[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vp / vr / w / wg / wsg / wsr / x / y ] [Home]
4chanarchives logo
An observation: I feel snes games really benefit from shaders
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.

You are currently reading a thread in /vr/ - Retro Games

Thread replies: 255
Thread images: 120
File: goyim-shader.png (542 KB, 1192x896) Image search: [Google]
goyim-shader.png
542 KB, 1192x896
An observation: I feel snes games really benefit from shaders that emulate the effect of a crt.
The pixels really melt together through the crt shader and create a higher res texture that gets sharp again through the scanlines. raw textures just look wrong

on the contrary i think nes games look ugly with crt shaders. the crt effect+scanlines dont add anything to the textures because they are so low res it's in vain. the scanlines make it look worse as opposed to on snes. i like the raw textures more, but a bit of blur and smudging through an ntsc filter seems to add some texture to the low res but this is more of a preference for some people. i personally like to play nes games without any filter but on a low res mode 512x480 with a natural blur by upscaling of the hdtv.

gameboy games feels also really wrong to add filters. dot matrix is a preference for some people and it is okay. without any filter it looks the most right. scanlines look really ugly here especially as they have no place on a gameboy, it looks atrocious as the original hardware never had scanlines and it just feels wrong i think we can all agree. yet blurring or any kind of filter also feels wrong. Playing in 1x or 2x integer scaling with a gameboy borders feels the most right to me as those games look wrong on a big screen, they are meant to be played small. Again a bit of blur is okay rendering it in a lower res mode. I play for example with the border in 768x576 mode.

So my point is snes games benefit from and look right when emulated with CRT+scanline shaders, while nes and gameboy games feel off with crt+scanline shaders as there is no texture enhancement gained by the effect, except hiding the pixels.

This again all this is just my preference and I don't claim this to be the right way to play it.
What do you guys think? WHat feels right for you?
>>
>yet ANOTHER crt thread
keep it contained in your general >>2842270
>>
>>2852456
it's a crt shader thread ~
>>
>>2852456
go back to posting final fantasy threads, fag
>>
File: logo-large.png (3 KB, 284x115) Image search: [Google]
logo-large.png
3 KB, 284x115
>Several paragraphs about playing games on CRT televisions
>Likely reached the max character limit for a single post
>>
File: say_no_to_blur.png (373 KB, 549x1192) Image search: [Google]
say_no_to_blur.png
373 KB, 549x1192
>>2852451
>because they are so low res it's in vain
They're the exact same resolution as SNES games (excluding the few rare exceptions like the naming screen in Secret of Mana).

And filters are entirely personal preference -- artists certainly didn't all intend their work to be seen with blurred pixels (pic related).
>>
>>2852451
Autism: the post
>>
>>2852451
Contains crt threads, they are fucking multiplying
>>
>>2852451
emulating is garbage in general.
>>
>>2852492
/thread
>>
The obsession with CRTs has nothing to do with graphical technicalities. It's just wanting to simulate the experience of a video gaming period in the past.
>>
>>2852492
Correctly configured emulator + correct hardware can give you something indistinguishable from the original except with better latency, making it objectively better. But this is very unlikely - doubt more than 100 people have ever done it.
>>
>>2852481
with los res i mean nes sprites are just simple looking, while 16bit snes sprites are more high res texture looking because they imitate high res texture and a shader or a real crt blends those 16bit texture together to a higher res texture, while with 8bit texture there is not a lot to blend together to begin with.
>>
>>2852496
A strobed display is the only way to get good motion quality from console/arcade games. CRTs give you strobing with zero latency penalty. (if you emulate you can use black frame insertion instead)
>>
>>2852497

>better latency

I'm not so sure about that.
>>
>>2852509
120Hz CRT + Retroarch + KMS or GPU hard sync + BFI + Frame delay + Parallel port input

It's possible.
>>
>>2852496
Ignorance is bliss.
>>
File: raw.png (20 KB, 1024x896) Image search: [Google]
raw.png
20 KB, 1024x896
Raw.
>>
File: ugly.png (562 KB, 1024x896) Image search: [Google]
ugly.png
562 KB, 1024x896
Ugly.
>>
File: geom-flat.png (318 KB, 1024x896) Image search: [Google]
geom-flat.png
318 KB, 1024x896
Good.
>>
File: hyllian.png (129 KB, 1024x896) Image search: [Google]
hyllian.png
129 KB, 1024x896
Good, too.
>>
>>2852451
SNES widely benefits for being the better made 16 bit console of its time. That's why it's so enduring today, despite what some on /vr/ would complain. The games look good and play well. That's why games of the quality of Super Metroid, as you posted, is almost expected of the later Metroid games.
>>
File: corrected_aspect.png (77 KB, 1192x896) Image search: [Google]
corrected_aspect.png
77 KB, 1192x896
>>2852519
Best, but wrong aspect ratio. The SNES does not have square pixels so some minor blurring is unavoidable unless you want highly restricted image sizes. Pic related.
>>
>>2852538
Nope, 4:3 is the wrong aspect ratio in Super Mario World. It's always game dependent. Some games were developed with anamorphic pixels for 4:3 ratio, others games were developed with square pixels in 8:7. Snes only outputting 4:3 is a flaw from the console, and adding 4:3 to any game despite having the option of 8:7 in an emulator is just nostalgia, unlike blending 16bit textures with a shader, which is adding of definition and resolution and not just nostalgia.
>>
File: pic-related.jpg (542 KB, 1679x677) Image search: [Google]
pic-related.jpg
542 KB, 1679x677
>>2852538
pic related.
(for the record: i wouldn't use snes9x - it's incorrect)
>>
>>2852514
That is totally ridiculous though.
>>
File: non-integer example.png (6 KB, 821x464) Image search: [Google]
non-integer example.png
6 KB, 821x464
>>2852561
The left pic is doubly incorrect. Not only is the aspect ratio wrong, but also it is using non-integer ratio nearest neighbor scaling.

>>2852570
The only uncommon things are the CRT and the parallel port controller. And even without them latency is very good.
>>
>>2852573
I'm just saying that's a lot of equipment to approximate a console experience.
>>
>>2852573
You can stay in denial about your nostalgia. The scaling is not the point about this picture. 4:3 is just nostalgia. But do remember, there are also games designed for 4:3 and they stretch out to 4:3. You see you grew up with the stretched picture, naturally you can't tell anymore subjectively if a picture is stretched or not. Me, as part of a younger generation, I have an unbiased view and naturally choose the format which looks more right to me. The human vision has a thing for automatically choosing approximately the correct aspect ratio. But for example if you watch all 4:3 movies stretched to 16:9 long enough you get used to it. Now imagine a whole generation growing up with 4:3 movies stretched to 16:9. They couldn't tell what's right and what's wrong in terms of aspect ratio years later anymore.
>>
File: ct_wrong.png (29 KB, 768x633) Image search: [Google]
ct_wrong.png
29 KB, 768x633
>>2852596
4:3 is objectively correct, and all the games with competent artists designed for it. See pic.
>>
File: ct_right.png (80 KB, 844x633) Image search: [Google]
ct_right.png
80 KB, 844x633
>>2852587
That's to *exceed* a console experience.

>>2852597
And with corrected aspect ratio, as it was displayed on the real hardware, the moon is the right shape.
>>
File: 1448745543829.png (72 KB, 966x383) Image search: [Google]
1448745543829.png
72 KB, 966x383
>>2852597
not true. not all games. It's a 50:50. Please stop with the denial. It's embarrassing. Just move and live on .
>>
File: 1449256959366.png (226 KB, 1668x672) Image search: [Google]
1449256959366.png
226 KB, 1668x672
>>2852597
>>
>>2852604
>>2852606
And neither of these have graphics as good as Chrono Trigger.
>>
>>2852607
I beg of you, for the love of god. Please stop the denial and just acknowledge the fact that not all games were designed for 4:3. It's just cringeworthy at this point.
>>
>>2852608
I never claimed *all* of them were, only the ones with the best art.
>>
>>2852609
Okay that's good with me. I'm not interested in game wars.
>>
>>2852451
>I really like scanlines and shit that interferes with the image
>>
>>2852604
>>2852606
These images are functionally identical.
>>
File: filtervs.jpg (396 KB, 1576x1044) Image search: [Google]
filtervs.jpg
396 KB, 1576x1044
>>
>>2852658
>posting a emulator bug example in a filter thread
>>
File: 1440707758761.jpg (138 KB, 900x675) Image search: [Google]
1440707758761.jpg
138 KB, 900x675
>>2852456
>>2852491
>>2852496
>>2852626

The best thing about this thread is all millennials jumping on OP with torches and pitchforks "WE HATE CRTS WE HATE CRTS", while misunderstanding the very fundamental criticism of his post, saying that while crts and scanlines add a lot to snes games, they are almost useless for nes and futile for gameboy games.
>>
>>2852504
>>2852518

I prefer crisp, sharp graphics. Latency is only an issue for twitch games. Due to the limits of human vision and the way most games are designed, optimizing for low latency will only yield imaginary benefits.
>>
>>2852705
>I prefer crisp, sharp graphics.
You can get this with a high resolution CRT + emulator or line doubler.
>>
I am confused by this thread.

If I use original hardware on 15 Khz CRT TV will my round objects be oval?
>>
>>2852773
They'll be the same as they were on the original SNES. Some artists accounted for the non-square pixels, some did not.
>>
>>2852606
>>2852604
>>2852561
What I am wondering is: If Nintendo knew this, why did they still make the aspect raio 8:7?
>>
>>2852606
second one looks round enough to me
>>
>>2852828
because they are kucks.
>>
>>2852828
>What I am wondering is: If Nintendo knew this, why did they still make the n64 have cartridges instead of a drive?
>>2852828
>What I am wondering is: If Nintendo knew this, why did they still make the wii a meme console for non gamers?
>What I am wondering is: If Nintendo knew this, why did they still make the wii-u a graphically underpowered compared to its competitors?

You see a kucked pattern?
>>
When the fuck did /vr/ start hating CRTs?
>>
File: amarec(20151201-214420).png (240 KB, 640x480) Image search: [Google]
amarec(20151201-214420).png
240 KB, 640x480
>>2852538
>>2852604
>>2852606

Here we go again!
>>
>>2852782
Does this mean I have to adjust my emulator display settings on a per-game basis?
>>
NES/SNES DISPLAYED GRAPHICS BY PRINTING TILES OF EVEN WIDTH AND HEIGHT (AKA SQUARE TILES)

HOWEVER THIS DOES NOT MEAN ANY TILE GRAPHIC IN ANY GAME WAS MEANT TO BE SQUARE OTHER THAN WITH ARBITRARY ASSUMPTIONS

However we may ever give the benefit of the doubt.
>>
>>2852978
only if u want to be an asshat
>>
>>2852979
>NES/SNES DISPLAYED GRAPHICS BY PRINTING TILES OF EVEN WIDTH AND HEIGHT (AKA SQUARE TILES)
They did not. The tiles had even horizontal/vertical pixel count, not width/height.
>>
>>2853003
Tomato and stuff.
>>
File: get.jpg (66 KB, 600x898) Image search: [Google]
get.jpg
66 KB, 600x898
>>2852481

I feel like every time this image gets posted I lose a few brain cells. Hear me out and consider this:

The artist was merely trying to represent what it looked like on the monitor on paper. In doing so, it would not make much sense to intentionally blur the pixels on paper. On paper, the more defined look makes sense. That however, does not mean that that was the intended look. They were just replicating it on paper.

You need to prove that the artist intended for that to look the way it did on paper because I have never seen any evidence for the claim nor have I seen it logically thought out at all. I have only seen people claim that 'WELL, SINCE THEY DREW IT ON PAPER THIS MUST MEAN IT WAS WHAT WAS INTENDED TO LOOK LIKE ON TELEVISIONS, SO THIS JUSTIFIES MY USE OF SHITTY FILTERS".

So yeah, consider that.
>>
I kinda get what you're saying, OP. I really only emulate back to Megadrive & SNES though, so i'm hardly very picky about earlier gens. I did like the way Shovel Knight came off, but much prefer SNES on muh PVMs, so maybe I agree.

>>2852525
>>2852530
>good
Good job holding a flyscreen up in front of your LCD, you mean. The first shader is godawful, and even if the second is better it's still worse than the raw shot. A real CRT, on the other hand, just works.
>>
>>2853189
I think that's a good point. I'd imagine that the manual images were based directly on the graph paper drawings that were used to design/plan the enemy sprites. This is easier to reproduce than trying to blur it or add scanlines, not to mention that it just makes more sense for the medium of paper. Not to mention that, if they were competent graphic designers, they planned for the typical method of viewing by the consumer. For instance, your sprite might look great and crisp and readable on graph paper, but if it looks like an incomprehensible mess when viewed as a 1 inch tall object on the average display of the time, then it was a bad sprite and should be redisgned to read better since players aren't looking at your graph paper.
>>
File: comparison.png (4 MB, 1280x1920) Image search: [Google]
comparison.png
4 MB, 1280x1920
I only know that raw looks like shit.
>>
>>2853448
I don't believe in scan lines/CRT shaders for 3d games. you can just render them in any higher resolution. integer scaling doesn't matter. You can play in 720p on a 1080p and it will still look better than both these messes. Please keep scan lines contained in 2D Pixel art games.
>>
>>2852553
could you shut up faggot?
>>
>>2853489
I don't give a shit what you believe. FFVIII is primarily a 2D game. The backgrounds are pre-rendered, and the upper screenshot is what it looks like unfiltered. Increasing the internal resolution will only make the 3D models stick out even more. Increasing the internal resolution of early 3D games is a no-no in any case. Do you think games like MGS, Dino Crisis and Vagrant story were exempt to phosphor blur and shadowmasks/scanlines?
>>
I always play vanilla windowed 2x or 3x the original resolution, why are you so obsessed about filter shit.
>>
>>2853520
the backgrounds are prerendered photorealistic graphics and not pixel art, thus bilinear filter looks better. the 3d models and full 3d battles can be uprendered to your prefered resolution. you don't have to go full blown 4K for ultimate sharpness. you can go with lower resolutions like 540p or 720p to blend them more in with the backgrounds. increasing internal resolution is a nice go-go for 3d games. You can get a nice look with a nice dithering filter and anti jittery filter. Last but not least it's really correct dithering that makes the ps1 games look nice. The dithering if not configured correctly in higher resolution can look wrong, like make the sky purple in silent hill. Shadowmasks and scanlines are just there to hide the low resolution and blend the dithering. Again it's all about the dithering, since it isn't pixel art there is no reason no hide the low res, as you can render in a higher resolution to smooth out the pixelated 3d.
>>
File: ffight.jpg (241 KB, 1280x720) Image search: [Google]
ffight.jpg
241 KB, 1280x720
lol at these plebs dicussing about SNES' inferior aspect ratio.
>>
File: 1430002147738.png (1 MB, 2884x1879) Image search: [Google]
1430002147738.png
1 MB, 2884x1879
>>2853520
>>2853547
here is what i meant with wrong dithering (or actually no dithering)
>>
>>2853547
>the backgrounds are prerendered photorealistic graphics and not pixel art, thus bilinear filter looks better
It really doesn't. The pre-rendered backgrounds are low resolution. Bilinear filtering will only make them appear even more pixelated. It doesn't matter if it's a low resolution pre-rendered CG background or pixel art; phosphor blur makes them look better all the same.

CRTs were commonplace when these games came out, and it's how they were intended to look. Go ahead and fire the PC version of FFVII or FFVIII up and you'll see that they look like absolute dog shit compared to how you remember them from when you played them originally on a CRT.
>>
File: 1437871701551.jpg (114 KB, 1280x960) Image search: [Google]
1437871701551.jpg
114 KB, 1280x960
>>2853520
>>2853547
>>2853550
and here is correctly upscaled + 16-bit colors + color dithering + Monitor Dot-Matrix filter
>>
>>2853550
Sorry I don't believe in 3d + CRT, it's sacrilege for me. But I can understand if you value it, it's just not my thing.
>>
>>2853551
Even with the latest plugins that remove z-fighting issues, I'll take the original low-res image either on a CRT or with a 4k CRT shader
>>
File: 1432766095069.png (4 MB, 2390x1792) Image search: [Google]
1432766095069.png
4 MB, 2390x1792
>>
>>2853568
Purfectodah
>>
>>2853567
That's okay, hunty. Everyone has his preferences.
>>
>>2853568
let's imitate that with correctly configured crt royale. i wanna see a comparison.
>>
File: crt_metroid.png (2 MB, 2048x1792) Image search: [Google]
crt_metroid.png
2 MB, 2048x1792
>>2852451
>NES doesn't benefit from shaders
>>
File: zelda_uhd.jpg (2 MB, 3840x2160) Image search: [Google]
zelda_uhd.jpg
2 MB, 3840x2160
>>2853520
Thank god there was a better console this gen, with some decent full 3d games, with some even looking gorgeous in 4K.
>>
>>2853585
no, it still looks pixelated and the textures don't form a higher res texture by melting together - they are just a pixelated mess with or without crt, unlike snes textures.
>>
>>2853590
Pretty dumb criteria you invented there, IMO
>>
>>2853587
Don't forget OoT also has some pre-rendered backgrounds, and they look like shit even on the original console.
>>
>>2853585
what shader is that? Royale?
>>
>>2853606
omg true -sorry, i forgot. but there are like 5% of the game, so no one bothers, because the advantages outweigh.
>>
>>2853580
It can be done, but making the blue colors pop out that much would probably require ReShade/SweetFX with Vibrance affecting only that color.
>>
>>2852481
>no dithering intended

It's the keys of a fucking typewriter
>>
File: RetroArch-1214-094527.jpg (3 MB, 2691x2016) Image search: [Google]
RetroArch-1214-094527.jpg
3 MB, 2691x2016
>>2853568
>>2853580
This is 4 minutes of tweaking. View them in fullscreen fully zoomed in if you don't have a 4k monitor. It's not perfect, but anything's possible with some effort. ReShade and further tweaking CRT Royale should take care of the color differences, for those who'd care enough to go through the trouble.
>>
>>2853646
hyllian has blue boost.
>>
>>2853694
this look gorgeous. a shame that royale looks so ugly in 1080p and it's really a beast of a filter. I think my PC couldn't handle it @4K. But it is definitely the most gorgeous filter judging from this.
>>
>>2853694
That's some shit tier bloom. Someone needs to rewrite their algorithm.
>>
File: 4x-integer-snes_raw-vs-hyllian.gif (598 KB, 1024x960) Image search: [Google]
4x-integer-snes_raw-vs-hyllian.gif
598 KB, 1024x960
>>2853541
>le crt shaders and scan lines don't add anything meme
>>
>>2853694
That is a piss poor simulation of a CRT.
>>
File: 4x-integer_raw-vs-royale.gif (775 KB, 1024x960) Image search: [Google]
4x-integer_raw-vs-royale.gif
775 KB, 1024x960
>>2853703
Here is the second comparison with Royale. Mind that it's only 1080p mode with 4x integer scaling. It's just the standard settings, I haven't configured it.
Okay so with standard settings and 1080p I think royale shader sucks when it comes to snes games, because it deteriorates the image more than adding something to it. Also the colors get messed up, which is a no-go for purist. Again Royale can do much better in ultra hd modes, see >>2853694 and compare with real crt >>2853568 for example.

Secondly I want to add while royale looks shit in 1080p when it comes to snes games, it kinda adds some resolution to nes games. More to that in the next post.
>>
File: 4x-integer-nes_raw-vs-hyllian.gif (161 KB, 1024x960) Image search: [Google]
4x-integer-nes_raw-vs-hyllian.gif
161 KB, 1024x960
>>2853710
>>2853601
>>2853585
Secondly I'd like to apologize from claiming in the OP that crt + scan line shaders don't add anything to the nes experience. I was wrong. The shaders definitely add some definition compared to the raw image.
>>
>>2853708
Not really. With some tweaking, that could definitely be more or less identical to the reference image. Don't be salty just because modern technology is making your archaic devices obsolete.
>>
File: 4x-integer-nes_raw-vs-royale.gif (326 KB, 1024x960) Image search: [Google]
4x-integer-nes_raw-vs-royale.gif
326 KB, 1024x960
>>2853713
Last but not least here is the royale shader on 4x integer scaling in 1080p. It adds some interesting detail to the textures that hyllian doesn't, altough it messes up the colors. But my point is royale works for nes in 1080p and adds something special to the image, while with snes it deteriorates the image. But I guess this may be game depending. In FFVI in definitely deteriorates the image in 1080p mode.
>>
>>2853701
Agreed. It's the weakest part of CRT Royale. It's not really visible when you're sitting a few feet away, but it could always be improved.
>>
>>2853718
1080p CRT Royale can look nice with some tweaking. Even if you don't have a native 4k TV or monitor, you could always downsample. It's just a few steps away if you have an Nvidia GPU. You won't get the quality of native 4k, but it'll still be a huge improvement over 1080p.
>>
>>2853448
>>2852451
>>2852530
Problem with such shaders is that they hurt eyes more than real crt. Maybe it's ok on hidpi display i don't know.
>>
>>2853723
I wish I could, but my hardware is really shit-tier. I basically play with 512x448 hyllian-fast for a noise free play.Hyllian get's loud, i hate that without headphones, so I only play with 512x448 hyllian with headphones on, on my 2012 meme laptop. (nvidia 540M)
>>
>>2853189
It's evidence *against* use of blurring filters.

>>2853448
Neither of those are "raw". Try it with integer ratio nearest neighbor.

>>2853687
Exactly, that's why no dithering is intended.

>>2853547
>prerendered photorealistic graphics and not pixel art, thus bilinear filter looks better
Bilinear is terrible for everything. For scaling photorealistic graphics see:
http://www.imagemagick.org/Usage/filter/nicolas/
>>
>>2853805
Bilinear is more natural than nearest neighbor.
>>
>>2853805
autismo
>>
>>2853805
>Bilinear is terrible
yeah but you get my point. any scaler is better than integer nearest neighbor.
if you wanna be picking straws i can tell you mitchell or catmull rom look better than integer scaling without any interpolation on photorealistic graphics = not pixel art.

>>2853810
exactly!
>>
>>2853854
If the graphics were purely photorealistic then you would be correct. But it's mix of photorealistic and aliased low poly 3D, so whatever filter you use will be wrong. The least worst option is integer ratio nearest neighbor, because that's the easiest to upscale in your own brain.
>>
>>2853862
hunty, if you render in 540p or 720p the 3d graphics get an upgrade and there is a natural filtering/upscaling like bilinear or catmull by your hdtv to 1080p for the backdrops. and you can additionally add filters/shaders for the dithering.
>>
>>2853862
>because that's the easiest to upscale in your own brain
No it isn't.
>>
File: noise-good.png (257 KB, 512x512) Image search: [Google]
noise-good.png
257 KB, 512x512
>>2852451
>The pixels really melt together through the crt shader and create a higher res texture that gets sharp again through the scanlines.


Absolutely not. All shaders, scanlines and CRT technology basically just throw a layer of noise over the pure image.

For some people they may like the effect it gives, and that's fine. But it's not adding any real detail or making it higher resolution at all. It's exactly the same as putting a grain filter on an image. The added noise gives an impression of detail, but it's an illusion. There's no more real information, what looks like more detail is in reality just meaningless noise.

Shaders, scanlines and filters can blur, blend and distort the original image in a bunch of ways. And again, if that's what you like then congrats. But none of them add more real detail.
>>
>>2852481
This faggot again. The Chrono Trigger shot doesn't even blur on a CRT with composite. Try again.
>>
>>2853710
Adds resolution? Do you even know what you're talking about? This shit is goofy, I'm out.
>>
>>2853989
>none of them add more real detail
>>> >>2853703
say again?
>>
>>2854008
>The Chrono Trigger shot doesn't even blur on a CRT with composite
You misunderstand the point of that image. It is demonstrating non-dithering detail using an identical pixel pattern to dithering. This proves that dithering is not supposed to be blurred, because it would also destroy non-dithering detail.
>>
>>2853989
The art was designed to be viewed through this noise. So not using crt is basically distortion.
>>
>>2854015
I'm saying I understand it, that Chrono Trigger on a real CRT with composite doesn't dither it so badly as to ruin the image. You can tell what it is.
>>
>>2854021
Which is more proof that everybody claiming blur filters are required to smooth out dithering are wrong.
>>
File: 1450083692815 copy.jpg (697 KB, 1024x960) Image search: [Google]
1450083692815 copy.jpg
697 KB, 1024x960
>>2854013
I shouldn't have to say it again. That's precisely what I'm talking about. It doesn't add any detail, just noise that you are imagining is detail. I'm not going to bother animating, but you can compare these yourself.

Again, if you like it better with the scanlines. Or hell even if you like it better with a noise filter, then enjoy. It's your game. Just don't claim it's adding detail when it's not. Noise and detail are not the same thing.
>>
>>2854018
>The art was designed to be viewed through this noise
No, it was not. See >>2852481
>>
File: 1449081164044.jpg (495 KB, 1920x1200) Image search: [Google]
1449081164044.jpg
495 KB, 1920x1200
More proof that sharp pixels were intended.
>>
>>2854018
I still fundamentally disagree that any of these artists were working with the crt scanlines and have never seen a convincing case for it. I'm not this guy >>2852481, but at least he has a bit of proof in some way.

Either way it really doesn't matter, people should make their games look however they like when they play them.

But for me and my tastes, I think all retro games look magnitudes better on modern screens than any CRT anywhere.
>>
>>2854031
It has got retconned because of pixel art meme
>>
>>2854026
You are retarded
>>
>>2853585
Filters still struggle to get the CRT "glow effect" right, crt dont glow that much.
>>
File: Super_Mario_Bros._box.png (100 KB, 280x383) Image search: [Google]
Super_Mario_Bros._box.png
100 KB, 280x383
>>2854046
It's no retrocon. Sharp pixels were always intended.
>>
>>2854052
But really they weren't.
>>
>>2854058
I've posted 6 examples supporting sharp pixels. Nobody has posted any evidence supporting blur.
>>
File: mario comparison.jpg (330 KB, 1160x684) Image search: [Google]
mario comparison.jpg
330 KB, 1160x684
>>2854046
Did they ever display Mario with scanlines and CRT blur though? Any advertisement they ever made either had pixel art Mario or a cartoon. Even here on the box, the picture they show is very clearly made to look like fancy pixel art with some motion blur added. But no indication of scanlines.
>>
>>2854052
This was true up to about 86/87 when 16 bit hit the arcades and developers started to take advantage of CRT's monitors, to do "tricks" in order to make the ilusion that more colors and detailed graphics were displayed, this was also used on home consoles later on.
>>
>>2854052
Even though consoles couldn't output them? You have to use imagination and envision perfectly suqare pixels in display buffer I guess.
>>2854064
See >>2852451>>2852658>>2853585>>2853710
>>
>>2854067
Which tricks are we talking about?
>>
>>2854067
Chrono Trigger typewriter - intended detail would be destroyed by blur
Alien Carnage waterfall - dithered transparency was never blurred because of line doubling
The "tricks" are done in the human brain. All blurring does is destroy information.
>>
>>2854070
That meant to be examples for >>2854061
>>
>>2854064
>>
>>2852658
>>2854067
fact
>>
>>2854058
Do you have a single example anywhere? Idealized classic video game art that shows scanlines? An artist saying they're intended part of the design? Anything like that at all?
>>
>>2854070
>Even though consoles couldn't output them?
Mode 13h/mode X DOS games
PC-88/PC-98
All portable systems
>>
There are seriously people denying that dithering exists?
>>
>>2854074
That's just showing what they looked like on the only displays people would have available. I'm talking about an idealized view of what Mario is supposed to look like that shows scanlines. It's always either full on cartoon or pixel art.
>>
>>2854074
That's not scanlines, that's halftoning. It's an artifact of the printing technique.
>>
File: tmnt.png (1 MB, 800x1200) Image search: [Google]
tmnt.png
1 MB, 800x1200
>>2854071
see this:
>>2852658
also... pic related
>>
>>2854085
>see this:
That's non-integer ratio nearest neighbor scaling

>also... pic related
That's an emulator bug
>>
>>2854085
The top is no more detailed looking that the bottom. Blurriness is not detail. The only real issue is that the color on the bottom pic is off.
>>
File: 21111602[1].jpg (284 KB, 850x1196) Image search: [Google]
21111602[1].jpg
284 KB, 850x1196
>>2854087
What bug exactly?
>>
File: 1446854257189a_01.png (2 MB, 772x1024) Image search: [Google]
1446854257189a_01.png
2 MB, 772x1024
>>2852658
I can tweak the color balance to make it look blue too. It has nothing to do with scanlines or blurring.
>>
>>2854091
The colors are wrong. There is no fake-scanline filter that changes colors like that. It's a deliberately misleading troll image.
>>
>>2854087
owning both CRT's and LED monitors, as well as some arcade cabinets i customized, i can say that in fact that pic is true, hooking up any arcade game directly to a pc monitor gives that pixelated washed out look, in contrast to the much richer colours and "pop" on both 15khz monitors and TV CRT's. some say it has to do with the fact that CRT displays true black colour, and also the tube inside but im no tech genius in that matter. so, my best advice is to get some CRT's and LCD'LED and try it for yourself.
>>
>>2854096
>LED monitors
You don't have an LED monitor. LED monitors are huge outdoor wall sized things. You have an LED backlit LCD monitor, which looks identical to a CCFL backlit LCD at best, and worse than a CCFL if they PWMed it.

>hooking up any arcade game directly to a pc monitor gives that pixelated washed out look
If it's a CRT monitor then it will look identical except for sharper pixels (using integer ratio nearest neighbor), which makes it better.
>>
File: rivercr_1.png (2 MB, 783x1005) Image search: [Google]
rivercr_1.png
2 MB, 783x1005
River City Ransom scan from Nintendo Power. Note that they didn't add scanlines to the pixel art they took right from the game.
>>
>>2854085
But that's a bug in the emulation code of early Konami arcade machines.

>>2854072
>dithered transparency was never blurred because of line doubling

No, it was never blurred because VGA was a RGBHV signal which is clear enough not to blend pixels HORIZONTALLY because raster scanning is done from left to right and not from above to below. So line doubling basically means nothing in dithering's name.

SNES's composite encoder just didn't meant to blur screens which were only 256 pixels wide.
>>
>>2854101
I do own those huge montiors you talk about too, my brother and i have a mobile publicity bussiness, but thats another story, what i meant was LED TV's and LCD pc monitors, my english is also not the best.
>>
>>2854101
>You don't have an LED monitor.
Yes he does. LED backlit LCD monitors are simply called LED monitors now. No one genuinely confuses LED monitors with outdoor display signs. You're being needlessly pedantic.
>>
>>2854110
>are simply called LED monitors now

Might as well call any piece of software "app" from now on.
>>
>>2854105
Of course, why the fuck would they add scanlines?
But note that sprites are blurred.
>>
>>2854114
Not in this board as only downloadable smartphone software were called apps pre-2000.
>>
>>2854106
>>2854101
Samefag here, as i said earlier here:
>>2854096
You should try hooking up Arcade systems to different kind of monitors and see the result's by yourselves, it interesting how the image changes so drastically even between different kinds of CRT's.
>>
>>2854114
Please don't tell me I have to explain how language evolves over time.

Imagine yourself in a normal situation. Your aunt buys a new TV, says something like "this new LED tv I got was on sale!" or something like that. No one is confused as to what that means. If you replied to that with

>You don't have an LED monitor. LED monitors are huge outdoor wall sized things. You have an LED backlit LCD monitor, which looks identical to a CCFL backlit LCD at best, and worse than a CCFL if they PWMed it.

You would come off as a douchebag of colossal proportions. But who knows, maybe that really is how you act in real life.
>>
>>2854115
The blur comes from the printing process so it's inevitable. But yes, clearly they wouldn't add scanlines to it. No one ever does because it makes it look worse.
>>
>>2852451
> i think nes games look ugly with crt shaders.
No.

>dot matrix is a preference for some people and it is okay. without any filter it looks the most right.
Not even remotely close.
>>
Nintendo Power's 1993 calendar. Pixel art all over the place. No scanlines or CRT blur and bloom.
>>
File: Nintendo1990Calendar-05-April.jpg (437 KB, 1000x1200) Image search: [Google]
Nintendo1990Calendar-05-April.jpg
437 KB, 1000x1200
Again, I'm not saying that anyone shouldn't play on a CRT or use scanline filters if that's how they like it and think the game looks best. But Nintendo's idealized view of what their classic games were supposed to look like has always been pixelated, not blurred and distorted by scanlines.
>>
>>2854114
Programs were always referenced by the term application, apps is just a shortening of that and the term apps has been around since even windows 3.11

>>2854110
>LED backlit LCD monitors are simply called LED monitors now.
They're not.

>>2854124
>You would come off as a douchebag of colossal proportions.
Actually the feel good motherfucker who has the balls to complain about someone being a douche when they're explaining why they did not receive a thing they said they did. Not every console is a Nintendo and not every car is a Pinto. They have names for a reason and an LED monitors are a different thing. Using the term to reference your shitty same old TFT panel monitor makes it worse for when and if LED monitors do become popular and not OLED, but say actual graphene based nano LED monitors. When people use the term LED monitors they're being a douche by fucking up the usage of the word to mean something it's not which will hurt a superior technology when it arrives at the market. Don't be that guy calling everything you see a playtendo.
>>
>>2854131
Though one could use a CRT like the super gameboy as well. That was a thing.
>>
>>2854132
Note that hq3x filter applied to these flying things.
>>
>>2854146
Are you drunk? HQX didn't even exist back then. The only actual pixel interpolation filter at that time was used on Macs to upscale 320x200 content to 640x480.
>>
just type on youtube (insert game here) Arcade cabinet and shut the fuck up everybody.
only guys that are too young to ahve actually owned, o played on a CRT think retro games look better on LCD displays.

https://www.youtube.com/watch?v=cVGQHhZuGas
>>
File: Nintendo1993Calendar-06-vgo.jpg (1 MB, 2000x3000) Image search: [Google]
Nintendo1993Calendar-06-vgo.jpg
1 MB, 2000x3000
>>2854146
>>2854146
They're not actual filters. All through the calendars they play with different pixel resolutions, that's clear. Point is though, it's pixel art and doesn't show any indication of scanlines.
>>
>>2854148
>>2854151
Relax, I am just joking about how badly these are drawn with some pixels 2 times smaller than other.
>>
>>2854151
You do realize that poster was made digitally? you CANT print a display effect, unless is photograph taken from the actual display.
>>
File: led1.jpg (52 KB, 409x192) Image search: [Google]
led1.jpg
52 KB, 409x192
>>2854140
Apparently you do need the continuing evolution of language explained to you, but I don't have the time or inclination so I'll just mock you. Also I was apparently spot on about you being a colossal douche in the real world. Fun.
>>
>>2854106
>line doubling basically means nothing in dithering's name
Line doubling meant each logical pixel was made of 4 physical pixels, which made them much sharper.

>>2854115
The blur is very minor. It's much less than people add with emulator filters.

>>2854124
And WTF are you going to call LED monitors when they are actually released? It's entirely possible that this will happen in your lifetime, because LEDs can be produced on silicon substrate now, making them much cheaper.

>>2854146
People apparently managed to fuck up integer ratio nearest neighbor even when they were drawing it by hand.
>>
File: Nintendo1990Calendar-12-November.jpg (557 KB, 1000x1200) Image search: [Google]
Nintendo1990Calendar-12-November.jpg
557 KB, 1000x1200
>>2854153
>you CANT print a display effect
Of course you can. If the game was really intended to look like it had scanlines all over the place it would be more than possible.

Actually I have to partially recant this, since after having gone through almost 10 years of Nintendo power calendars, I found one example. The fireballs and wings/eyes of the metroid there show scanline artifacts reproduced.

So it's not never ever, it's just almost never.
>>
>>2854126
Lol thats a lie. I work in print, You can print pixel art, you need a magnifying glass to see any blur in print.
>>
>>2854154
By promoting this misnomer you are hurting the chances of getting real LED monitors. People will assume it's the same LCD shit they already have and refuse to upgrade.

>>2854164
>you need a magnifying glass to see any blur in print
Not everybody pays for the most expensive super-fine halftoning, and certainly not back then.
>>
>>2854164
He's talking about this image. >>2854105 Those sprites weren't blurred on purpose, they only look blurry because of the printing. And if you don't think Nintendo power had crappy printing that looks blurry when you photograph it close or scan it, then you've never seen one first hand.
>>
>>2854170
Text are clearly more sharp than sprites.
>>
>>2854162
You obviously know nothing about graphic design, or whatsoever, those posters where obviously made in the early to mid 90's i suposse, back then digital image editors didnt have the capcity to add "scanline filter" or "CRT filters" the images where REPRODUCED by the artist on software not taked directly from a photograph of a monitor display dumbass.
>>
>>2854171
The text isn't halftoned.
>>
>>2854168
"Real" LED monitors are an inevitability. That people now call backlit LCD, LED won't slow that process down one bit and when they do come out they'll be named something distinctive.

You can be upset about it if you want, but languages are ever changing and fighting against that happening is futile.
>>
>>2854159
>Line doubling meant each logical pixel was made of 4 physical pixels

Horizontal, raster scanning.
R, a, s, t, e, r, , s, c, a, n, n, i, n, g.

Rasterization. Electron gun. Line by line. Parallel. Systematic. Rake.

In case you still don't get it, the only thing it did was hiding scan lines better and give out, of course, a more sharp, uniform picture. Blur has nothing, and I mean nothing to do with line doubling.
>>
>>2854172
> the images where REPRODUCED by the artist

That's exactly what I'm saying. They are all idealized "takes" on what the games looked like. As I showed there, it's perfectly possible to paint them in such a way that they look like they're made of scanlines. But the simple fact is that virtually no one ever does. They almost always represent them with big clear pixels.

And I know quite a lot about graphic design, by the way.
>>
>>2854179
>"Real" LED monitors are an inevitability.
And people thought SED was an inevitability.

>>2854183
VGA is designed for 640x400. When you display 320x200 using line doubling, the size of the pixel is big compared to the size of the mask features (4 times bigger than it would normally be). This makes the pixels sharper.
>>
>>2854189
>VGA is designed for 640x400

Utter nonsense, unless you're talking about the graphic standard.

> the size of the pixel is big compared to the size of the mask features (4 times bigger than it would normally be)

Then you'll also need to stretch the pixel horizontally. Why don't you also mention this openly instead of just focusing on line doubling? Maybe because pixels are not a thing anymore when traveling through analog signals?

>This makes the pixels sharper.

No. it makes them more compact vertically, which is not the way graphics are drawn printed to the screen. Because raster scanning. Again I assure you sharpness has nothing to do with line doubling.
>>
>>2854185
Let me rephrase it, take an screenshot of you playing a game on an emulator, then print it on your pc printer, how does it look? thats how they did that.
>>
File: mario_gb_border.png (236 KB, 768x672) Image search: [Google]
mario_gb_border.png
236 KB, 768x672
>>2854131
I don't like that shader. It looks incorrect to me. This is not even remotely how the gameboy looked. This is more like gameboy pocket screen, but not really close. It doesn't even look like a gameboy pocket to me, more like a kids toy imo.
Btw if you read the post correctly it said that all this is just preference so saying something like
>Not even remotely close.
is totally missing the point because by that you claim to know better than everybody else and sound like a total douche. I also like to add that I apologized for saying nes feels more right without shaders, as I corrected myself here >>2853713

Last but not least here is a pic of how I play Gameboy games. I play with 2x integer scaling and a border in 768x576 and then upcale by the TV to 1080p. The extra black space if for SGB borders. For me b/w without any filters with the lightest shade being white looks the most right to me. Again this is just my preference. Previously I was also playing with shades of yellow/green, but now I changed my mind and think b/w is more correct after thinking a long time, coming to the conclusion that dot matrix or crt/scanlines shades don't add anything to the experience of the gameboy and where never meant to be. It is just nostalgia and the technical limitations of that time, that led to the display having a dot matrix and the screen a green tint. In my opinion of course.
>>
>>2854203
Well that's what they used as the basis of the paintings, yes. But that's beside the point. I am talking about the way in which they interpret what the game is ideally meant to look like. That's what all this is about.
>>
>>2854208
The ideal way is on your display, thats how it was meant to look, other art are artist representations, most of the times saida rtist's didnt even play videogames man, a bunch of them were punk artist, that didnt even owned a VCR of game console, like the guy that did the art for OCEAN games.
>>
>>2854198
>the graphic standard
That's what VGA is.

>Then you'll also need to stretch the pixel horizontally.
Obviously.

>instead of just focusing on line doubling
Because focusing on line doubling make it very clear why mode 13h had sharp pixels.

>I assure you sharpness has nothing to do with line doubling
So you think 640x400 non-line-doubled pixels on a VGA CRT are just as sharp as 320x200 line-doubled (logical) pixels on a VGA CRT? Anybody with working eyes can see this is false.
>>
>>2854205
Can someone photograph a Game Boy?
>>
>>2854216
See >>2852481
>>
File: example_01.png (76 KB, 510x258) Image search: [Google]
example_01.png
76 KB, 510x258
>>2854027
This is low-tier bait without any effort put into it, but I'm gonna bite. Maybe I can educate some kid. If not, nevermind.

Please compare these textures and see how the limited 16bit low-res textures of the snes are melted together underneath the crt shader and also in real CRTs, creating colour gradients that were not possible with low-res 16bit textures. The result is of course very blurry. The scan lines add sharpness and definition again to the image and hide the flaws, creating an image very similar to the sharpness of the original (just a bit less sharper and bit less bright than the original - a real CRT of course doesn't do this minial degrading!)
>>
>>2854213
>are artist representations
That's exactly what I'm talking about. When the artists were told to make art based on the games, this is what Nintendo commissioned them to do. They asked for a bunch of pixel art, because that's how they wanted the games represented. It doesn't matter that they weren't gamers, they were artists commissioned to do pieces that their client directed. That's how commercial art works.

Point is, when they commissioned them, they didn't ask for scanlines. Or maybe they did once in that case, but it was rare.
>>
>>2854221
Raw is best, the others are disgustingly blurred, which we have already established was not intended. And all three are the wrong aspect ratio.
>>
>>2854216
>>2854219
But that's not how you look at a gameboy. No one looks at the screen with a macro camera. I have a real gameboy here and when you look at it the dot matrix is still visible, but not like that. It's kind of autistic simulating a gameboy on your TV with the perspective of a macro camera instead of a normal distance perspective like you would experience when playing the real hardware. Technically correct is not always the most authentic.
>>
>>2854221
>The scan lines add sharpness and definition again to the image and hide the flaws,

They absolutely do not. There is no actual detail added at all. This isn't bait in any way at all.
>>
>>2854215
>That's what VGA is.

No, it's also a video standard which relies on five connectors, namely Red, Green, Blue, Horizontal Sync, Vertical Sync. These signals and their pulse intensity are directly connected to the CRT's electron gun so they are as clear as they can possibly be, provided the connector does good conducting.

http://www.kramerelectronics.com/academy/?keyword=HorizontalAndVerticalSyncPulses

>Because focusing on line doubling make it very clear why mode 13h had sharp pixels.

No, it only makes clear why scan lines were thinner and pictures were more compact.

>So you think 640x400 non-line-doubled pixels on a VGA CRT are just as sharp as 320x200 line-doubled (logical) pixels on a VGA CRT?

Yes I do and it's not something wild for everybody who lived the era and saw how stuff worked.

>Anybody with working eyes can see this is false.

Just so I know, what is 640x400 going to do to my precious pixels?
>>
>>2854223
you do realize the time it would take to make pixel art with scanlines back then?
>>
File: example_02.png (180 KB, 600x634) Image search: [Google]
example_02.png
180 KB, 600x634
>>2854027
>>2854221
Example number 2. Now tell me, which one looks sharpest and most high-res to you?
>>
File: dragonfly in a meadow_ss.jpg (376 KB, 800x619) Image search: [Google]
dragonfly in a meadow_ss.jpg
376 KB, 800x619
>>2854230
I'm an artist, I'm well aware of how much time various types of paintings take. As shown here >>2854162 it's possible. And realistically wouldn't have taken a lot longer. Certainly if that was what Nindendo had wanted it would have been done.
>>
>>2854231
Raw, without question. Scanlines add noise, not detail.
>>
>>2854228
You realize the electron beams are made with real analog components? They're not some mathematical abstractions, they have non-zero size and imperfect focus. The mask is likewise a real physical object. The imperfections of the hardware cause blur, and this blur is consistent regardless of whether the scan lines represent one pixel or half a pixel. If the pixels are bigger compared to the hardware displaying them then the pixels are sharper.
>>
>>2854227
They do. It's in their nature. They make the blurry edges edgy again. That is how the CRT shader works. On a real CRT there are blank lines and they don't hide anything. Blur is depent on source and TV set, but the effect is similar nevertheless. Please do understand that people don't just add scan lines, even when emulating for just the sake of nostalgia. There is actually technical finesse involved.
>>
>>2854221
>how the limited 16bit low-res textures of the snes

You're funny.
>>
>>2854235
I always wanted to meet a live troll. You are funny, I like you.
>>
>>2854241
He actually believes it.
>>
>>2854241
If you genuinely think the image on the far right has more detail than the raw one I can't help you. You're simply wrong though. Adding a layer of noise on top of an image doesn't make it more detailed. Detail has to relate to something, otherwise it's just noise.

I can't explain this to you any more simply. If you don't understand such a basic concept then there's no point in continuing in any way.
>>
File: pixel size and blur.png (3 KB, 196x84) Image search: [Google]
pixel size and blur.png
3 KB, 196x84
>>2854228
Line doubling == bigger pixels == less blur per pixel == sharper pixels
>>
File: dusterror-arena-WEB.jpg (583 KB, 965x963) Image search: [Google]
dusterror-arena-WEB.jpg
583 KB, 965x963
>>2854234
I really doubt nintendo owuld commision specifically "No Scanlines" art. but anyway im an artist too, tough my style is different
>>
File: gameboy.jpg (231 KB, 1600x1200) Image search: [Google]
gameboy.jpg
231 KB, 1600x1200
>>2854205
>It looks incorrect to me.
Well it's a good thing no one gives a shit what some retard thinks.

>to know better than everybody else and sound like a total douche.
Too bad, stop being a douche yourself.

>>2854216
Pic related
>>
File: 1450110064115_blurred.jpg (176 KB, 965x963) Image search: [Google]
1450110064115_blurred.jpg
176 KB, 965x963
>>2854253
Stippling is the low-tech version of dithering. But nothing would claim attached pic is artists intention. The stipples are meant to be seen, just like the individual pixels were meant to be seen.
>>
>>2854257
Oh, boy. Now having a different opinion is considered being a douche.
Look how he went marco with the camera in the reflection. I bet that is how you play a game boy in real life.
>That close
No.

>>> >>2854225
> It's kind of autistic simulating a gameboy on your TV with the perspective of a macro camera instead of a normal distance perspective like you would experience when playing the real hardware. Technically correct is not always the most authentic.
>>
File: Capture.png (1 MB, 1142x874) Image search: [Google]
Capture.png
1 MB, 1142x874
>>2854257
Different gameboys have different screens though, most of them look similar but vary in tint. GB color looks more tan and black, the original had variations of darker green or golds.
>>
>>2854253
What I'm saying is that Nintendo commissioned a bunch of art, when a client does that they explain to the artist what kind of an image they want. Nintendo very clearly asked for the art showing their games and characters to have the blocky look of raw pixels as opposed to something that replicates the way the games look on a CRT. That's fundamentally what this is all about.

Nice work btw
>>
File: gameboy color.jpg (684 KB, 1934x2579) Image search: [Google]
gameboy color.jpg
684 KB, 1934x2579
>>2854271
> I bet that is how you play a game boy in real life.
No, I put mine in it's dock and sit on the couch ten feet away and squint at it like every retro gamer did, which I know because I'm gigantic flaming hipster whose qualified to tell everyone younger than 18 what it was like back when I wasn't even born yet.
>>
>>2854257
You are also totally ignoring the main argument
>>> >>2854205
> coming to the conclusion that dot matrix or crt/scanlines shades don't add anything to the experience of the gameboy and where never meant to be. It is just nostalgia and the technical limitations of that time, that led to the display having a dot matrix and the screen a green tint.
The desired effect doesn't add anything to the gameboy experience but nostalgia. Nothing wrong with that. It's totally okay. I'm just saying that CRT+scan line shaders actually add definition to SNES and to some extent to NES games unlike the case of the gameboy.
>>
>>2854279
GBC (and GBA) have very ugly interlaced screens, so trying to reproduce that look is a bad idea.
>>
File: retroarch 2015-02-24 20-20-02-03.jpg (406 KB, 1600x1200) Image search: [Google]
retroarch 2015-02-24 20-20-02-03.jpg
406 KB, 1600x1200
>>2854280
>>
>>2854289
>the butthurt is strong with this one
:P
chill dude.
>>
>>2854292
Shit guy, relax you don't have to get all frustrated about shit.
>>
>>2854280
>I'm just saying that CRT+scan line shaders actually add definition to SNES and to some extent to NES games unlike the case of the gameboy.

YOu're really just proving you have no point and that the only reason you think SNES looks better with scanlines is because of nostalgia.
>>
>>2854297
I didn't even own a snes lmao. I'm a millenial.My first device was a Gameboy. But my first console was a Nintendo 64. But I playes snes ocassionaly at my friends place as a child, but I don't remember exactly how it looked, yet alone scan lines. Yet this is my preference. No reason to be mad about it. I like it that way you like it the other way. No reason to get angry, it's okay to have differentiating opinions. I don't have any hard feelings.
>>
>>2854237
>The imperfections of the hardware cause blur, and this blur is consistent regardless of whether the scan lines represent one pixel or half a pixel.

So you're basically saying it causes blur just because it does. While it's true that a higher pulse intervals do lead to less smooth phosphor variation, verticality (your "line doubling" argument) has nothing to do with it because the rasterized lines don't talk to each other.

And more over, this has JACK SHIT to do with how dithering is blurred. Never, in any way, with signals competent such as VGA, can 15KHz 320x200 be blurred.

>>2854238
>They do. It's in their nature

They are only there so that the electron gun can fold back to the other side of the screen to draw the next (horizontal) line.
>>
File: V.png (548 KB, 616x540) Image search: [Google]
V.png
548 KB, 616x540
I know a lot of people will get angry because of this, but this is how I play and always have played Gameboy games - from approximately this viewing distance.
>>
File: V2.jpg (729 KB, 1632x1410) Image search: [Google]
V2.jpg
729 KB, 1632x1410
>>2854320
... and not like this! (I'm not myopic!)
>>
File: Regen 2015-12-14 19-30-20-503.png (638 KB, 1280x480) Image search: [Google]
Regen 2015-12-14 19-30-20-503.png
638 KB, 1280x480
>>
>>2854310
>verticality (your "line doubling" argument) has nothing to do with it
It does, because it contributes to the pixels being bigger, which makes them sharper because the blurring does not change.

>Never, in any way, with signals competent such as VGA, can 15KHz 320x200 be blurred.
Exactly, but with 640x400 there is more blur per pixel.
>>
>>2854327
Left is wrong, just like >>2854264 is wrong. The dithering is supposed to be visible. You don't throw away all that detail just to hide dithering (which in any case cannot be automatically distinguished from intended non-dithering detail, eg. CT typewriter).
>>
>>2854327

NTSC filters are my favourites
>>
>>2854336
Do you have any proof that CT typewriter blends into solid color like intended dithering areas on real console and tv or using emulator shader simulating that?
Where exactly do you see losing detail?
>>
>>2854348
>Where exactly do you see losing detail?
Any blurring filter with limited bit-depth loses detail.
>>
>>2854328
>which makes them sharper because the blurring does not change.

Jargon.

>Exactly, but with 640x400 there is more blur per pixel.

So you're arguing against your own point now?

>>2854336
Left is right when taking the composite signal from a Genesis, while the right side is also correct when using a basically uncompressed signal with little to no low pass.
>>
>>2854352
>So you're arguing against your own point now?
Blur is roughly constant per unit screen area. Bigger pixels means less blur per pixel. Smaller pixels mean more blur per pixel. Line doubling means pixels 4 times as big as the smallest pixels the display is designed for, so of course they will be sharp. Which is what mode 13h games looked like.
>>
>>2854358
I'm done, since you're still adamant on using "pixels" when talking about analog interfaces. I'm taking your post on the CRT general so people more knowledgeable than us can give us a straight answer. Please forgive me but I'd also like to know.
>>
File: capture.jpg (65 KB, 533x800) Image search: [Google]
capture.jpg
65 KB, 533x800
>>2854320
Oh wow, that's super close. You should see an ophthalmologist. This is how far you should be from the screen. The added monitor is useful for friends watching from their houses.
>>
>>2854323
Eh? That's the same distance, just the top one is what blind people see.
>>
>>2854364
I'm pretty sure he's just talking about upscaling, not CRTs necessarily.

If something is being upscaled badly so it's blurry then adding scanlines can bring a bit of definition back. But optimally it should be upscaled so there is no blur.
>>
>>2854364
"Pixel" means the smallest element of the image that is intended to be updated independently. Analog systems can have pixels. Mode 13h has bigger pixels than mode 100h on the same CRT, therefore those pixels will be sharper. See >>2854249
for demonstration of the concept.

>>2854378
No, I'm talking about CRTs specifically. Mode 13h, the popular DOS game mode, had pixels much bigger than the minimum size pixels the CRT could display. This means they were always sharp.
>>
>>2854381
I didn't even reply to >>2854249 because I failed to see its point. One text was upscaled using nearest neighbour, whilst the other with a bilinear filter. What exactly are you proving?
>>
>>2854383
Look at the text box here and see the effect of scan lines in action >>> >>2854231
>>
>>2854387
Everything in that image but the left side is phoney. Those aren't real scan lines, nor does the center pic serve any purpose other than to show how a shader renders stuff internally.
>>
>>2854383
There is no bilinear filtering. The right side is the left side with gaussian blur applied. Note how the bigger version looks less blurred, despite having the exact same blur applied. This is because there is less blur per pixel.
>>
>>2854387
That image doesn't mean anything. The left side is how it should look. The middle is upscaled to be blurry. The right side is meant to imitate a CRT but looks awful.
>>
>>2854284
>GBC (and GBA)
> interlaced screens
What the fuck are you talking about?
>>
>>2854432
They are interlaced. You can clearly see the combing artifacts.
>>
>>2854446
That's not interlacing.
>>
>>2854474
Yes it is. They update odd and even lines separately. Interlacing is not exclusive to CRTs.
>>
>>2852451
> I don't claim this to be the right way to play it.

After four paragraphs you finally said something I agree with.
>>
>>2854495
Got any proof? I'm sure people would love interlace LCDs, then we wouldn't have to deinterlace everything. Except they don't exist, its just an artifact of the way those things display.
>>
>>2854336

>The dithering is supposed to be visible.

So you're saying that the developers who used dithering were too dumb to understand that it would be blended by a standard consumer tv?

How did they expect people to see the dithering if it would get blended by most people's tvs?
>>
>>2854390
That's exactly the purpose of the middle picture that was intended. You got it right! Congrats
>>
>>2854503
>its just an artifact of the way those things display
>interlacing is just an artifact of interlacing
Look at one yourself, it's obvious. And you can do interlacing in software if you really want to, there are shaders for this. But it doesn't look identical to CRT interlacing because of the high persistence.

>>2854526
Developers, like everybody else who cared about quality, had RGB monitors. RGB SCART was very common in Europe. And even on composite connections the blurring wasn't nearly as bad as people set it in filters.
>>
>>2854529
Who cares when both of them look worse than the plain image anyways? Sure blurring made it look like shit but the scanlines still don't make up for it.
>>
>>2854534

>Developers, like everybody else who cared about quality, had RGB monitors.

But they were designing it for people who had composite. They knew that most of the people who would play their games would not see the dithering.
>>
>>2854573

I don't play SuFami so I do not care
>>
>>2854573
Are you retarded? The image you're cycling in and out there is a drawing based off of the in game graphics. The original was an Amano drawing.
>>
>>2854534
It doesn't make the display any less progressive even if it looks like interlace.
>>
>>2853715
Dude, I use a CRT every day. I am well versed with their quirks and shortcomings. Colors do not look like that unless you are an idiot that can't even into basic calibration, that curve filter does not accurately portray what it is like to look at a curved tube, and the attempt to simulate flare is just pathetic.
>>
>>2854638
I disagree with them, but I can at least theoretically understand why some people would want to recreate scanlines. But I can't imagine wanting to recreate CRT glare like that.
>>
File: IMG_0101.jpg (2 MB, 2448x1836) Image search: [Google]
IMG_0101.jpg
2 MB, 2448x1836
Another CRTfag here.

The thing that kills me about most filters is that they don't look like CRTs. They look like photographs of CRTs.

I'm not against using some sort of filter on emulator to prevent chunky pixels, but imho the shaders tend to oversaturate color and add too much glow.
Thread replies: 255
Thread images: 120

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

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.
If a post contains personal/copyrighted/illegal content you can contact me at [email protected] with that post and thread number and it will be removed as soon as possible.
DMCA Content Takedown via dmca.com
All images are hosted on imgur.com, send takedown notices to them.
This is a 4chan archive - all of the content originated from them. If you need IP information for a Poster - you need to contact them. This website shows only archived content.