash a lot of the time instead of simply stutter.
>>
>>100732649
whats' your screen resolution?
>>
>>100732756
1080p. Video scaled from 720p. xy-VSFilter renders at 720p, xySubFilter at 1080p.
>>
>>100732756
I'd dare to guess it's 1920x1080, since he's watching 720p video at 1080p.
>>
>>100732832
>>100732847
then it's normal. xyvsfilter will be faster that xySub on that scenario
>>
>>100729601
>but CCCP does more checks and stuff
I was using CCCP for awhile, kept getting frame rate problems, so I switched to my own set-up, same problem.

Tried stable KCP and not sure if it was re-clock but the frames were not sticking and it worked well.

I do recommend KCP, even though Daiz has said time and time again that it's shit.
Though I do agree that people shouldn't use beta, since we don't need bleeding edge as VS is fine for now and Sub is not ready completely. And little changes to filters or the base MPC is a waste of time.
>>
What's next after 10bit?

16bit?
>>
>>100733084
10bit h265 will be the next standard.
>>
>>100732569
I'm using standard CCCP with madvr added.
>>
>>100733139
>not x265
>>
Does branding of GPU effect anything involving playback.

Like Nvidia works with some adobe products better, and in some cases you need Nvidia for 3D modeling within Photoshop Extended.
>>
>>100732569
What >>100732711 said. If you really want to fuck up your video with madVR as well then install FFDShow and use the raw video filter.
>>
>>100733084
10-bit is probably going to be used for a while. The HEVC even has a 10-bit main profile.

>>100733169
H265 is the format/standard, x265 is an implementation of that standard, i.e. an encoder that outputs h.265 video.
>>
>>100733169
h.264 is the codec, x264 is an encoder for it.
So it would be the same for h.265.
>>
>>100733293
>>100733338
Speaking of which

http://www.nyaa.se/?page=search&cats=0_0&filter=0&term=x265
http://www.nyaa.se/?page=view&tid=514712
>>
>>100733402
It's still retarded to use x265 for anything other than testing as 10-bit x264 still produces better results at relevant bitrates.

Also, laughed at http://www.nyaa.se/?page=view&tid=514712
>>
>H.265 (x265). Bitrate: 284 Kbps. Size: 67.2MB
http://ifme.sourceforge.net/page/future.html#264vs265

I don't know if I can believe this
>>
>>100733715
It's a re-encode of the HS release.
>>
>>100733715
There's nothing impressive about low bitrates when the quality obviously suffers from it.
>>
>>100729868
>>100730264

realtime nnedi3 will probably require over 9000 sli'd/xfire'd GPUs
>>
>>100730203
>as long Jinc(8 taps, AR/LL) & Catmull-Rom(AR/LL)

- using more than 3 taps for Jinc is nonsense
- using LL when upscaling is nonsense
- using AR when downscaling with CatRom is nonsense
>>
File: Untitled-3.jpg (763.89 KB, 1440x2160)
763.89 KB
763.89 KB JPG
>>100733144
EVR and madvr have identical colors for me. I only notice the difference when upscaling to 1080p, and then it's just some extra artifacts with EVR.
>>
what settings should I be using for xysubfilter?
>>
>>100731615
>Why is CR rips tagged as 23.810, anyway?

some bullshit rounding error in whatever HS is doing (42 ms frame duration instead of 41.7 ms)
>>
>>100734068
Default.
>>
>>100733926
Yeah, I'm curious of how he's handled that.

>>100734068
You shouldn't really change anything in xysubfilter settings.
>>
>>100732569
>Anyone know, how i can get more vivid colors with madvr?

Your EVR cap looks fucking terrible, but if you actually like that, you might want to look into using pixel shaders like "vibrance"
>>
>>100734050
>EVR and madvr have identical colors for me.

Chances are >>100732569 is running an AMD GPU and getting his EVR output ruined by "enhancements"
>>
>>100728575
>that queue size
that'll munch RAM/gfxRAM like no tomorrow
>>
>>100731225
>>100731615
Doesn't look like you'll be able to use filenames for that.
http://forum.doom9.org/showthread.php?p=1271417#post1271417
Perhaps you could e-mail him about it?
>>
>>100734183
If he likes over-saturation, he can also just use ctrl+alt+shift+i in madVR to double expand TV levels.
>>
File: HLv1qmj.png (3.12 KB, 534x196)
3.12 KB
3.12 KB PNG
>>100734042
Not him, but I followed this chart. How is it?
>>
>>100734481
Jinc 3taps AR is pretty standard if you can handle it. I have my downscaling set as the "Highest" row in your pic suggests, but I've yet to see a reason why I should give a shit about downscaling if you've got a 1080p display.
>>
>>100734312

Yep AMD, I'll better check the catalyst settings, hope the drivers won't cause a bsod.

>>100734477

Will see what it does.
>>
>>100734445
Either that, or just post in the thread.

Just don't mention HorribleSubs or anything or neuron2 will ban you!

>>100734481
It's good. But Jinc for chroma is really overkill, so that should be the first thing to change if your GPU can't handle it.

Also, it can be somewhat subjective. Some prefer spline over lanczos, for example.
>>
Is there a good way to get rid of the banding?
>>
>>100734752
Use testing build of madVR, or wait a few days for stable release.
>>
>>100734752
Besides downloading better videos next time, I haven't found a good way. ffdshow can do it, so for videos where it's an issue I can check it in my external filters.

Install ffdshow first before CCCP and it'll show up as an option.
>>
>>100734690
2k and 4k content
if you don't go fullscreen for 1080p content
>>
MPC's internal sub renderer still fucks up the OP of Monogatari with Kaiki.
>>
>>100735034
>4k anime
We're not living in the future yet.

>2k
You keep using that term. 1080p is 2k. Our marketing overlords are flipping the axis on us and doing that shitty rounding-up thing they always do. Widescreen "4k" is 3840x2160, or 2160p.

There are other random formats that count as 2k and 4k, but smart encoders won't deal with that shit.
>>
File: Vector_Video_Standards2.png (127.94 KB, 749x599)
127.94 KB
127.94 KB PNG
>>100735034
>2k and 4k content
Do you know how long it will take to content like that?
I don't even think dual-layer Blu-ray discs can handle 4k.

>>100735389
2k and 1080p are different. Different aspect ratios though.
>>
>>100734445
Didn't even notice he had posted that. But yeah, doesn't say anything about file names.


if (srcBits == 8) "Strong debanding profile"

Should work for detecting HS, but of course will also get most western-scene film content as almost everyone use 8-bit. The (srcFps == 23.810) hack will catch that, but Funi content will still not be detected.

File names should be useful outside anime content too, for example having different profiles for names containing "BDrip" and "WEBRip", so there should be enough use cases for madshi to consider adding it. madVR should be able to get the file name at least, iirc it already uses it for some other things.
>>
>>100735389
It's called "4k" because you can fit 4 1920x1080 images in one 3840x2160 image.
It's still dumb as fuck, though, and is only misleading since there's an actual 4k standard that has nothing to do with consumers.
It'd be much easier and more consistent if we just stuck with vertical resolution to name them.
>>
>>100735702
>The (srcFps == 23.810) hack will catch that, but Funi content will still not be detected.
The trick is to just not watch funi rips. On top of shit video, the subs are horrible too. Have you seen Nourin?

http://www.crymore.net/2014/01/21/funimations-no-rin-02-is-the-worst-simulcast-script-ive-seen-this-season/
>>
File: mantis shrimp.jpg (164.27 KB, 1024x768)
164.27 KB
164.27 KB JPG
>next madVR version
>error diffusion dithering
>OpenCL NNEDI3 prescaler

>mfw
>>
>>100735852
Yes, you can fit four 1920x1080 images in 3840x2160 but the name 4K comes from the fact that its horizontal resolution is approximately 4000 pixels. Thus the "K".
>>
>>100735694
>I don't even think dual-layer Blu-ray discs can handle 4k.
the physical disks can handle it just fine, the standards currently don't allow 4k content. the consortium is currently considering to change that as many newer players could already handle it anyway.
>>
>>100735876
Sure, I rarely watch Funi, but it'd still be useful to catch them. And I wouldn't say it's ideal to rely on HS's 23.810 for CR rips.

Of course, one can still just use ctrl+alt+shift+1 to toggle between the debanding modes, so it's not like it's a hige deal. Debanding is the only reason to have different modes for webrips and normal fansubs, as far as I can see.
>>
>>100735852
>It's called "4k" because you can fit 4 1920x1080 images in one
No, that would be 4x. When you say "4k" it means 4000, as in...

>>100736077
Yeah, what he said.
>>
>>100736077
Strange, I swear the reason I posted is why it was called 4k.
Either way though, it's still pretty dumb.
>>
>this thread
>plebs discussing 10 year old programs
>extended up their ass with filters and codecs
mpv is the future, step it up senpai
>libass for full res sub rendering
>opengl-hq for muh colours
is all you need
>>
>>100736199
>the standards currently don't allow 4k content
Standards of the containers or the Blu-Ray Association not updating the standards of their discs?

>newer player could already handle it anyway
So we won't have to wait for the price of quad-layer discs to lower and them to release quad-layered players?

I honestly thought it had to do with the size but with Hi10p and the h.265 the sizes has shrank enough to keep the same amount of episodes to a disc and movies too I guess.
>>
>>100736449
While mpv is great, madVR is still a superior renderer than anything mpv uses. And it will just get better after the next update.

And you shouldn't use libass as a argument for mpv as long as fansubbers target VSfilter.
>>
>>100736628
While fansubs do target VSFilter, libass just about has full compatibility with it nowadays.
>>
File: end25.jpg (350.85 KB, 1145x1600)
350.85 KB
350.85 KB JPG
>>100723426
Use MPC player they said.
MPC is better than VLC they said.
MPC with CCCP will improve your animoo experience 10 fold they said.

It can't even play my Godannar DVD right. Thanks, slovak cheeki breeki slavi lavi.
>>
>>100736628
libass is pretty much 100% compatible with vsfilter
the renderer mpv uses is basically madvr with sane defaults
you can even configure it in the same way as madvr, set the scaling algorithm, the parameters, etc, etc.
there is no reason to use old shit like mpc-hc, which let me remind you is based off of Windows 2000's WMP
>>
>>100736499
I highly doubt we'll see improvements in physical media for home video for some time unless they manage to be fully backward compatible with underpowered, first generation BD players.

The reality is we'll get higher resolution once the Internet is fast enough for Netflix and the like to push it.
>>
>>100736827
>you can even configure it in the same way as madvr, set the scaling algorithm, the parameters, etc, etc.
It doesn't even offer Lanczos4 + AR, much less Jinc3 + AR or (soon) nnedi3. It also doesn't have equivalent of smooth motion.
>>
>sub flashing the entire time

This shit drives me up a wall and it only happens during OPs and EDs.
>>
>>100736721
>MPC is better than VLC they said.
If you think for a second that VLC is better, you're a fucking idiot.

It's a lot better than it was 3 years ago, but still lags behind.

It's the most compataible, user-friendly media player that's great at audio streams, cross-platorming from desktop to mobile and has that sweet little option for a remote so you don't even have to leave bed to pause it, skip some boring scenes or god-awful OP/ED when the playlist is not enough.
But people value quality over that.

Also for some reason, I can't load a video without it crashing all the time, so I personally don't use it.
>>
>>100736965
Stop using ISR then.
>>
>>100736965
Stop using internal subtitle renderer.
>>
>>100736708
Yeah, it's much better than what it used to be. But VSFilter still has a lot of quirks which fansubbers might rely on. Recently you had that newline thing in UW's KlK (which was more of an libass issue than VSFilter, actually). Of course it was a retarded use of newlines, but since, again, fansubbers target VSfilter things like that might happen.

>>100736827
>old shit like mpc-hc
>based off of Windows 2000's WMP
Don't be retarded, man. The GUI is based around old WMP, but that's about it.

>>100736965
Just use the bundled player from CCCP.
>>
>>100736970
>and has that sweet little option for a remote
So does MPC-HC.
>>
Why do you guys praise smooth motion so much? All it does is slow down playback.
>>
>>100737177
Unlike you, we don't watch anime on toasters.
>>
>>100737177
Because slowing down playback isn't all it does.
>>
>>100737177
You shouldn't use it if you GPU can't handle it.
>>
>>100736914
>once the Internet is fast enough for Netflix
Didn't some news break a few months ago about Netflix downscaling all 1080p content?

No doubt one day, we'll be able to purchase/download 4k content online, we won't see broadcasting/streaming 4k for quite awhile though.

Fiber internet will be how we get to downloading the filesizes. I don't believe Netflix will be much of a helper to push the 4k standard.
>>
>>100737073
Yeah, the newline is the biggest problem remaining with libass, but as Daiz himself said, the issue is fairly avoidable with pos.
>>
>>100737177
You're thinking of reclock, and if you're using reclock correctly, it doesn't do that unless you're speeding down a 25fps video back to 23.976fps.
>>
>>100737102
Wait seriously?
I'm googling that shit now.
>>
>>100736827
>there is no reason to use old shit like mpc-hc, which let me remind you is based off of Windows 2000's WMP
MPC-HC is a fork of MPC which is based largely off of ffmpeg(specifically ffplay) which mplayer, mplayer2, and mpv are also based on. The only thing MPC/MPC-HC have in common with WMP is the interface, it was intentionally modeled after it's interface because it was simple
>>
>>100736947
>It doesn’t even offer
>Lanczos4, Jinc3, nnedi3
>even

Shit nigga, have you ever compared those algorithms? I bet you can’t even see the difference between lanczos2 and blackman4. Spline36 is more than enough placebo, anything beyond that is insanity.
>>
>>100734481
>Not him, but I followed this chart. How is it?

Not too terribly wrong, I guess. Some choices are kinda dubious.

AR does just about fucking nothing when downscaling with Catmull-Rom, so it might be wise to spend that performance on better chroma upscaling instead.

Also in my opinion there are better tradeoffs between sharpness on the one hand and the least amount of aliasing/ringing on the other hand in some cases, for example Spline without AR over Lanczos without AR, and Lanczos with 4 taps plus AR over Lanczos with 3 taps plus AR.
>>
>>100738843
>Shit nigga, have you ever compared those algorithms? I bet you can’t even see the difference between lanczos2 and blackman4. Spline36 is more than enough placebo, anything beyond that is insanity.

http://screenshotcomparison.com/comparison.php?id=4121

Your move.
>>
>>100739413
The first one looks slightly better. The second image appears to have artifacts in and around most of the hands.
>>
>>100737177
>Why do you guys praise smooth motion so much?

It pretty much eliminates judder when watching video on a display whose refresh rate is not an integer multiple of the content's frame rate.

Like just about every flatscreen computer display, many TVs, or any and all VFR sources.
>>
I had problems with MadVR, kept on crashing on certain scenes, so I went back to using EVR, at the end of the day I'm liking it more because of the darker colours.
>>
What's the deal with the improved EVR? I remember a month or so ago there was a commotion because it started to support xy-vsfilter, but it wasn't quite up to par. Is it still inferior or has it continued making progress?

I can easily run madVR but it's still nice to let the machine rest if given the opportunity.
>>
>>100739413
Now do the same with mpv -vo=opengl-hq:lscale=spline36

Your second screenshot is bilinear upscaling, that one is really really REALLY easy to spot. Oh, and try to get the same frame.
>>
>>100738843
difference 4tap spline+AR vs. 4tap jinc+AR
>>
File: fullscreen.png (680.44 KB, 1920x1080)
680.44 KB
680.44 KB PNG
How do you even take screenshots of fullscreened mpv?
>>
>>100739703
you can just use madvr and use one of the cheaper upscaling algorithms if you don't want to use much power.
>>
>>100739626
There shouldn't be any difference in the colors.
>>
>>100739982
>Filtering
Yeah, sure. You can do filtering in avisynth too. I was talking about the algorithms themselves. But oh, they probably look worse on their own because they ring too much, right?
>>
File: a at her computer.jpg (96.23 KB, 800x600)
96.23 KB
96.23 KB JPG
>>100740081
They are bit darker, probably I touched something.
I'm a retard when it comes to computers so for me as long as it just werks I won't complain.
>>
File: 3m3G1.jpg (76.73 KB, 626x481)
76.73 KB
76.73 KB JPG
>>100739744
post you're mpv config please
>>
>>100740232
You probably have some video "enhancements" turned on in your GPU drivers that fuck up your video with EVR.
>>
>>100740232
Well, if there is any difference, it's usually madVR that does it right. Might be your graphic card interfering with the colors when using EVR-cp.
>>
>>100740292
>>100740314
I'll start using it again and save the crash report then.
>>
>>100739744
>Your second screenshot is bilinear upscaling

I don't think you've ever seen bilinear upscaling if you think that frame actually suffers from it.

>Oh, and try to get the same frame.

It is the same frame, subpixel offset is due to resize filter implementation differences.
>>
>2014
>People still having issues with playback of their animu.

Plebs. My fucking Surface Pro handles everything I throw at it with KCP and madvr set to second highest settings with LAV.
>>
>>100740450
Or you could just disable those "enhancements" if you want to keep using EVR.
>>
File: spline36.png (1 MB, 1920x1080)
1 MB
1 MB PNG
Also I wasn’t saying that there is no difference, just that splerging about that minor stuff as essential, saying “doesn’t even”, is incredibly stupid. A real world comparison between lanczos2 and spline36, with a very noticeable difference:
>>
File: lanczos2.png (1.02 MB, 1920x1080)
1.02 MB
1.02 MB PNG
This is actually a “pretty damn big” difference.

>>100740485
I just read the top and didn’t see any specification of any algorithm used, so I just thought that was pretty retarded.
>>
>>100733402
>>100733655
>>100733715
>>H.265 (x265). Bitrate: 284 Kbps. Size: 67.2MB
>http://screenshotcomparison.com/comparison/57559

Well, it sure looks the part. Can I please get some bleach for my eyes now?
>>
>>100740546
>>100740714
They look the same. Calm your autism down.
>>
File: azumanga delta.png (233.64 KB, 1920x1080)
233.64 KB
233.64 KB PNG
>>100740714
>>100740546
>>
>>100740546
>>100740714
>This is actually a "pretty damn big" difference

hahaha holy shit.
>>
>>100740258
Nothing special, because I don’t care about placebo, just the usual demuxer-mkv-subtitle-preroll, keep-open, force-window etc.

>>100740822
And what was my point again?
>>
>>100740793
looks pretty good for how low the bitrate is
>>
>>100740546
>>100740714

If you were to try that on a better source with cleaner edges that allow scalers to play to their strengths instead of drowning their impact in compression noise, you'd probably see more of a difference.
>>
File: deal with it fabulous.gif (1.90 MB, 400x300)
1.90 MB
1.90 MB GIF
>tfw have never had these problems /a/ is always talking about with MPC and subtitle flickering

or any other problem really

feels good man
>>
>>100740960
That doesn't really mean anything. Do the same with x264 and get similar results.
>>
>>100741069
>implying
>>
I installed cccp last night after about 8 months of vanilla MPCHC, is there a way to turn off subs using a keybind? I can't seem to get it to work.
>>
>>100741069

I actually doubt that. x264 is not quite as well-optimized for compression this strong.
>>
>>100741029
Well, I was posting a real world example on purpose, and even on the best source possible it’s really hard to spot a difference. Spot one on a not moving screenshot with all the time in the world to look for differences. Point is, better scaling beyond lanczos does tend to get really fucking useless and noticeable differences are due to other stuff with questionable usefulness, like filtering the video.
>>
>>100741125
S
>>
>>100741333
>Those grapes sure must be sour.
>>
>>100741123
x265 is still relatively early in it's development, it will take time before it gets usable.

>>100741125
I don't think there's a keybind for on/off with xy-VSfilter (which comes with CCCP), but using 's' will switch to the next embedded subtitle, or turn them of if there isn't one. So if there's only one subtitle available it will work as an on/off button.

>>100741261
Maybe, but I still think it will give something similar like that screenshot. Maybe action scenes will look different.
>>
>>100741125
w in MPC-BE. MPC-HC sadly only supports it for ISR.
>>
>>100741067

Never had any problem too with mplayer2/mpv
>>
File: 1388448367567.jpg (55.81 KB, 453x294)
55.81 KB
55.81 KB JPG
You seem to be having a Windows problem, OP, you can start to correct this by only using free software from now on.

Compiling mpc isn't that hard.
>>
>>100741764
*mpv
>>
>>100741806
I thought that was the joke. Would have been funnier.
>>
Any mpv users here get youtube videos working? It's supposed to play them, but I think I missed a dependency when compiling. Everything else is working, but I'm tired of having to combine mpv and youtube-dl to get around using Flash.
>>
>>100741574
So what the hell does "BE" stand for?
>>
>>100742045
You probably missed libquvi. Or Lua, don’t know.

Why not just write a small script for it? Youtube-dl is better than quvi anyway.
>>
>>100741333
>Well, I was posting a real world example on purpose

Not from the world the rest of us is living in, fwiw.
>>
>>100742133
Blood Enema
>>
>>100742163
>Not rewatching Azumanga every week
Pleb
>>
>>100741764
How is replacing his subtitle renderer with a shittier one going to fix OPs problem?
>>
>>100742133
Black Edition.
>>
Is there any way to change madVR's fullscreen overlay back to the default mpc-hc one?
>>
>>100742146
>Why not just write a small script for it?

If you've got a link to some scripting documentation I'd love to try, I've never scripted anything before, but I've always wanted to.

Is it anything like a programming language or am I far off?
>>
>>100742319
Yes, disable automatic fullscreen exclusive mode.
>>
>>100742262
Subs that won't properly render in mpv are the fault of the fansubbing group, as has been said in this thread, compatibility has come along way.
>>
>>100742333
Bash:
mpv $(youtube-dl -g $1)

And that’s it.
>>
>>100742418
Has libass finally caught up with xy in speed again?
>>
>>100742318
Edgy as fuck. Does it even have any features that make it worth?
>>
>>100742438
>not quoting your arguments
Enjoy getting fucked in the ass by neo-browsers that sometimes have you copy-paste actual spaces in URLs.
>>
>>100742565
It's unlocked.
>>
>>100742492
I'm not so sure about that, the subs properly render though, which is what you generally see Micro$hills throwing at it.

I've never had any subtitle rendering problems in mpv, or even in years old repo versions of smplayer/mplayer that I install sometimes before I get around to compiling mpv.

>>100742438
Now I just feel dumb. I was the same way with compiling, afraid of it, then it turned out to be easy. Much like GNU/Linux in general.
>>
>>100742565
Aforementioned turning off subs, configurable compression of screenshots, works better with mouse, preview on seekbar.
Overall, it simply feels nicer to use to me.
>>
>>100742565
it has a black interface
>>
>>100742727
>Aforementioned turning off subs

There's S/Shift+S, most of the time (namely when using the first subtitle track) Shift+S to disable and S to re-enable works just fine.

Wonder why W seems to control the subtitle renderer instead of the splitter in vanilla, though. That's just silly.

>configurable compression of screenshots

I want to stab whoever is responsible for not adding a similar feature to vanilla in the face via IP. Seriously.
>>
>>100742492
It hasn't, but it's getting speed developments now. It needs a better font management for Windows to be considered included in CCCP and the likes, though.
>>
>>100743015
mpv has configurable compression of screenshots, format choices for screenshots, and the ability to take screenshots with and without subs on the fly by modifying your screenshot key with shift.

Does MPV not have this stuff yet?
>>
>>100743015
>There's S/Shift+S, most of the time (namely when using the first subtitle track) Shift+S to disable and S to re-enable works just fine.
True, but that still pauses the video for a second, unlike W.
>>
>>100743278
>>100743015
*MPC
And I can't type today apparently.
>>
>>100743346
It can only save in JPEG and PNG, and its JPEG compression is overly harsh. As for screenshots with/without subs, it depends on the renderer, but no, no it's not as easy as in mpv.
>>
>>100741333
Chances are that Azumanga frame is too blurry to see any significant difference even with bilinear (maybe except for a slight decrease in sharpness)
>>
>>100742640
Not to mention & in urls, yeah. Don’t do this kids, or do it but know what you’re doing.

>>100743178
>it's getting speed developments now
Nice, ETA when?
>>
>>100743278

What >>100743503 said, sadly.

Does mpv have framestepping yet?
>>
>>100743278
MPC-BE has first two, and you can sort of do the latter by fullscreening/defullscreening with XySubFilter.
>>
>>100743554
I dunno man, I'm not really following the development closely.

Here's the branch of guy doing a lot of the optimizing, though: https://github.com/11rcombs/libass/commits/noucome-proof
>>
>>100743278
MPC-HC has JPG screenshots with aggressive compression, resulting in shit screenshots and PNG screenshots with crappy compression, resulting in bloated PNGs.
>>
>>100743178

That's good news. Maybe some competition will finally wake up (xy)vsfilter devs again.

>XySubFilter 3.1.0.546 Beta
>Released: 2013-07-18

I mean, dammit.
>>
>>100743740
mpv is the only player I've used that doesn't spit out 3mb .pngs
>>
>>100743760
http://forum.doom9.org/showthread.php?t=168282&page=13
>>
What specs do I need to do the KCP highest settings?

I have a 550ti and an i5-3570k.
>>
>>100743740
I've solved this by having MPC-HC take screenshots as PNG and then I have a script running in the background which uses ImageMagick to make a decent quality JPEG and recompress the PNG.
>>
>>100744012
I've solved this by using mpv.
>>
>>100744042
His way is cleverer.
>>
>>100744012
Or you could just use MPC-BE. You can disable the black UI if you prefer MPC-HC's look.
>>
>>100744112
Not really.
>>
>>100743740
I just take PNG screenshots and batch-convert/batch-optimize every now and then. Can't say that I'm happy with the situation though.

>>100743877
Ouch.

>>100743278
>mpv screenshot capabilities

MPC-HC tries to retrieve the current frame from the video renderer, and it's up to the renderer to decide whatever it deems suitable, leading to discrepancies between screenshots taken with different renderers including but not limited to: PC vs TV levels, AR correction vs 1:1 pixels, subtitles vs no subtitles, playback resolution vs source resolution.

What about mpv?
>>
File: 1.jpg (9.05 KB, 326x259)
9.05 KB
9.05 KB JPG
>>100744120
Why do you have a UI at all?
>>
>>100744236

It's pretty convenient when using a mouse.
>>
>>100742732
Great, like my soul. Downloading now.
>>
>>100744339
To what end? You're using your video player wrong.
>>
>>100744339
You only need mouse for seeking.
>>
>>100744403
Not everybody has their hands on keyboard all the time.
>>
>>100744434
But you can use it for everything else
>>
>>100744475
Why would you? It's much slower.
>>
>>100744502
How?
>>
File: mpvsnap002.jpg (261.95 KB, 1280x720)
261.95 KB
261.95 KB JPG
>>100743713
Looks good.
>noucome-proof
No, not really, zapped through it a bit, picture related. It really is much better than before though, both in terms of speed and accuracy. Stellar improvements. 10/10, will download frequently.
>>
>>100744042
>>100744112
I did the same thing before I started using mpv, but by hand, if I remember correctly

convert *.png .jpg
>>
>>100744502
>A flick of the wrist is slower than moving your entire arm.

Sure, whatever you say, bro.
>>
>>100744502
Yeah, by like millisecond. Mouse is literally all you need to operate your media player. Well, personally I also use space to pause sometimes and arrow keys to navigate volume for convenience. But no more than that is necessary.
>>
>>100744663
You are using your keyboard wrong.
>>
>>100744339
>It's pretty convenient when using a mouse.

Mice are a crutch. You have a giant customizable 104-button controller sitting in front of you. Use it.

Or not, whatever you feel like man.
>>
>>100744714
You're using your head wrong.
>>
>>100744679
Scroll wheel to control volume, and left click on video to pause.
Keyboard is pointless.
>>
>>100744679
>Mouse is literally all you need to operate your media player.
A keyboard is literally all you need to operate your media player.
>>
I bet some of you own Razer gaming mice...

haha
>>
>>100736628
>>100736947
VDPAU is better than MadVR.
>>
libass is sometimes incredible. Using XBMC on a tablet and 99% animated subtitles work beautifully on a DualCore CortexA9 ARM processor.
>>
>>100744769
>not using whichever one is closer
>>
>>100744848
Logitech MX518. Try again.
>>
>>100744737
>104-button controller
Get out, America.
>>
>>100744769
Mouse vs keyboard fags

taco_girl.jpg

Shit people, we use both. Keyboard is better if you just have one or another but goddamn we're fighting about peripherals.

Take it to the gentoo forums.
>>
>>100744622
Yeah, it's still in development. Doesn't seem like anything's been merge in to main branch yet either. Glad to see it actually improves stuff, though. I think everyone can agree that the faster one can ditch VSFilter the better.

>>100744852
Not really.
>>
>>100744964
VDPAU is better than MadVR.

Yes really.
>>
>>100745058
How?
>>
>>100744852
>inferior scaling
>no smooth motion

Try again.
>>
>>100744663
In what world are you watching a show with your hand on your mouse?
>>
>>100745198
In what world do you have a hand to spare for your keyboard while watching a show?
>>
>>100744946
I'll defend my Model M to my grave.
>>100744947
Using a keyboard is a skill no matter what you're doing with it, it requires training to be used, and even more training to be used on the level of the average GNU/Linux fanboy.

The end result is that keyboard arguers consider mouse defenders to be lazy which, to some extent, is true, using shortcuts or typing the names of programs to open them will always be faster than navigating a menu, especially if your hands are already on the keyboard, which is where they should fucking stay.
>>
>>100733402
>>100733655
>>100733715
>low bitrate x265 encode

http://check2pic.ru/compare/34555/

These are adjacent frames. I don't want this kind of DivX ;-) 3.11 tier shit anywhere near myself, my computer, or even my internet.
>>
>>100743593
mpv has has great forwards and backwards framstepping for a while.
>>
>>100745310
>especially if your hands are already on the keyboard, which is where they should fucking stay.
>where they should fucking stay
>they

Dude, even at the worst of times I need at least one of my hands elsewhere.

And by "elsewhere" I do not mean my mouse.
>>
>>100745304
I have neither but if I'm going to move a hand either way it's way, it's going to be reaching for the giant spacebar.
>>
>>100745084
>>100745186
Open source and developed by GPU engineers and lots of people around the world. MadVR is proprietary and created by some random weeb.

It takes up less resources than madVR and is far more efficient.

It provides equal or better rendering quality.

It provides rendering options madVR in itself doesn't:

Options:

deint Integer (-4 to 4) (default: 0)
chroma-deint Flag (default: yes)
pullup Flag (default: no)
denoise Float (0 to 1) (default: 0.000000)
sharpen Float (-1 to 1) (default: 0.000000)
hqscaling Integer (0 to 9) (default: 0)
fps Float (default: 0.000000)
composite-detect Flag (default: yes)
queuetime_windowed Integer (default: 50)
queuetime_fs Integer (default: 50)
output_surfaces Integer (2 to 15) (default: 3)
colorkey Color
force-yuv Flag (default: no)

Working with mplayer/mpv on linux is far cleaner. MPC-HC and the third-party plugins you need are a mess of spaghetti and extremely bad software design. With this it's clean. Configuration is done in a simple one or two lines in a small text file, you can take this text file anywhere and have the same experience on any install or computer with mpv/mplayer2 and VDPAU. MPC-HC's best decoders are a bad port of native ffmpeg libraries made to work with Directshow, and for everything else you need some crappy third-party plugin, making the entire thing a big mess of software spaghetti.

It can handle GPU based decoding, and GPU based rendering.
>>
>>100745554
Even the keyboardiest of keyboard fags use the mouse for that.

Even if they have to use it with their left hand, fucking amateurs.
>>
>>100745669
Not when the site supports WASD control
>>
>>100745661
>It provides equal or better rendering quality.

Except it doesn't, because of inferior scaling and no smooth motion.

>It provides rendering options madVR in itself doesn't:

It's funny because instead madVR "just werks"
>>
>>100745661
>It provides equal or better rendering quality.
[citation needed]
>>
>>100745661
>MPC-HC and the third-party plugins you need are a mess of spaghetti and extremely bad software design.
This. So fucking much. MPC-HC are some of the worst examples of the downsides of proprietary software and the absolutely logic-defying way windows deals with the installation of programs and codecs.

This isn't freetardation at all, pretty much everything manages applications better than windows, including your phone.
>>
>>100745581
If I want to jump to next chapter, reaching mouse and pressing side button is much faster than using keyboard.
>>
>>100745661

Oh sure, I'll enjoy my madVR on my Windows while you enjoy your ringing, aliasing, and judder on your GNU/SourGrapes. At least you have your Freedom.
>>
>>100745926
>MPC-HC are some of the worst examples of the downsides of proprietary software
MPC-HC is free software.
>>
>>100746070
You're right. The problem is that /a/'s windows users recommend you install a ton of proprietary software with it.
>>
>>100745871
>It's funny because instead madVR "just werks"
MadVR is nowhere near as stable as VDPAU. I don’t really use VDPAU myself because I just don’t see the merit when I’m only watching 10bit anyway, but I have to admit that it’s an impressive piece of software. Probably even on radeon drivers, nowadays.
>>
>>100746204
The only proprietary software I've seen people recommend is madVR.
>>
>>100746204
>one piece of proprietary software
>a ton
You're retarded.
>>
>>100746070
>>100746204
Directshow usually forces players to load the worst kind of proprietary software: commercial video codecs.
>>
>>100746204
Like madVR? And Haali comes with CCCP I guess, though it's not used anymore. Hardly "a ton".

Also, as much as I like to support FLOSS, in the end I usually choose what works the best. madVR is the best renderer around, with features like smooth motion which works incredibly good. That's the main thing that's keeping me from using mpv.
>>
>>100746204
madvr is the only proprietary component and it's optional
>>
>>100746401
Uh no.
>>
>>100746427
And Haali was shipped with CCCP for so long only because of ffmpeg bug.
>>
>>100746450
>accidentally install AcmeCorp Super Fast Video Decoder
>break your mpc-hc setup
>don't even know why
>just reinstall windows!
>>
>>100746401
Assuming that were true: what's the problem?
>>
>>100746037
Page Down
>>
>>100745871
>Except it doesn't, because of inferior scaling

Except it does provide equal or better rendering quality. At the very least better than anything you will get by default with any video player on windows.

>no smooth motion.

This is like complaining about a screen with a higher resolution than a rMBP, but saying it's shit because it's not a "retina screen", retina being a buzzword that means high resolution, not unlike how "Smooth Motion(tm)" means something else.

Can you go ahead and explain to me what "Smooth Motion(tm)" actually does?

>It's funny because instead madVR "just werks"

I wish those third-party plugins all "just werked". There's a reason that stuff like CCCP/KCP, and giant guides on manually installing them exist though.

>>100746047
Okay I will enjoy my GNU/Linux. You go enjoy Windows, the greatest operating system for video games and Adobe software, you true connoisseur of computing.

>>100746306
I believe SVP, Reclock, CCCP itself (not what's included), Haali, and windows itself, is proprietary.
>>
>>100746529
>accidentally install software

There is no excuse
>>
>>100746529
>I'm too dumb to use Windows!

Now that's a new excuse.
>>
>subtitles
>requiring any real horsepower
What? It's not like we're trying to run Crysis here, it's just putting some words at the bottom of the screen. Isn't all this over-the-top typesetting just missing the point?
>>
>>100746606
>Can you go ahead and explain to me what "Smooth Motion(tm)" actually does?
It gets rid of motion judder. You know, the thing that happens when the refresh rate of your display isn't a multiple of the video's frame rate. Which is the case with most computer monitors.
>>
>>100746606
Smooth Motion smooths out the playback of panning shots, which historically have cause stuttering problems.
>>
>>100746606
>Except it does provide equal or better rendering quality. At the very least better than anything you will get by default with any video player on windows.

Backpedaling should be an Olympic Discipline.
>>
>>100746606
>Can you go ahead and explain to me what "Smooth Motion(tm)" actually does?
It blends frames to simulate infinite refresh rate when the frame rate and refresh rate don't match. In such situations without it you get judder.
>>
Smooth motion is a good idea right?

Which option should I check?
>>
>>100746700
>Isn't all this over-the-top typesetting just missing the point?
Kind of. Stupid people doing stupid things with the wrong tools. Not that the right tools would exist, though.
>>
>>100746534
I'd say finding Page Down is slower than just putting your hand on mouse and pressing button under thumb.
>>
>>100746802
>no judder
>get blur instead
>apply additional sharpening on the video
>best video setup ever
>>
File: typesetting.gif (2.76 MB, 320x180)
2.76 MB
2.76 MB GIF
>>100746700
Not when it goes with the show.
>>
>>100746868
>finding page down
This fucking guy
>>
>>100746841
"As long as there would be judder without it" or however its phrased should work fine, but the exact multiple one can also be used.

And yes, it's a good idea.
>>
>>100746709
>>100746728
>>100746802
VDPAU can make your 4 key frames of anime drawings interpolated to 24 frames per second look like they're being played at 60 frames per second.

Source: ftp://download.nvidia.com/XFree86/vdpau/doxygen/html/index.html

>>100746841
It's not.

>>100746792
How did I backpedal? I said it provides equal or better quality in the first post I made.
>>
>>100746868
>having a mouse with extra buttons
Pleb as fuck.
>>
>>100746869
Should I or shouldn't I enable smooth motion for anime?
>>
>>100746438
>>100746401
>>100746332
>>100746306
One piece of proprietary software is enough.

The whole point is to not use proprietary software unless you absolutely have to. Use of it alone advocates its use and further propagates it. Using proprietary software anymore than you have to (BIOS) is contributing to the exploitation of the user by software monopolies and their anti-competitive practices.
>>
>>100746869
>get blur instead
>apply additional sharpening on the video
Two different kinds of blur. Hell, blur from smooth motion isn't even visible.
>>
>>100737177
Maybe on a toaster.
>>
>>100746606
I believe SVP, Reclock, CCCP itself (not what's included), Haali, and windows itself
None of those are recommended or required. Except for Windows, obviously. The only thing you need is MPC-HC and MadVR, now that LAV filters is built into MPC-HC.
>>
>>100746995
No, it's crap and just makes things blurry.
>>
Watch anime on a screen with a refresh rate that's a multiple of 24. Problem solved.
>>
>>100746869
You haven't actually tried it, have you?

No real visible blurring on 24fps content on a 60hz screen.

>>100746995
You should.
>>
>>100746700
Sometimes shows have a lot of words on the screen at the same time.
>>
>>100746975
>How did I backpedal?
A rather long way from "equal or better quality than madVR" to "better than anything you will get by default with any video player on windows" which pretty much equals "toaster-safe bullshit"
>>
>>100746995
Decide it for yourself. Sora no Woto OP is a good video for such test.
>>
>>100747025
The software monopoly of a dude who puts out work for free.

I'm all for FOSS but madshi isn't exactly Apple over here.
>>
>>100747072
>Watch anime on a screen with a refresh rate that's a multiple of 24. Problem solved.
>vfr
>>
>>100746995
Try ti and find out for yourself. It's not like ticking and unticking a checkbox is hard.
>>
>>100746975
>Source: ftp://download.nvidia.com/XFree86/vdpau/doxygen/html/index.html
TL;DR
>>
>>100747156
>of a dude who puts out work for free
Free as in speech, not beer.
>>
>>100747025
Sure is /g/ in here
>>
>>100746975
I didn't see anything there that's close to what madVR's smooth motion does.
>>
>>100747209
I'm well aware. You're still being ridiculous by referring to madVR as exploitative and anti-competitive.
>>
>>100746975
>motion interpolation
>on anime

Fuck off, SVP a shit.
>>
File: Untitled.jpg (545.41 KB, 1920x1080)
545.41 KB
545.41 KB JPG
>>100747035
That didn't show what I wanted it to show for some reason.
>>
>>100747116
>A rather long way from "equal or better quality than madVR" to "better than anything you will get by default with any video player on windows"

I did not switch, I used both at the same time, and both of those points can stand at the same time.

>which pretty much equals "toaster-safe bullshit"

I don't know how you got there. xv renderer is toaster safe. VDPAU decoding/rendering requires a decent GPU actually.

>>100747156
He doesn't put it out for free, it's proprietary.

I prefer Apple by the way.

>>100747255
I assure you that it can make your 4 key frames per second interpolated by the studio to 24 frames per second look like they're being played at 60 frames per second, just like Smooth Motion(tm) does.

You know what I also enjoy doing? Taking 1280x720 video and upscaling it to 1920x1080. It makes it higher quality and you're really getting all the pixels you missed out on, just like frames you are without Smooth Motion(tm).
>>
>>100747265
The use of a proprietary software license further spreads the harmful notion that proprietary software doesn't hide anti-features and exploit customers.

He is not personally anti-competitive, but he buys into supporting practices put in place to keep big firms big and small firms non-existent. I'm going to stop /g/ing out though, because I doubt you want to hear me post about how pirating software is still supporting the developer's monopoly.
>>
>>100747465
You seem to be misunderstanding what smooth motion does. It doesn't generate new frames between existing ones to artificially increase the framerate. It just blends the existing frames in a way that gets rid of the motion judder.
>>
Regarding madVR's smooth motion FRC (frame rate conversion), there's been some attempts at explanation recently in the madVR thread on doom9:

http://forum.doom9.org/showthread.php?p=1661243#post1661243

nb: there's some Reclock discussion going on at the same time, don't get them mixed up.

Basically if the duration of the current screen refresh does not fall completely within the duration of one frame, it'll blend in part of the next frame proportionally.
>>
>>100747465
You really have no idea what you're talking about.
>>
>>100747465
>I prefer Apple by the way.

Fucking justify that hypocrisy.
>>
>>100747465
>both of those points can stand at the same time.

Except "equal or better quality than madVR" cannot stand at any time.
>>
>>100747583
Err, disregard that "part of"
>>
>>100747465
Smooth motion doesn't do interpolation.
>>
>>100747627
I'd like to take this time to point out, as the other free software advocate in this thread, I don't like Apple and probably don't agree with the justifications he'll reply with. It's a human rights issue, they can have a walled garden all they want.
>>
>>100747255
>I didn't see anything there that's close to what madVR's smooth motion does.

I didn't see anything there that's close to the motion interpolation bullshit the other guy is going on about either, for that matter.
>>
>>100747555
> It just blends the existing frames in a way that gets rid of the motion judder.

Cool, so it's really like you're getting those frames you missed out on.

>>100747590
Nice argument.

>>100747627
They develop for stuff like CUPS, Clang, and push a Unix system that is currently the second in terms of desktop usage market-share next to Windows, which is better and less locked down than windows and consists partly of FOSS.

>>100747677
Okay, it has better quality than madVR.

>>100747779
I know, it draws the frames that you were missing before.
>>
>>100747898
>missed out on
But that's not even its objective.
>>
File: gob.png (155 KB, 625x626)
155 KB
155 KB PNG
>>100747898
>>
Is this what /g/ looks like?
>>
>>100747898
>Cool, so it's really like you're getting those frames you missed out on.
>I know, it draws the frames that you were missing before.
What.

Seriously, read up on what it does before you start talking about it.


>They develop for stuff like CUPS, Clang, and push a Unix system that is
And Microsoft is one of the biggest contributors to the Linux kernel.
>>
>>100748140
Used to look like.
Now it's full of retarded macfags and phonetards.
>>
>>100747898
I hate that people give Apple credit for forking off of BSD and then fucking it all up.
>>
>>100748140
not anymore, /g/ is pretty shit now
>>
>>100748140
Nah, on /g/ this thread would've turned into OS wars and arguing about software freedom a lot faster.
>>
>>100748151
>And Microsoft is one of the biggest contributors to the Linux kernel.

HAHAHAHAAAAAAAAAA no they're fucking not, literally they are at the very fucking bottom of people who contribute. They have barely contributed anything whatsoever at all.

Main contributors from companies would be from people like Redhat, IBM, Intel, Google, and Oracle. Those are the main companies you're going to see, people like HP and Microsoft have barely contributed anything whatsoever at all, and when they did it's for their own benefit.

In fact I just ran grep microsoft /usr/src/linux/CREDITS

There's ONE microsoft email.

raymondc@microsoft.com
>>
>>100748510

You should have a doctor look at that anusaggravation of yours.
>>
>>100748510
>>>/s4s/
>>
>>100748510
Based microsoft doesn't even ask for credit, that's how much they care about freedom.
>>
>>100748510
>Jewgle
Botnet as fuck
>>
>>100749582
At least it’s an open source botnet.
>>
>>100749644
They say it's open source.
>>
>>100748956
Every time I hear the microsoft tablet commercial with the girl singing "Honestly, I wanna see you be free!" I can't help but giggle.
>>100749644
Open Source exists entirely as a movement to remove the freedom from free software while keeping the inherent benefits from publicly reviewable code.
>>100749743
Chromium is open source, Chrome is Chromium + Proprietary Google shit.
>>100748510
>>100747898
"Of course, I explain to the local people why they should migrate the machines to free software, but I don't push them hard, because annoying them is not the way to convince them."
-Richard Stallman
>>
>>100749907
Wow, Stallman is less autistic than /a/. That’s something.
>>
>>100750014
His followers aren't.
>>
madVR's smooth motion and XySubFilter are conflicting. What do?
>>
>>100752526
Are you sure it's smooth motion?
>>
>>100752628
Yeah everytime it's off, the subs are back to normal
>>
>>100752786
I don't know, maybe go ask madshi on the forums.
>>
>>100752526
>>100752786
Do you have "optimize subtitle quality for performance instead of quality" checked in the trade quality for performance tab?
>>
>>100753306
Oh thanks, man
>>
>>100752526
Why in the world are you watching gg's JoJo release?
>>
So what filter to be used with mphc?

I'm using Directvobsub and all the LAVs (Splitter, audio & video decoder). I'm also using madvr.
>>
>>100755357
Switch from DirectVobSub to XySubFilter
>>
>>100755621
I did that, I removed it from external filters in options but that little fucker keeps creeping back in.
>>
>>100756222
Add it back and set it to block.

That should work.
>>
Hey guys I'm using CCCP+MPC.It is good but some scenes are not smooth.What should I do to make it smoother ?

Answer please
>>
>>100756651
You enable smooth motion in MadVR settings, and that helps. Though from my experience it does not remove all the judder.
>>
>>100756839
So should I use MadVR instead of Haali ?
>>
Is there a way to change the way MadVR handles things in fullscreen? I'm using KCP and whenever I try to fullscreen a 720p show to my 1080p monitor it maxes out my cpu and skips and stutters.
>>
>>100757005
Yeah
>>
>>100757005
Man you do not know what you are doing just use CCCP.
>>
>>100757049
MadVR Settings > Scaling algorithms > image upscaling
>>
>>100757102
>>100757070
What the hell ? Explain please.
>>
>>100757162
How do I get to the MadVR settings? I'm retarded I'm sorry.
>>
I just follow the guide that an anon posted some time ago.

It just werks.
>>
is there anyway to make subtitles load instantly while seeking through mpc-hc?
>>
What's the current /a/ recommended configuration? I'm running MPC-HC nightly + LAV Filters (vid, audio, splitter) + madVR + xysubfilter. Should Reclock be used?
>>
>>100757971
>[HorribleSubs] Inari, Konkon, Koi Iroha - 01 [480p].mkv
>Let's try high settings as possible
>Jinc all (8 taps AR/LL) + Catmull-Rom(AR/LL)
>Max CPU & GPU queue
>Play without frame drops but this

I can't save her this quality guys, what should i do?
>>
>>100759319
Dude, are you having a stroke? What are you trying to say?
>>
Xysubfilter is shit with madvr. The latter blocks the former.
>>
>>100759742
The latter was actually used to help development with the former, and was the first to support it.
What problems are you having?
>>
>>100759742
I've been using it basically since it came out and I haven't had an issue.
>>
>>100759822
>>100759841

Subtitles don't show.

Alright, so when the user right clicks and hovers over fliters, what should be showing up?

I installed and registered VSfilter (I'm on Windows 32 bit).

I'm betting it's CCCP that's screwing things up. The settings say that it will launch VSFilter but it doesn't. That shit...
>>
File: Untitled.png (53.07 KB, 776x486)
53.07 KB
53.07 KB PNG
>>100760145
Turn that off.
>>
File: Untitled.jpg (98.71 KB, 654x499)
98.71 KB
98.71 KB JPG
>>100760252
Wait shit, wrong picture. Disregard that.
>>
>>100760304
It's already off.

And so are the madvr internal decoders.

I've done everything said in xysubsfilter troubleshooting page except for resetting CCCP settings and "re-registering filters" (I can't find the options in CCCP settings, they probably removed it).
>>
File: 243524526.jpg (276.03 KB, 947x1080)
276.03 KB
276.03 KB JPG
>>100760145
you should see something like this
>>
>>100760514
>>100760145
I'm confused now. Are you trying to install VSFilter or SubFilter?
>>
>>100760518
Thanks. I don't see it.

>>100760552
VSFilter.

Also when I add VSFilter.dll in the external filters options of mphc, what that shitty program adds instead is DirectVobSub (auto-loading).

So basically, Directvobsub = xysubfilters.

It's like the program itself is retarded.
>>
>>100760720
Okay, enable that option I just told you to disable and try blocking DirectVobSub in the external filters.
>>
FUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUCK...

I confused xysubfilter with xyVSfilter.

It works now. Mistake: Assumed both were the same.
>>
>mfw i see windowsfags struggle to get their players working

>mfw mpv works out of the box and is better in every way
>>
>>100761405
Don't you like to struggle? It's fun when you learn things.

Also calm down on your greenspeak, you're not even using it correctly.
>>
>>100762081
>struggle to play a video that would work out of box with even vlc

>i-im l-learning!!

LOL
>>
>>100762118
Who are you quoting?
>>
>>100762157
you
>>
File: 1260317796716.png (23.45 KB, 400x342)
23.45 KB
23.45 KB PNG
>>100762081
I do this for vidya as well. I seem to enjoy it a lot despite the frustrations sometimes, but I'm guessing it's just autism. I cannot rest unless I get it to work.

>take a few hours to mod something
>finally got it to work properly without crashing
>only play 15 minutes of it and then never touch it ever again
>>
>>100762290
I do that too. I love optimizing stuff for the best possible experience.



Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.