[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y ] [Search | Free Show | Home]

Last thread >>61036325 Install mpv: https://m

This is a blue board which means that it's for everybody (Safe For Work content only). If you see any adult content, please report it.

Thread replies: 255
Thread images: 58

File: mpv2.png (54KB, 500x268px) Image search: [Google]
mpv2.png
54KB, 500x268px
Last thread >>61036325

Install mpv:
https://mpv.io/installation/

Wiki:
https://github.com/mpv-player/mpv/wiki

Manual:
https://mpv.io/manual/stable/

User Scripts(including opengl shaders):
https://github.com/mpv-player/mpv/wiki/User-Scripts

input.conf:
https://github.com/mpv-player/mpv/blob/master/etc/input.conf

Vulkan(Linux only for now):
https://github.com/atomnuker/mpv

Test vulkan and post logs if it give you any kind of problems.
>>
how do you set config files for specific directories? wiki and manpage only says how to do file specific configuration
>>
I have tried interleaving two video clips in avisynth for easier video comparison, however have come across a couple things that confuse me.

Why isn't the length of output video the combined length of both videos? When loading the script into MPC, the time reflects that of a single video, even though frames are indeed interleaved. I was under the impression that since frames are interleaved, the video duration gets extended too as frames are effectively doubled. Is this not the case or is MPC just having a visual bug?

Also, is there any way to mark each video clip so that when interleaved and traversing through frames I know which frame belongs to what video?
>>
>>61065766
What is vulkan?
>>
File: 1470851011633.jpg (277KB, 1920x1280px) Image search: [Google]
1470851011633.jpg
277KB, 1920x1280px
Can someone please add a proper file picker thing to the android port?
Thank
>>
is anyone running mpv on linux on skylake/kabylake? can you confirm that vaapi hw acceleration for vp9 video works? my mpv is compiled with the hwaccel vaapi option, but doesn't accelerate vp9 videos.
>>
A faster SSimSuperRes.glsl like shader when?
Even when reducing the tap variable, it's still too much for my toaster.
Adaptive-sharpen works but it's really bad. :(
>>
>>61065810
use-filedir-conf
>>
>>61066842
I write that but what's the filename for the config file. i try writing the directory name.conf but it doesn't do anythin
>>
>>61066865
https://mpv.io/manual/master/#file-specific-configuration-files
>>
how can I automatically download subtitles from opensubtitles.org?
>>
>>61066953
subliminal
filebot
>>
>>61066931
doesn't work. I put that in the main config file and the directory one.
>>
>>61066561
Dude jsut use opengl-hq if your computer is shit. Sharpeners without prescaler like nnedi or ngu AA look like shit.
>>
File: 1472515382202.png (2MB, 1920x1080px) Image search: [Google]
1472515382202.png
2MB, 1920x1080px
CHALLENGE: produce an equally sharp mpv upscale to this MadVR upscale with as little noise and artifacting as this upscale, and post repeatable mpv settings for the upscale. MadVR settings for this are NGU AA medium image upscaling, and adaptive sharpen 1.0 + crispen edges 1.0 + anti-ringing filter under upscaling refinement. This image utilizes AMD denoising but you can use mpv's dn3dHQ for an equivalent effect. This challenge has been posted on /g/ many times and /g/ has failed every attempt at it.
Note: your subjective opinion of how nice this looks is irrelevant to this challenge. This is a technological experiment to see if mpv can produce an upscale this sharp without noise or artifacting.

Refusing to attempt the challenge because of ambiguous buzzwords will be interpreted as an admission of defeat. Posting BS images without an accompanying mpv conf settings will be interpreted as trolling.
>>
>>61067869
here's a raw 720p screen. note the timecode if you want to attempt the challenge.

anyway, your blurry meme mpv player cannot even come close to the parent pic. mpv
>>
>>61065854
post script
>>
Remember to hide and ignore this retard.
>>
>>61067957
>linking to 17 of your own shitposts
Wow you sure showed him.
>>
>foozoor is back at it
Nice.
>>
File: 1477104311486.png (122KB, 245x245px) Image search: [Google]
1477104311486.png
122KB, 245x245px
>>61067869
I like said before, madvr is such a mess that even winfags don't want to use it.
>>
File: 1497639831281.jpg (3MB, 7680x4320px) Image search: [Google]
1497639831281.jpg
3MB, 7680x4320px
>>61067967
Where do you live?
I can give you the phone number of a pretty good ophthalmologist.
>>
>>61067869
https://desuarchive.org/a/thread/158793993/#q158795343
>>
File: 1481545606774.gif (2MB, 320x240px) Image search: [Google]
1481545606774.gif
2MB, 320x240px
>>61068036
Even /a/fags bash him on how bad it looks.
Even anons using madvr.
>>
File: madvrmpv.jpg (1MB, 1920x2160px) Image search: [Google]
madvrmpv.jpg
1MB, 1920x2160px
I'll just leave this here.
>>
File: 1478410700402.png (655KB, 721x436px) Image search: [Google]
1478410700402.png
655KB, 721x436px
>>61068093
>not even trying to hide the shilling
Stay mad ngufag
>>
>>61068111
Enjoy having a player that can't play a playlist. Seriously manually selecting each episode in a over memed player in 2017 is laughable.
>>
File: 1498149589157.jpg (313KB, 612x716px) Image search: [Google]
1498149589157.jpg
313KB, 612x716px
Why does mpv still not have a user friendly interface almost four years after it was released?
>>
>>61068140
But it does
>>
File: nagi.jpg (42KB, 341x361px) Image search: [Google]
nagi.jpg
42KB, 341x361px
>>61068144
>>
>>61068128
>can't play a playlist
It can, but you're a retard that can configure it.
>>
>>61068128
:^)
>>
>>61068140
Because it's a lightweight player.

>>61068159
Don't pay attention to retards.
>>
>>61068159
>not downloading Moozzi2 releases whenever they're available
I can see mpv users don't care about picture quality.
>>
>>61068190
>Moozzi2
>>>/d/oom9/
>>
>>61068190
Now this is some next level shitposting.
>>
>>61068190
nice meme
http://screenshotcomparison.com/comparison.php?id=145999
>>
File: 0be.png (365KB, 673x680px) Image search: [Google]
0be.png
365KB, 673x680px
>>61068220
That's a tremendous improvement. Are you blind?
>mfw grain
>>
File: 535353535.png (43KB, 723x599px) Image search: [Google]
535353535.png
43KB, 723x599px
>>61068190
Stay madvr
>>
>>61068234
>he doesn't like grain in old anime
>>>/r/eddit/anime/
>>
mpv build package or git package for Arch, anons?
>>
>>61068254
I bet you like tape hiss and vinyl hiss, too. Grain, tape hiss and vinyl hiss are all literally distortion resulting from antiquated analog technology.
>>
>>61068255
mpv-build-git
Uses up to date packages
>>
>>61068264
The grain in original image doesn't look bad, the other image looks over-processed and some details are kill
>>
>>61068287
>The grain in original image doesn't look bad
It looks like distortion to me, which it literally is.
>>
>>61068255
https://aur.archlinux.org/packages/mpv-git/
https://aur.archlinux.org/packages/mpv-build-git/
>>
>>61068278
Thank you!
>>
>>61068293
And what do you call the disappearing details? Enhancement?
>>
>>61068220

whoever is defending this filter shit is retarded

there's nothing wrong in the original image that needs "fixing"

if it needed any additional "fixing", whoever remastered the bluray release would have done it already (an actual field PROFESSIONAL), they know best.

kill you are selves.
>>
Regarding the Moozzi2 version,
PROS:
- No grain (grain = distortion)
- Sharper lines
- More vibrant colors
CONS:
- Some autistically tiny details are lost
3 pros and 1 con. That means the Moozzi2 release is objectively better.
>>
>>61068356
Embarrassing.
>>
>>61068356
meant for
>>61068328
>>
>>61068356
SHAME!
>>
File: smugtooru03.jpg (41KB, 358x373px) Image search: [Google]
smugtooru03.jpg
41KB, 358x373px
>>61068329
>if it needed any additional "fixing", whoever remastered the bluray release would have done it already (an actual field PROFESSIONAL), they know best.
Is that way so many "PROFESSIONALS" use fucking Q-Tec for upscaling?
>>
>>61068329
I prefer the original but >>61068379 is right, many of this "professionals" end up fucking the image instead of fixing it.
>>
>>61068412
It's pretty embarrassing for the anime industry that all they ever use for upscaling is bilinear, bicubic or Q-TEC when much better solutions exist.
>>
>>61068356
>grain = distortion
Dumbfuck
>>
>>61068444
It's literally the video equivalent of tape hiss. It's distortion.
>>
>>61068458
Christ.
>>
>>61068458
Abstain from posting.
http://farm5.static.flickr.com/4065/4526515334_480d964309_o.jpg
>>
File: 1485216463690.gif (3MB, 445x247px) Image search: [Google]
1485216463690.gif
3MB, 445x247px
>>61068093
>the city of mpv
>>
>>61068093
Any more samples? Bottom looks better IMO.
>>
>>61068581
t.blindmpvdev
>>
>>61068093
>Not even posting source image
You're trying too hard m8
>>
File: 1495875207038.png (42KB, 500x268px) Image search: [Google]
1495875207038.png
42KB, 500x268px
>>61065766
Why is mpv producing blurry images when enlarging anime?
>>
>>61067134
I'm also having trouble with this
>>
File: 1470024159120.jpg (65KB, 310x310px) Image search: [Google]
1470024159120.jpg
65KB, 310x310px
>>61068743
You're blind?
You should wear your glasses when using a computer you know.
>>
>>61068743
Myopia.
>>
File: 1478073341760.jpg (168KB, 1280x720px) Image search: [Google]
1478073341760.jpg
168KB, 1280x720px
>>61068775
>>61068782
Well, my vision is 20/20. Issue is that other players make sharper anime images. mpv is just so damn blurry!
>>
File: 1469538657199.png (1MB, 1920x1080px) Image search: [Google]
1469538657199.png
1MB, 1920x1080px
>>61068870
>I'm a retard and I don't know how to configure mpv
Fixed
>>
File: 1494268793928.jpg (60KB, 600x474px) Image search: [Google]
1494268793928.jpg
60KB, 600x474px
>>61068140
I thought it was very rude when it wouldn't start up unless you loaded a file, but it still fucks off as soon as your media ends.
>>
>>61068943
There's a way to solve this, read the manual.
If don't want to read it, just use madvr
>>
>>61068910
Please do share you magical config that doesn't make blurry anime.
>>
>>61068971
>in b4 slow-ass NNEDI3 or slow-ass Vapoursynth or blurry-ass ewa_lanczossharp
>>
File: 1492770096260.png (77KB, 500x500px) Image search: [Google]
1492770096260.png
77KB, 500x500px
>>61069088
>buzzwords
>no config
>>
File: 1497537240471.png (239KB, 549x560px) Image search: [Google]
1497537240471.png
239KB, 549x560px
>>61069153
>slow and blurry are buzzwords
>>
Are these timings alright?
>>
>>61069210
How do I get those numbers?
>>
>>61069210
Why is that upscale so blurry?
>>
>>61069210
Depends.
>>
>>61069210
If your monitor is 60hz you should keep it below 16ms

>>61069222
Stats.lua
https://github.com/Argon-/mpv-stats/

>>61069231
That upscale is fine, if you want a sharp mess like >>61067869 stay with madvr
>>
>>61069210
look pretty blurry kohai.
>>
>>61069279

I agree.
>>
>>61069279
>>61069299
>>61069231
Looks fine to me

>>61069274
Below 16ms is below 16000 right?
>>
File: 1473947174535.jpg (23KB, 600x337px) Image search: [Google]
1473947174535.jpg
23KB, 600x337px
>>61069328
>Looks fine to me
check your prescription. you're blind as a bat.
>>
>>61069368
Like >>61069274 said, if you want a sharp mess like >>61067869 just use madvr with the same settings as the "CHALLENGE".
Some people don't like to fuck with the image to the point it doesn't even look like the original art style.
>>
>>61068943
Good, but I would have preferred a binding to toggle between stopping at the end of the current file or continuing to the next file, with the latter as the default as it is with the default --keep-open=no.
>>
File: 1489388392557.jpg (77KB, 720x960px) Image search: [Google]
1489388392557.jpg
77KB, 720x960px
>>61069499
meant for >>61068964
have a birdie
>>
File: MPC-HC.png (1MB, 1920x1080px) Image search: [Google]
MPC-HC.png
1MB, 1920x1080px
Stock MPC-HC vs stock mpv
http://screenshotcomparison.com/comparison/213607
>>
>>61069445
why is this so blurry
>>
>>61069525
MPC-HC vs VLC
http://screenshotcomparison.com/comparison/213608
>>
>>61069525
ctrl+s to take upscaled ss in mpv
Also download the BDs
>>
>>61069562
>>61069368
>>61069279
>>61069299
>>61069231
samefag >>>/out/
>>
>>61069562
Doesn't look that blurry to me.
>>
File: 1389245650395.jpg (9KB, 302x225px) Image search: [Google]
1389245650395.jpg
9KB, 302x225px
>>61067869
>mfw on linux and can't get it to look this nice
>>
>>61069562
>>
File: 1498256099817.png (155KB, 623x651px) Image search: [Google]
1498256099817.png
155KB, 623x651px
>>61069574
Why would I take an upscaled screenshot if I'm going to be viewing the anime as is? Are all of these circle-jerking screenshots really upscales for the sake of being able to upscale without blurriness or pixelation? What's the purpose if you never need the upscale?
>>
File: 1493823155485.jpg (124KB, 1280x720px) Image search: [Google]
1493823155485.jpg
124KB, 1280x720px
>>61069620
I know. I boot into Winshit to watch anime. Thats the only time I do it. I feel dirty for it but it's worth not having to see blur everywhere.
>>
>>61069622
He probably doesn't even know what a BDRemux is.
He is just shitposting
>>
>>61069641
MadVR and mpv are both capable of producing better upscales than the anime studios do. You'll get better results upscaling 720p BD rips with MadVR or mpv than you will watching 1080p BD rips at native resolution when the anime in question is listed on anibin as having 720p lineart.
Anime studios typically use bilinear, bicubic or Q-TEC upscaling, all of which are inferior to some of the options available to MadVR and mpv.
>>
>>61069641
>What's the purpose if you never need the upscale?
you have a 1080p screen and you download 1080p anime? waste.
>>
File: 1469854725166.png (380KB, 800x450px) Image search: [Google]
1469854725166.png
380KB, 800x450px
>>61069620
>>61069656
Shilling this hard
>>
>>61069664
That makes sense, thanks for the explanation.
>>
>>61069672
>>61069664
This,most anime is produced in sub-1080p resolution and anime studios use shit upscalers, just some studios like kyoani produce their anime in 1080p,
>>
File: fuck.png (12KB, 363x481px) Image search: [Google]
fuck.png
12KB, 363x481px
>>61069664
>>61069672
>>61069689
Welp, now I can delete all of my anime and replace it with 720p rather than 1080p.
>>
>>61069750
Use anibin
He analyses almost all airing anime and determinates the resolution.
Is in japanese tho
>>
>>61069664
>>61069689
>>61069672
These guys know what's up. 720p is also better because most of the times they are even higher bitrate encodes than 1080p since many groups don't want to upload 1GB files.

1080p has about 50% more pixels than 720p yet most 1080p encodes are maybe 30% larger.
>>
File: done.png (34KB, 126x155px) Image search: [Google]
done.png
34KB, 126x155px
>>61069769
Thanks, every day is a school day. Some are just more painful than others.
>>
>>61069750
Not all of it. For the stuff listed between 720p and 1080p on anibin, if you already have the 1080p version, you should probably download a single episode of the 720p version and compare yourself to see which looks better. KyoAni anime are always worth the 1080p version.
>>
>>61069947
>KyoAni anime are always worth the 1080p version.
Er, that is, Hyouka and later KyoAni anime. Their pre-Hyouka HDTV productions are 720p lineart.
>>
>>61069956
So is Kyoto making their anime in 1080p so there is no need for upscaling? I guess I'll need to do research before downloading every series now. I'll probably end up collecting physical BDs like a legit weeaboo.
>>
>>61069672
1080p freeleech man.
>>
>>61070010
Why they do it is anyone's guess.
I'm not too sure about this but I think I've heard that they use larger paper to animate, too. More resolution in the final composite would warrant the expense of larger paper somewhat.
>>
Hook limit is still 16 lol.
>>
>>61070200
https://github.com/mpv-player/mpv/blob/master/video/out/opengl/user_shaders.h#L25
>>
>>61070217
https://github.com/mpv-player/mpv/blob/master/video/out/opengl/video.c#L48
>>
File: 1470261160252.png (309KB, 512x772px) Image search: [Google]
1470261160252.png
309KB, 512x772px
vulkan on windows when?
>>
File: 1475207585773.png (44KB, 911x235px) Image search: [Google]
1475207585773.png
44KB, 911x235px
>>61070200
it's not like it matters much... most of these options are placebo and don't matter to mpv output quality anyway. still blurry.
>>
>>61070662
That's because you are blind
>>
>>61070695
maybe so but image diff doesn't lie.
>>
>>61070707
What the mpv image tell me is that you like sharp mess
>>
>>61070707
That's not how you compare sharpness, retard
Use vqmt for that
see
http://forum.doom9.net/showthread.php?p=1803263#post1803263
>>
>>61068190
>Moozzi2
L-looking good, there
http://screenshotcomparison.com/comparison/213614
>>
Please add support for more hooks. Please.
>>
>>61072290
Make a fucking issue on github, jesus.
>>
>>61072336
I dont know how. Some anon said that they were working on increasing it already!
>>
>>61072418
https://github.com/mpv-player/mpv/issues/new
There you go, create an account if you don't have one and submit your issue.
Since it's a feature request, don't forget to describe what your use case is.
>>
>>61072480
Actually it's a bug now, not a feature request
>>
>>61072608
Then report it as such, nothing is going to happen if you keep it to 4chan.
>>
tfw fell for sharp madvr meme
https://diff.pics/kQTNPY9PKmEa/1
>>
What's a hotkey in MPV for going to specific frame or specific time?
>>
>>61073009
Not sure which scalers you used, both images are terribly aliased. Try NNEDI3 for mpv, or NGU AA for madVR.
>>
Give me the most basic but best conf available.
>>
>>61073201
I don't care about aliasing
>>
>>61073181
.,
>>
Did anyone use doom9 or videohelp prior to /g/ ?
>>
>>61073226
Wow. Ok. Then why don't you use Nearest Neighbor sampling? ;)
>>
>>61073578
Pixelation and aliasing isn't the same thing.
>>
>>61068190
Moozzi's Flip Flappers release appears to be better than BDMVs since they look somewhat washed out.
http://screenshotcomparison.com/comparison/213629
>>
>The authors of SRCNN have published another paper about accelerating SRCNN(FSRCNN).now it can run in real-time performance

Can mpv devs implement it? Ive heard its supposed to be good with anime especially.
>>
ewa_lanczossharp is for a markov model on the popularity in steam emoji
>>
>>61073857
This shitty sharpening may look okay in some cherry-picked scenes, but it's going to be oilpaint-fest in more detailed ones.
Besides I don't understand why you would want to bake this destructive filtering instead of doing it at runtime.
>>
>>61074834
In this comparison it also helps reduce some banding in the source.

What do you say a proper encode should target then? Compress the source without sacrificing too much quality and let users apply their own filters per video source? I don't think most people cherry tune their setup for each series they watch and just stick to whatever configuration came with the set up guide, whereas people that care will probably encode their own releases either way.
>>
>>61073857
>warpsharp
yeah no that's completely indefensible trash
>>
File: 1491593911749.png (1MB, 1920x1080px) Image search: [Google]
1491593911749.png
1MB, 1920x1080px
Is this a good mpv upscale? It's mostly defaults.
>>
>>61075736
No. It uses bilateral on defaults or something. Use profile=opengl-hq and scale=ewa_lanczossharp.
>>
File: 1498356062357.png (2MB, 1920x1080px) Image search: [Google]
1498356062357.png
2MB, 1920x1080px
>>61076080
He's the one samefagging about blur and using bicubic. Not even bilateral looks like that.
>>
>>61076385
Oh, i see. He is getting subtler.
>>
File: 68708461.jpg (15KB, 251x206px) Image search: [Google]
68708461.jpg
15KB, 251x206px
Please add FSRCNN to mpv!
>>
why are people constantly begging for more meme shaders
nothing is ever good enough for you people
>>
>>61073857
That's a massive improvement, as are most of Moozzi2's releases. mpv users will disagree since they like blur.
>>
File: yuzuconsider.jpg (361KB, 1506x997px) Image search: [Google]
yuzuconsider.jpg
361KB, 1506x997px
/g/ will never fairly evaluate MadVR because it is proprietary, Windows-only software.
>>
>>61076849
I kinda starting to hate madvr because of that blur spamming retard.
>>
>>61076656
THIS. ANIME DEMANDS IT.
>>
>>61076748
It's just one shitposter.
>>61076849
It being proprietary and windows-only is part of the evaluation.
>>
>>61076849
I used madVR for years before switching to mpv because mpv is simply better software.
>>
Is there a way to have mpv somehow display a movie in black and white? Without transcoding the video?
>>
>>61077001
Yes.
>>
>>61077001
lower saturation?
>>
>>61077001
--vf extractplanes=y
>>
>>61077044
That's not universal.
>>
>>61077250
yeah cause you've got so many RGB movies right
>>
>>61076656
If madshi does it first mpv is finished.
>>
File: 1462962288258.jpg (184KB, 1486x840px) Image search: [Google]
1462962288258.jpg
184KB, 1486x840px
>>61076656
I support this
>>
File: 1340134431361.png (120KB, 225x286px) Image search: [Google]
1340134431361.png
120KB, 225x286px
>>61077274
>he doesn't know about my DCI remuxes
Keep watching straight to dvd anime, pleb.
>>
>>61076873
This, I have used madvr for over a year now but that retard is beyond salvation.
>>
>>61076944
Both have their advantages and disadvantages
>>
>>61077468
>JPEG2000
isn't that shit still YUV
>>
>>61071102
Moozzi2's release here is such a dramatic improvement, it's almost as much of an improvement as MadVR over blurpv.
>>
How to please shitposter-san:
1. warpsharp my shit up
2. crank the saturation way it
Congrats you're now a respected doom9 member
>>
>>61077691
There are people who like sharpening on /g/ and doom9. There are people who hate sharpening on /g/ and doom9. Same with warpsharp.

Saying that all doom9 members like warpsharp and saturation boosts is like saying that all Muslims are terrorists. It's both not true.
>>
>>61077691
>people with different opinions than me are shitposters
"No."
>>
>>61077623
t.Too Retarded To Configure MPV
>>
>>61077848
Stay blind fag
>>
>>61076799
>>61077623
Stop samefagging.
>>
>>61076385
>>61075736
those look extremely blurry. why is mpv so blurry???
>>
>>61078151
Because he is using the default configuration.
>>
>>61078151
That's the purpose, retard.
>>
>>61078171
Stop responding to the retard. If someone says "blur" just ignore him, he is autistic or baiting.
>>
I bet that's the same ogm autist.
>>
OK, I've updated my mpv config and it now looks like this.

COnfig:

profile=opengl-hq
scale=ewa_lanczossharp
cscale=ewa_lanczossoft
dscale=ewa_lanczos
video-sync=display-resample
interpolation=yes
tscale=oversample
sigmoid-slope=10


Any good?
>>
File: 1495140327531.png (2MB, 1920x1080px) Image search: [Google]
1495140327531.png
2MB, 1920x1080px
>>61078324
Sorry, forgot pic. Here it is.
>>
>>61078354
Blurry.
>>
File: mpv-shot001.png (2MB, 1920x1080px) Image search: [Google]
mpv-shot001.png
2MB, 1920x1080px
>>61078354
:^)
>>
The one thing I don't understand is why he keeps using the same bad image for his challenges. We don't even have a high-res reference frame to compare to, so judging which upscaled image is "better" is all subjective.

Anyone who posts a challenge should have a non-biased high-res reference frame, to make objective judgement possible. E.g. the Passengers Blu-Ray / UHD Blu-Ray screenshots posted earlier would make for an excellent challenge.
>>
>>61078461
Yuru Yuri has 720p lineart according to anibin. The 1080p BD release is a standard shitty anime studio bilinear/bicubic upscale. It's not worth posting.
>>
File: 1480829875196.png (1MB, 1920x1080px) Image search: [Google]
1480829875196.png
1MB, 1920x1080px
>>61078422
>>61078439
I added this:
scale=ewa_lanczossharp:scale-antiring=1

Better?
>>
>>61078439
Blurry.
>>61078559
Blurry.
>>
>>61078568
BlurriYuri
>>
>>61078568
Hmm, OK. Can you show me one that's not blurry? I'll then know what to shoot for.
>>
>>61078559
This >>61078439 is with your configuration, stop using angle backend.
>>
File: kyoukolaugh.gif (140KB, 285x286px) Image search: [Google]
kyoukolaugh.gif
140KB, 285x286px
>>61078585
>>
>>61078601
see
>>61067869
>>
>>61078633
That looks like shit.
>>
>>61078641
I'm sorry it doesn't satisfy your blur fetish.
>>
File: 1476388598188.png (2MB, 1920x1080px) Image search: [Google]
1476388598188.png
2MB, 1920x1080px
>>61078633
>see
>>>61067869
Hmm... I'm not sure how to get anything sharp like that.
Here's another try with
scale=haasnsoft


PS: This thread has reminded me how much I love Kyoko.
>>
>>61078709
You really need to stop pretending you're taking actual screenshots.
>>
Is this autism?
>>
>>61078803
yes, who's asking?
>>
File: 1483484348222.png (1MB, 1920x1080px) Image search: [Google]
1483484348222.png
1MB, 1920x1080px
>>61078773
I can try some other anime if you want.
Here's one of the new ones... 720p source --> 1080p. Using the settings above from >>61078324

How does this one look like? Better?
>>
>>61078888
Come on man. It's clear that you're taking these screenshots with the default settings.
>>
>>61078920
Just ignore.
>>
File: mpv-shot0001.jpg (391KB, 1920x1080px) Image search: [Google]
mpv-shot0001.jpg
391KB, 1920x1080px
>>61078888
Fixed
>>
File: 1473595110101.png (1MB, 1920x1080px) Image search: [Google]
1473595110101.png
1MB, 1920x1080px
>>61078920
Nope. With the config above.
Why do you try to deny truth?
>>
File: 1483687373940.png (88KB, 522x473px) Image search: [Google]
1483687373940.png
88KB, 522x473px
>>61079021

you just applied unsharp mask in GIMP. look at the edges... they're full of unsharp mask fingerprints.
>>
File: 1463816727066.jpg (38KB, 362x346px) Image search: [Google]
1463816727066.jpg
38KB, 362x346px
Why shinchiro updates his builds more often than lachs0r?
Any difference between the two?

>>61078920
Just ignore that samefagging retard.
>>
>>61079066
lachs0r just updates when there is a release, whereas shinchiro whenever he wants to.
>>
>>61078888
Enable log and search for "dumb". If youre not trolling maybe angle is broken for you.
>>
>>61079058
That's compression artifacts, fag
And that's a crop of the first original pic
>>
>>61079150
name the anime, EP and timecode.
>>
>>61079190
Just try to argue with a retard
>>
>>61079190
Idk, I'm not him >>61078888
>>
File: 1492000437427.png (1MB, 1920x1080px) Image search: [Google]
1492000437427.png
1MB, 1920x1080px
>>61079208
>>61079215
Then quit naming your bullshit troll images with "mpv-shot00x.jpg" names since none of these images in this thread with those names came from mpv.

Here's another test but this time with haasnsoft scale.

The hostility of some people in this thread when presented with actual screenshots and settings is quite shocking to me!
>>
>>61079280
>since none of these images in this thread with those names came from mpv.
retard detected
>>
File: 1497316580737.gif (2MB, 320x320px) Image search: [Google]
1497316580737.gif
2MB, 320x320px
>>61079280
>since none of these images in this thread with those names came from mpv.
>>
Original challenge guy here. >>61067869 isn't even me, so there's now more than one of us helping Windows users make an informed choice of media player.
Thanks >>61067869 as you're doing God's work.
>>
>>61079868
>you're doing God's work.
By showing to everyone how shitty madvr can be?
>>
>>61079868
Thanks, just installed mpdn.
>>
>>61079868
It's amazing how butthurt mpv fanboys are getting. They refuse to post settings and screenshots that those settings produce and instead post fake image that are Photoshopped. When someone posts real output of mpv, they accuse you of lying.

There's two or three insane people on here who can't deal with (blurry) truth.
>>
>>61080084
t. a retard
>>
>>61079868
Madvr is such a mess than even winfags don't want to use it.

>>61080084
>Photoshopped
Fucking wow! Madvrfags are simply retarded
>>
File: chitoseshrug.jpg (47KB, 475x371px) Image search: [Google]
chitoseshrug.jpg
47KB, 475x371px
>>61080084
I didn't expect the hostility my challenge has generated, and I didn't know whether to expect mpv would be able to equal/surpass my challenge or not. I merely posted it out of intellectual curiosity as to the relative technological capabilities of the two media players. I'm continually baffled by those who interpret as a contest to produce a subjectively better looking image rather than a technological experiment to see which media player is capable of producing a sharper image without noise or artifacting.
>>
>>61080203
That's not a challenge, you're simply blind if you really think that's a good upscale.
>>
>>61080203
>sharper image without noise or artifacting.
your definitions of those things are all as warped as your upscale
>>
>>61080222
It is a sharp, noise-and-artifact-free upscale, regardless of your subjective opinion on how nice it looks. mpv users have thus far failed to produce an equally sharp, noise-and-artifact-free upscale, regardless of your subjective opinion on how nice it looks. This suggests that MadVR's technological upscaling capabilities are currently more advanced than mpv's, regardless of your subjective opinion on how nice it looks.
>>61080255
Can you point out objectively definable artifacts in >>61067869 such as ringing, aliasing, etc.?
>>
>>61080290
>your subjective opinion doesn't matter because my methods are objective
your image is an objectively oversharpened piece of shit and the hideous flattened lineart is absolutely an artifact that reflects that
>>
>>61080290
>over-sharpened mess is an advanced upscaling capability
Confirmation that winfags are retards
>>
>>61080337
>hideous flattened lineart
This is a subjective improvement, not an objectively definable artifact. Removing blur from the edges of the lineart is the entire point. You may not like the way it looks, but others do.
>>
>>61080337
>>61080402
Also, the width of lineart is easily customized in MadVR via the anti-bloating filter and the thin edges filter. It's easy to enable/disable sharpening repeatedly, adjusting the above two filters each time, until the width of the sharpened lineart matches the width of the unsharpened lineart, since MadVR updates the picture in realtime as you apply new settings. This enables one to produce exceptionally sharp, crisp, clear upscales without altering the width of the lineart.
>>
>>61080402
>objectively definable artifact are only those that mentioned on madvr upscaling tab, which are blur, aliasing and ringing.
Another confirmation that winfags are retards
>>
This is the kind of autism the retard party at doom9 produces.
Remember kids, stay away from doom9.
>>
>>61080203
Your challenge is generating hostilty because people are just not interested in it (besides your friend from doom9), yet you keep posting it over and over.
>>
>>61080575
I haven't posted it for a while since it appeared nobody was going to complete it. The guy who posted it in this thread probably did so precisely because of the hostility it has generated prior to this thread.
>>
>>61080607
It generated massive butthurt because mpv produces blurry output no matter what kind of placebo meme settings you add.
>>
>>61081225
>mpv produces blurry output
see >>61079021
mpv is just not for imbeciles
>>
>>61079021
Blurry and noisy.
>>
>>61081318
t. blind retard
Make a better one in memevr
>>
>>61081318
>Blurry and noisy.
That's not even an mpv output. He just took my image, sharpened it and saved it as jpg with that "mpv-screenshot" name. He's just a butthurt troll. He's been screeching like mad from his immense autism in this thread.
>>
>>61081387
This >>61079021 is my first post in this thread, retard
Also I don't see any proofs that your memevr screenshots aren't a photoshop.
>>
>>61076656
>>61076891
>>61074682
I hope they will not wasting their precious time with neural upscale, at least this year because there are tons of things to improve in mpv and I know all those shits take time to write, debug, test and so on. I checked the github activity of all the mpv "commiters" and it's easy to see they are pretty active people and mpv is just a side project for them.
>>
File: mpv_2017-06-25_19-35-10.png (2MB, 856x1030px) Image search: [Google]
mpv_2017-06-25_19-35-10.png
2MB, 856x1030px
>>61065766
Wtf,

Here's the stats on window mode (1920x1080):
http://i.imgur.com/Qr3RIRT.png

The stats on fullscreen (2560x1440) is on pic related.

Have I done something wrong?

profile=opengl-hq
opengl-backend=dxinterop
opengl-shaders="~~/SSimSuperRes.glsl,~~/adaptive-sharpen.glsl"
scale=ewa_lanczossharp
video-sync=display-resample
cscale=ewa_lanczossharp
dscale=mitchell
tscale=oversample
>>
>>61081318
So far I've only seen screenshots where madvr looks like an oversharpend shit or blurry mess like >>61073009
madvr is literally a meme from what I've seen.
>>
>>61081593
>about 10 commits/day.
>side project
>>
>>61081593
Upscaling is the #2 thing mpv needs to improve behind a user-friendly interface.
>>
>>61081658
Why not just increase oversharp in SSimSuperRes instead of using sharpener on a sharpener.
>>
>>61081705
>user-friendly interface
Already done.
>>
>>61081705
It is already near perfect, the only problem is aliasing.
>>
>>61081705
GUI is literally never going to be implemented in mpv.
The great thing about its design is that libmpv allows you to create whatever GUI you want, and people have already done that (see IINA).
>>
>>61081681
Except wm4. ;)

>>61081715
Is it really better than adaptive-sharpen?
>>
>>61081715
How I would go about doing that? I'm a recent Ex-VLCfag.
>>
>>61081658
blurry blurry
>>
>haasn
>1 contribution in private repositories
Is it libplacebo or nextgen vulkan renderer?
>>
>>61081786
You open it in a text editor, search for oversharp and change the number.
>>
>>61081796
Stay blind madvrfag
Thread posts: 255
Thread images: 58


[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y] [Search | Top | Home]

I'm aware that Imgur.com will stop allowing adult images since 15th of May. I'm taking actions to backup as much data as possible.
Read more on this topic here - https://archived.moe/talk/thread/1694/


If you need a post removed click on it's [Report] button and follow the instruction.
DMCA Content Takedown via dmca.com
All images are hosted on imgur.com.
If you like this website please support us by donating with Bitcoins at 16mKtbZiwW52BLkibtCr8jUg2KVUMTxVQ5
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.
This is a 4chan archive - all of the content originated from that site.
This means that RandomArchive shows their content, archived.
If you need information for a Poster - contact them.