/vr/ - Retro Games

>>6384518
I think there's a general confusion when people compare certain multiplatform games like Mega Man Legends and think that bilinear filtering must be really crap because it's made the textures look so undefined compared to the PS1 original.

But it actually has nothing to do with the filtering. Mega Man 64 uses textures which are 1/4 the size of the PS1 version. So when you consider that, the filtering is actually doing a pretty remarkable job.

While filtering is better in most cases, it doesn't go well with pixel art at all. That much is true.

>>6384518
>>6384530
Because what is the point when a good CRT shade does a better job and looks nicer, otherwise just play on a real one

Fat crispy pixel > blurshit

>>6384571
Texture filtering has nothing to do with that "CRT effect".

>>6384518
>Why would anyone not prefer bilinear interpolation in 3D games over point sampling?
It has been scientifically shown numerous times that point sampling is exclusively for queers.

>>6384530
Yeah, filtering gives a nice softness to smaller textures, to a point. But the N64 textures were *so* tiny, there’s barely any information to work with and it ends up blurry. It’s still better than without the filtering mind you, if you’ve seen until there’s N64 texture screenshots it’s pretty abhorrent.

Because the simple minded hipster drones that are everywhere in the "retro" community nowadays see sharp squares = pixels = omg so retro.
The N64 look doesn't fit the modern narrative of what is "retro", as decided by youtube ecelebs, so it gets shat upon.

>>6384579
Except it does, because you're having dot-crawl and other CRT artifacts ontop of whatever texture filtering you're doing. Just looking at that example the linear filtering has way less contrast than the nearest so it likely impacts how the CRT draws them. The N64 just had such low res textures that a CRT alone isn't enough to pretty it up (pic related). Maybe someone can make a comparison of how CRT's handle the different textures for a more scientific example.

>>6384759
Yes.

>>6384579
Indeed it does, you would know if you had one, non RGB of course

>Playstation games benefit greatly from the added filtering on the PS2
Do they?
https://www.ign.com/articles/2000/10/28/fun-with-backwards-compatibility-2

>>6384865
>Vagrant story with texture filtering
Absolutely disgusting.

>>6384530
there is really no need to filter pixel art. textures should be filtered because after geometric transformation (translation, scaling, rotation, shear) there is no guarantee that screen pixels align with texture pixels. however pixel art pixels are pretty much always aligned to screen pixels since the only transformations you apply are translations in screen space.

>>6384798
considering that interpolation is a low pass filter, i'd guess that dot crawl would be less of an issue for interpolated textures than not, since usually dot crawl is caused by high spatial frequency chroma signals aliasing to luma

>>6384798
>Dot crawl
*laughs in RGB*

>>6384518
I remember when any sort of texture filtering was considered incredibly desirable and we all wanted it, around 1996 when the second batch of 3D accelerator cards for PCs came out after the first offerings in 1995.

>>6384574
This all the way.

>>6384998
If I don't want dot crawl I have a PVM. For a lot of non-PC retro 3D games I prefer to have dot crawling. It does a lot to hide the shortcomings of early 3d.

>>6385018
Now look at everything that’s not the characters

>>6385151
I love the way the whole game looks. I think the blurry version in OP's pic looks like shit and the reason so many N64 games looked terrible.

>>6384518
The issue isn't filtering itself. It's the combination of Filtering + Extremely low resolution and low quality textures + crude AA that blurs everything that makes some people prefer the look of the PS1 and Saturn over the look of most N64 games.

>>6385427
Linear is not as much blurry as nearest is overly sharp.
There is no loss of information, pixel colors are still represented faithfully, it's just that there is now a smooth gradient between them.
For what it's worth, the only true representation of pixels are infinitely small points, and that's not really useful.

>>6385445
I'm not saying you have to agree, but to me there's nothing I consider overly sharp. I also hate CRTs though and display all my games with sharp pixels.

I'll certainly agree this looks bad >>6384798 but that's because it was never meant to be seen that way and is an ugly game already.

>>6384986
Thanks for the enlightening information that no one understands, Dr. Pixel.

>>6385490
https://www.youtube.com/watch?v=InrDRGTPqnE

>>6384518
>It was without a doubt one of N64's biggest boons
It hurt the N64 more than helped it, but the whole system is a mess of premature features on underperforming hardware so it ended up needing it.

It looks worse with low res textures but better with high res

>>6385469
based asha-poster

>>6384798
I feel like for n64 the artists here definitely planned around the fact that the textures would be filtered. In a lot of n64 games the textures feel like they are there to convey color and lighting rather than, well, texture

>>6385018
Blur sucks, but non-integer nearest is worse.

>>6385445
>the only true representation of pixels are infinitely small points
This is false in the case of pixel art.

>>6384865
It does look pretty good. You don't lose any detail, it just smooths over sharp edges. Looks like a Dreamcast port of a Playstation game.

>>6385630
Looks like ass with major losses of detail

>>6385639
What detail was lost?

>>6385640
put on some glasses

>>6385648
So no detail was lost and you're just a pixel lover.

>>6385651
Put on some glasses and stop talking like a woman, faggot

>>6385653
Why keep replying if you have no argument? What detail is lost in that comparison?

>>6385639
>>6385640
I agree there's no real loss of detail but the unfiltered one looks way way better.

>>6385639
To be fair, Quake's 8-bit textures are a poor match for filtering since there is not sufficient information to create smooth gradients.
I wouldn't say that detail is lost, but the discontinuities between various shades are a lot more evident.

>>6384518
i generally prefer ps1's larger, sharper textures over n64's blurry blobs

also, the n64 did trilinear texture filtering, not bilinear

>>6385678
>there is not sufficient information to create smooth gradients.
Yeah, texture filters are better for textures with more colors and higher resolutions.

>>6385630
i get where you're coming from, but i don't think the right is necessarily a net positive
there's much less aliasing, sure, and smooth gradients make sense smoothed out, but at the same time, you also lose sharp edges, making everything look a bit out of focus, which contrasts with polygon edges, which i think looks weird, like everything is coated with vaseline. this effect is only really a problem with low resolution textures... which the n64 is even worse at for various reasons

>>6384518
Look at how you can clearly see the the bricks pop out from the mortar in the nearest filtered texture. How you can see darkness between them. It looks like literal shit soup in the filtered version. Low detail textures generally benefit from nearest, unless you're smartly using a blur over an organic surface or fabric.

>>6385630
>just fuck my contrast and illusion of depth up, senpai
With low res, 5th gen textures there's no real winning but there just isn't enough resolution there to smear it over and not have it look undefined shit.

Bilinear + low res makes everything look someone sharpied in details on cardboard.

>>6385721
This.

>Not using bicubic

>>6385820
Wow that looks like ass.

>>6384518
it works well with low-res textures
>>6384530
>pixel art
no one asked. and 2XSuperSai was almost perfect for most PC monitor pixels. not so much for consolefaggy tv games.

>>6384574
Not the way it was meant to be prayed. PC monitors had crisp pixels. Console fags got tv scanlines and composite dithering and defects.

It really depends on the game, I think. For example, I like Mario Maker's way of displaying retro styles. But I also like to play SMB3 and SMW on CRTs.

>>6384518
Because it looks smeared. It looks wet and oily. It's frankly fucking disgusting looking.

>>6384518
N64 graphics always remind me of this comic

>>6384518
I use nearest on Quake 2, Half-Life 1 and Counter-Strike 1.6. At a distance, it almost looks HD compared to bilinear filtering, which is horrible and whenever possible I disable it in retro games.

>>6384798
>dot-crawl
>somehow this is CRT's fault
Composite video and RF have dotcrawl, you can see it in modern displays too.

>>6385860
Love this backwards thinking. It was meant to be played crisp but the fucking horrible limitations of the time meant we got a blurry fucking mess.
But keep rationalizing your horribly ugly fucking filters because you kids cry every time you see an actual pixel.

>>6385860
It was meant to be played on whatever you had available, and if you had a TOP OF THE LINE device you would have enjoyed its crispy benefits.

>>6384518
Soul vs soulless

>>6386528
Console games were all designed with composite dithering in mind. Devs of that time knew most TV sets would look like that. Dumbass.

Most copypastaed SANIC example.
https://youtu.be/x0weL5XDpPs

>>6386956
>Console games were all designed with composite dithering in mind.
that's like saying all artists are masters at their craft

>>6385490
>interpolation
connecting the dots
>is a low pass filter,
blurs things out,
>dot crawl
some goofy tv behavior
>would be less of an issue
wouldn't happen
>for interpolated textures,
for blurred pictures,
>since dot crawl is caused by high spatial frequency chroma signals
since goofy tv behavior happens when color changes fast
>aliasing to luma
and gets misinterpreted as brightness

>>6386956
oof

>>6384518
>bilinear
>N64's biggest boons
N64 didn't use bilinear but 3point filtering, a shittier approximation.

everyone in this thread is retard

bilinear filtering is intended for CRT monitors, because they're already really sharp

LCD monitors in comparison are rather soft unless its like an IPS OLED or whatever

in my opinyun:
>using a CRT? enable bilinear or trilinear
>using an LCD? disable filtering or use a more advanced version of texture filter

simple as that
this thread is now closed

>>6387029
3 point bilinear filtering is a legitimate variation on bilinear filtering, it still applies the bilinear algorithm but based on triangulation rather than a quad grid.

It’s not blurrier than regular bilinear filtering. It’s actually slightly sharper, being closer to nearest neighbour because it is prone to generating diamond shaped aliasing in textures

>>6386956
Based.

RGB niggers eat shit.

Composite master race coming through.

>>6387082
>>6387087
>>6387070

>>6387070
Okay, dunning-kruger dope, the surface of a texture is still massively affected by the filtering applied to it even up close, where the resolution/pixel density/mask of the display doesn't matter.

>>6387087
>coming through

More like flickering through. Muh unintended transparencies do not excuse everything else also being a blotchy blur.

Trying to use dithering as argument for blurcables is pretty bloody stupid considering that PC games extensively used dithering as well. Furthermore, this whole composite nonsense has been brought up only in recent years, if it was actually intended, the notion would have been around for far longer. Good luck digging up ANY of this shit before 2010.

>>6387169

>>6387159
>Muh unintended transparencies do not excuse everything else also being a blotchy blur.

they don't look like that on CRT. and CRT/Composite filters look fine on Modern LCDs unless you fake-accuracy fag..

>>6387175
Source?

>>6387183
Earthworm Jim Making Of book.

>>6387169
>Good luck digging up ANY of this shit before 2010.

SANIC was before 2010, duh. Many PSX games also made use of dithering with composite TVs in mind.

>>6385018
Undead Hero be jamming.

>>6387070
>LCD monitors in comparison are rather soft
In what context?
When upscaling, sure, but LCD's are the very opposite under native resolutions. In fact, the inherent softness of the CRT is a benefit to image quality that has been lost.

>>6387187
>Published December 2019 by Interplay Entertainment Corp. (first published 2019)

Not exactly before 2010.

looks fine to me

>>6387289
Yeah but it's talking about their old games lol

>>6387296
unnfff yeah now thats the stuff...big retro pixxxellllsss....just like my favorite youtubers told me to like....

Clearly it differs from game to game. Quake, for example, looks better in "Chunky Mode" because Adrian Carmack is first and foremost a pixel artist. I mean just look at Doom! No one in their right mind would ever say Doom looks good filtered. And again like what >>6385678 said the lack of colors made the filtered textures look like shit anyway. Not to mention the software rendering, which was what the game was originally developed with, lacked filtering. They only put filtering in for glquake later.

However, other games didn't think of textures as sprites or pixel art from the 2D era and designed them to be filtered, like in >>6387296. Again it depends on what the artists intended for each game.

But whatever I guess it's more fun to join a team and argue.

>>6388589
You know this “low colors makes for bad filtering” thing is completely anecdotal and lacks any scientific basis right? There’s been plenty of research completed on 3D graphics and this is the first I’ve heard of it.

The classic 3D API manuals don’t mention anything like that in their guidance, even though they do mention situations where filtering and not filtering may be suitably used.

>>6388606
lol you don't need scientific research to look at Quake's filtering and think "wow this looks like shit"

>>6384518
I guess you can just literally be wrong if you want. Crisp pixels of textures always look better in older games. This is why Quake 2 looks best when you run it in hardware mode for the lighting, but disable texture filtering to give it the crisp look of software mode.

Uppity N64fags at it again.

>>6388609
It’s still anecdotal garbage. You might think they look bad due to th unfitting artstyle... or maybe it’s just your personal familiarity with the pixels making you think anything else looks off. Confusing nostalgia for quality, etc. I don’t know. But I can say most people talk about filtering without understanding what it does.

>>6387296
i'm sorry the n64 can't afford too many pixels

>>6388636
Not him but I agree. Most PC games look extra terrible with filters on.

what i want to know is how to force unfiltered textures in arbitrary pc games

>>6384518
>Why would anyone not prefer bilinear interpolation in 3D games over point sampling?
just look at your picture to find out why

>>6389337
Lol that's exactly what I've been thinking. I guess I prefer it because it doesn't look like a blurry mess. I wonder if eyesight factors in, it would be interesting to hear if people who like filters wear glasses. Maybe they're more used to blur in general so find it comforting? Probably not but it's a thought.

>>6388636
>>6388606
>Aesthetic preferences are not scientifically verifiable
Wow, we have our next Einstein here.

Pixel and low res art is cartionish in nature. As such you don't want to apply a photographic filter, you want something that blends similar colors and sharpens and connects edges.

There are several algorithms that do this and many are great.

>>6389403
I have terrible eyesight and blurriness is frustrating. I generally prefer unfiltered.

>>6389403
No glasses, filtered.
Just look at the transition from the glove to the skin.

>>6389470
Example https://en.m.wikipedia.org/wiki/Hqx
Overall the game was made at the resolution it was made.

>>6387025
>let me take a picture with a shit camera
do you honestly think games look that faded when you are looking at the screen?

Question: If old games are better off without texture filtering, why make an exception for the lighting?

>>6389840
because lighting is generally not sharp detail (except shadows, which back then were done with either their own blob textures, or polygons which were necessarily sharp)

>>6391057
Same could be said about many textures.

Although a case could be made about overblurring. Bilinear filtering on old hardware is technically wrong as seen here.

>>6384518
The textures have to be made with texture filtering (or without) in mind. N64 games need filtering. PS1, Quake and stuff like that should not. There are some other PC games that are debateable.

>>6389840
If the scale is fucking huge like N64 textures yeah it needs filtering.

>>6389840
how did you do it?
I've been trying to do that with thief
how can i get pixelated look in og deus ex with the opengl renderer?
I saw a screen shot of unreal looking all pixelated as well

>only static images itt
linear looks better in slight motion
nearest has a shimmer on the pixels when slowly looking around

>>6384518
I don't like my video games looking like I took my glasses off.

>>6392467
The filtered version just looks better in every way. Why are there people that want the image to be as noisy as possible?

>>6392495
It's really game dependent, but resolution is another factor. Playing Quake 1 in 16:9 1080p or higher, the nearest textures bring out the detail in texturework and enemy graphics. Playing in low resolutions, linear keeps the visuals smoother and denoises the image. I for one usually prefer keeping things unfiltered, but you get some rare exceptions where things are too messy, and in my emulation I just try to keep things pixel accurate or the N64 3-point filtering on.

>>6392495
it is completely dependant on the scale and intention of the textures

>>6384518
left looks awful and anxiety-inducing, right is nice and relaxing

>>6392425
Not my screenshot, but I managed to do it in Quake and Unreal by hex editing. Look up the definition of the OpenGL nearest filter string, then search for its little endian value in the exe or dll.
It is quite insane how well bilinear masks the low resolutions.

>>6387029
>>6387082
Huh, the more you know.
I always thought something was off. Quake on the N64 looks especially awful.

>>6387082
Call me biased but I love how the slightly jagged textures look with 3point. It tends to give games a warmer, fuzzier look than either sharp pixels or complete vaseline.

The filtering, more like smearing, seems awful as it robs the picture of its crispness and colors.

>>6384518
Because at a certain point, it does make sense when the resolution of the textures are high enough and you have enough mipmaps that make it look good. My rule of thumb is that for retro games, there are almost no games that fit the above. 128x128 might look good in a racing game, depending on the person but 256 x 256 is around the spot where that shift happens but very few games hit that target. Legacy of Kain Soul Reaver is one exception for me where nearest sucks compared to bilinear and above filtering. Of course, you would play the Dreamcast version. Shame the PC port hasn't been fixed but it is what it is.

>>6384518
Linear filtered always looked like blurry garbage to me. I actually ran quake 2 and half life in software just to get them to look better and sadly it was a trade off between shit performance and some lost effects vs looking at garbage textures. I'm not sure if its one of those "mind fills in the details" but the blocky textures always looked better and more detailed to me than the vomit soup.

>>6393104
>Smearing
There goes that buzzword again.

>>6385630
Right looks better

>>6393804
>buzzword
There goes that cope again.

>>6393804
I wouldn't call "smearing" a buzzword as it accurately describes the effect he is referring to.

Unreal software mode filter was the best, sadly it was never used after.

>>6396634
>>6396681
Smearing is loaded with implications of quality reduction. "Robbing a picture of crispness and color" is not a very useful description academically. At the very least call it a blur, which is colloquial for low-pass filtering.
Both forms of interpolation, bilinear and nearest, are technically speaking filters with various trade-offs. The difference lies in the area in between samples that isn't explicitly defined, and thus it must be reconstructed with said filters. None of the filters are categorically better, it all depends on what one is attempting to reconstruct.

But why should I expect more than subjectivistic rambling on a board like this, a board fueled by nostalgia that had best not be perturbed for the sake of mental health?

>>6398708
You would be right about just filtering alone but mipmapping solves the in between sampling problem especially when combined with bilinear filtering to give us common trilinear filtering. That combined with ansiotropic filtering basically renders your tradeoff obsolete at the cost of lots of GPU power, which we have today.

In a retro situation, it's just better to use nearest because the tradeoff there is much easier to justify and you might as well use the extra resources required from bilinear for other things.

>>6398896
Mipmapping is a benefit to the performance if anything due to better cache utilization.
Hardware capable of bilinear filtering usually does it for free anyway since pixels are drawn in quads.
Either way, the argument about texture interpolation has little to do with mipmapping, since anisotropic filtering is orthogonal to it. You can enjoy crisp and stable textures at steep angles without blurring them close up.

>>6396851
This is a cool effect, what other games have this?

>>6398984
does any other game use dithering?
you are on /vr/ you know

>>6399134
software rendering with dithering to fake the texture smoothing smart ass

>>6399264
it's called the PS1

>>6393165
I was waiting for this game to come up. It's not simply a problem of one or the other, but how the games were made. Soul Reaver is a standout example. Compared to the DC version, the PS1 version looks like shit. Back in the day, I didn't know there was a PS1 (or PC) version, so I only had the DC game. Now I have all three (it's one of my favorite series/ entries in the series). Playing the PS1 version was like hell on the eyes. Looked good for what it was, but a big no thank you. Despite the upgrades the DC port got, I really do with they'd have made a proper DC game out of it. Same for Nightmare Creatures II. A game that is plenty fun, but suffers on the DC port from being almost entirely unupgraded from the PS1 original.

>>6399271
doesn't look the same

>>6399464
looks similar to me

>>6384518
I don't see any difference?

filtered looks objectively better

unfiltered looks more "video game-y"

/thread

I'm seriously considering making a "bounty" for an N64 clone system that completely disables any anti aliasing or blur, or at least makes it a toggle
I'm thinking of putting up 15k for it but I need to think about it more

>>6399585
do you mean like the hdmi mod?

>>6399603
The HDMI doesn't do it good enough for me, it doesn't totally disable blur and only kinda fakes disabling the horizontal blur

>>6399610
I think we can all agree that every one of those images looks like shit.

>>6399615
You can't just talk about Obama like that

>>6399610
I mean the one on the right looks exactly like it, just zoomed in

>>6399610
yes it does

>>6398984
none
>>6399134
>>6399271
>>6399557
It's a texture filter, dipshit. Not a fullscreen dither.

>>6396851
>>6400447
I think what this is, is basically a software implementation of what was used as an alternative to trilinear filtering on the original 3dfx Voodoo.

Basically if I recall, the original Voodoo needed two cycles to execute trilinear filtering because the linear filtering hardware had to be run twice and it only contained one unit in the pixel pipeline capable of linear filtering, which was kind of a bummer.

To implement a second linear filtering unit would have eaten up too many transistors. So what 3dfx did is they also implemented a simple texture coordinate dither unit instead which could run in the pixel pipeline after the linear filtering unit in the same cycle. That way the Voodoo could bilinear filter textures, and dither coordinate filter mipmaps in the one cycle. Way less hardware extensive than a real bilinear filter, but acceptable results.

I wonder if Epic got the idea by reading 3dfx patents?

>>6386956
>Console games were all designed with composite in mind
Technically, this also is true for early PC games, as they used composite artifacts to generate more color compared to the awful CGA palette.

>>6399884
Yes but you need to enter a GameShark code every time

>>6400658
>generate more color compared to the awful CGA palette.
few people even realised some games used it.
https://youtu.be/rSJzYwPhKhU

>>6400901
just use an everdrive with an already downloaded set of cheats for that

>>6399557
Idk what this conversation is about but the right looks way better.

>>6402064
why do you think it looks better, do you even know what the purpose of dithering is?

>>6402078
to sell games by making them look nicer

>>6402078
not him, but i can understand adding a bit of noise can make things look subjectively nicer, especially in blurry/low-detail scenarios, since it gives the impression of detail where there is none, it can make your brain see more than there really is

pic has some noise added to the left side

>>6402119
I mean no offence by this but god damn that is a dumb opinion
there is a reason that CRT's and CRT shaders are used for dithered images

>>6402126
there's also a reason why some people opt to use noise-adding shaders in video players

>>6402132
why would you add noise filters to videos?
go watch standard def videos on CRT if you think the quality is bad, adding noise seems completely pointless and film grain does not make anything look good

>>6402078
Dithering is used to trick the brain to see more colors. In that instance it also adds more depth to the image. The one on the left looks like a blurry mess compared to the one on the right.

>>6402119
They kinda look the same now, although there are a few cases where I still prefer the right. Particularly on the giant gem. The left looks better in the darker areas that are one solid color like the inside of the arch and the window-ish thing.

>>6402126
He's right though. It'd also explain why I prefer to use my shitty consumer RF TV for a lot of retro games over my PVM for certain games. It emulates dithering.
>there is a reason that CRT's and CRT shaders are used for dithered images
Because most electronic devices could show enough colors that they didn't need to dither around the time CRTs started being phased out?

>>6402154
>Dithering is used to trick the brain
what the fuck are you talking about dithering is used for blurred signal, it has nothing to do with your brain

>>6402180
>Didn't grow up when dithering was used on everything
>Doesn't understand basic art 101
Someone else can try educating this child.

>>6402191
>Didn't grow up when dithering was used on everything
>Doesn't understand basic art 101
maybe you should educate yourself on why you think middle looks better than the right

>>6402180
>>6402191
It's both.
https://en.wikipedia.org/wiki/Divisionism
https://en.wikipedia.org/wiki/Pointillism

Many CRTs would blur images a bit, but others didn't really.
But that's irrelevant because it's also the basic fucking operating principle of a CRT, and you're both retarded.

>>6402154
>No noise = blurry mess
I feel sorry for you and your damaged brain.

>>6384518
because when I see blur I want puke
literally
when I was a kid and played those blurry games I went to toilet and puked and I never played them again

ya cause I'm sure the dithering is meant to be visible and not smoothly blended, amirite?

>>6403149
Technically, a dithering filter on top does decrease the proportion of blur.
Granted, it's a very artificial high-frequency spike, so it does not actually sharpen the overall image.

>>6389493
Unfitered is still better.

>>6384518
I did when the technique was new, but now I don't like it that much.

>>6384518
That texture looks pretty shitty and smudgy with the filter.
N64 games were generally designed around it, so it can look pretty odd with it turned off, but generally I think almost all PC 3D games look better with it off.

>>6384759
I'm indifferent to retro or not, I just like the crisp and clear graphics (I also had shitty TVs as a kid, so I don't have any nostalgia for those CRTs). People say it ruins the effect of dithering, but I think it conveys anyway.

>>6385639
Quake's low res texturing is too beautifully done to be smeared out.

>>6385704
You're right, the filtered one has less depth and just looks flatter.

>>6386519
That's pretty good.

>>6386524
Wasn't the filtering in GoldSrc games like faulty or done really poorly?

I guarantee you that no kid who owned a N64 at the time had any idea what the term "bilinear filtering" even meant

>>6403641
And? I doubt any kid (read: under 13) knows what temporal anti-aliasing or anisotropic filtering is on their PS4 but it’s still important.

>>6402191
It’s like a bell curve curve of soul

>>6389403
I have poor eyesight and I hate blur filters, makes me feel like someone took off my glasses. I also REALLY hate chromatic aberration effects for the same reason.

>>6399585
Jesus I've seen this image and post like 30 times in the last week you have an obsession

>>6402119
>a bit of noise can make things look subjectively nicer, especially in blurry/low-detail scenarios, since it gives the impression of detail where there is none,

that's what VLC player does for muh animes and old movies.

>>6403149
>Doesn't know most consumer CRTs added some form of noise
>Doesn't know retro 3d relied heavily on this
>Thinks retro games are suppose to be played at 4k on a crystal clear LCD screen
Zoom zoom

>>6403441
Why are you try to explain retro basics to zoomers?

>>6384518
I'm honestly at a loss trying to comprehend how anyone could prefer that blurry brick texture to the crisp detail on the right.

>It was without a doubt one of N64's biggest boons

Except it wasn't a boon. N64 games were frequently derided as blurry. People called it 'the vaseline filter'. It was particularly awful for text.

Bilnear filtering on textures tends to look awful until your texture resolution is high enough - usually around 256x256 - or 128x128 in cases where the textures stretched across a wide area of the screen.

What I really wish we saw more of was pixel dithering - its the best of both worlds and almost nobody uses it.

>>6403882
>Y-yeah, there's supposed to be noise, it's totally not an artifact of suboptimal video signals and hardware
>Never mind the fact that dithering is still relevant to this day and that new innovative ways have been found to do it better
>Never mind the fact that CRT softness is a desired effect to this day emulated through terribly inaccurate gaussian blurs, and that it isn't merely a matter of MUH RETRO
>Z-zoom!
Sure thing, Mr. Dunning Kruger

>>6404121
>there's supposed to be noise, it's totally not an artifact of suboptimal video signals and hardware
Consumer CRTs weren't PVMs
>Never mind the fact that dithering is still relevant to this day and that new innovative ways have been found to do it better
I'm starting to think you're confusing 1 of my comments between 2 different pictures. I was calling the 32-bit color one the blurry mess and said the dithering one was better.
>CRT softness
Is that the new term for noise?

>>6404148
>Consumer CRTs weren't PVMs
Not TV's, but the whole PC market basically had PVM-tier screens on their desks, and they used dithering religiously.
>I was calling the 32-bit color one the blurry mess
Exactly, and that is a braindead opinion.
>Is that the new term for noise?
Okay, I am starting to realize why you think the way you do.
Noise and blur mean rather specific things formally, and they are not the same. Noise is characterized by disturbances in the frequency spectrum either concentrated in peaks or showing up randomly. Blur is characterized by a gradual attenuation of high frequencies in a signal.
CRT pixels are blurry by nature, and that's a good thing since it ameliorates the noise introduced by dithering (an undesired effect).
The 32-bit image is NOT blurry, and even if it were, it would arguably be preferable over dithering noise. The Sonic waterfall is a classic example.
If you really want the image to be more sharp, there is a filter dedicated to that, dithering is NOT the solution.

>>6404237
>Not TV's, but the whole PC market basically had PVM-tier screens on their desks, and they used dithering religiously.
This. I always felt dithering looked fine (good even) even when not smeared/fuzzed out, the effect still conveys, it's not like you're pressing your face up close and like you're looking at a checkerboard.

>>6404237
crt's don't have pixels

>>6405537
Computer generated imagery has pixels that are displayed on the CRT, smartass.

>>6405537
Bet you felt really smug as you typed that out.