because pros, intuitively, understand that anisotropic filtering means adding unnecessary visual clutter and making it harder to spot enemies when they are close to detailed walls/floors at an angle it's not because "pros don't know any better lmao placebo"

it"s true on one hand, but some pros just don't know any better and copy each other. zywoo was recently asked what his video settings were and his response was basically "idk, I just leave it as is, he never bothered to change them". it's from an ESL clip on YT

yes, I am not saying that the average pro can even tell you what anisotropic filtering does. neither can the commenters under that post. but their characterization that “anisotropic filtering makes things clearer” (wrong, it only makes TEXTURES AT AN ANGLE clearer, not mesh boundaries), therefore “clearer = better” (not if it causes high frequency features near an enemy making its edges harder to immediately notice…), therefore “pros are clueless” is what really triggers me. Pros should be used as a large scale evolutionary algorithm. Assume they all use settings at random. If a setting ends up prevailing, maybe before writing it off, at least analyzing whether it could have some counter intuitive benefits is IMO the way it should be done: if the majority of pros use it, it could be that it’s because it provides a nonzero advantage, basically making them more “fit” for the environment like a random mutation would make you survive more generations in this genetic algorithm analogy I am drawing. They don’t need to know WHY it’s advantageous. Whereas just because pro’s “don’t know why they use a given setting”, a ton of people go to the “therefore it’s placebo” conclusion WHICH IS NOT the logical conclusion

I fully agree with you

You're seeing stuff neither setting makes the CT harder/easier to spot

and if my word is insufficient, here’s a paper on nature that explains that indeed, similar to how a CV algorithm would struggle to do object detection if more high frequency features are visible, so do human brains so I might be “seeing stuff” but maybe I am seeing the right stuff

Can't believe I'm reading ts instead of studying gn bro

Damn you showed everyone that overanaylising a screenshot for a minute means you would notice the difference in the moment. If you're looking at stationary pixels and not the moving ones, you might just be stupid.

yes because of course the extra details on the walls would make it much easier to see the enemy if I came out running from this angle and the enemy was there. how are so many of you missing the point? it’s concerning, really

Then why not post the whole screenshot of your view instead of a hyper magnetized view where you can hyper focus on the 15 pixels that change when the entire CT model is unironically a 3 times larger than your 2 pixel by 2 pixel crosshair?

It doesn’t matter. All the pros use 800 DPI as well which isn’t the best choice. They also don’t use OLED either(which is better than TN)

sigh. this is the same issue with the people criticizing anisotropic filtering. you saying “800 dpi isn’t the best choice”, to me, just equates to a shallow understanding of the topic. a statement like that does not stand on its own. go look at my other recent tweets.

There are exceptions to my statement but speaking in generalities is valid. “1600 DPI is better” isn’t disproven by “well AcTUaLlY in this specific application it isn’t” higher polling rate is better for example BUT YES there are limits/exceptions and other variables that limit that statement.

tell me how 1600 dpi is better.

Assuming you have no hardware limitations. Lower input latency to being with. Better accuracy at higher resolutions. (Better pixel reporting) so for example 0.625 at 1600 DPI > 1.25 at 800 DPI. I’m not going to write an electrical eng. dissertation on why. It can be googled.

maybe actually read my post and understand why “better accuracy at higher resolution” is a baseless claim that shows, as I said at the start, a shallow understanding of the topic. Same thing with the overloaded “input latency” terminology without understanding what latency means in this context. Again, read the posts I linked

I did. You are just simply incorrect. There are more data points. More accuracy (assuming the sensor can handle it) You just keep telling me I don’t understand the terms I am using. Btw your noise statement doesn’t apply to digital inputs the way you’re using it. Mice aren’t analog but I digress.

trust me, you don’t understand. you did not read my actual post or at the very least you don’t understand it, or you don’t understand what the dsp does in a mouse. and if you think signal processing concepts don’t apply in this case you don’t really understand the engineering part either. The DSP runs an optical flow algorithm to reconstruct motion based on individual samples that are inherently a discretization of an analog signal (infrared light). You just don’t understand basically a massive part of computer science that deals with real world data once digitized if you say a vague ill defined sentence like “doesn’t apply to digital inputs”. I am just so annoyed because you seem actually well intentioned and you seem to have some level of engineering expertise but you seem to refuse to put your biases aside and actually understand the topic. Btw, here’s francois morier from logitech also using a lot of signal processing terminology, specifically also explaining the (obvious) fact that higher DPI increase the noise floor. Might have something to do with the fact that we’re talking about a SENSOR? And using signal processing terminology in comp sci is a widely common way to describe a lot of statistical properties in machine learning for example as they are really good ways to describe some phenomena there, even though ALL the inputs in this case are “digital” unlike a mouse where the “inputs” are photons passing through a lens. I just don’t understand why you don’t want to even spend 10 brain cycles to actually stop and think that maybe I am NOT clueless but I actually understand the topic, and read my posts?

funny that in the example you give its easier to see the player model when x16 is applied lmao

i can’t believe how some people are willing to literally ignore what is in front of their eyes if it helps their confirmation bias

This is mental illness lmao. 99.99999999% the same look at the 2 lines and 2 tiny blocks on the floor. For sure THATS what makes a difference.

It's harder to see a guy because of some lines on a wall???

yes.

Skill issue

am i crazy the only difference is how detailed that brick wall is

the wall and the floor, yes, you’re not crazy. my point is that in the absence of the extra detail, the enemy becomes easier to spot. ofc people are not being very bright (not you, but the other commenters also noticing the difference is tiny but framing it as proof of this being irrelevant) and being unable to take the very tiny logical leap that if you were to peek towards A site from main at speed here, you wouldn’t see the CT as obviously as in this zoomed in, stationary screenshot, but you’d have to quickly identify it in the full scenery, and fewer black edges from the bricks to its right make it easier to spot fast. But yes, you are indeed correct that that’s all anisotropic filtering does. It does not change the appearance of the enemy whatsoever, which is what people tend to miss. It does not change the mesh boundaries as rendered on screen (aka the silhouette of the 3d models). It just adds detail to textures at an angle which in some cases (like this) makes the enemy stand out a bit less.

look im just going to be nice by not saying anything lol

sorry bro im actually a dad of 13 who works 90 hours a week and i play cs2 40 minutes every 3 months. So here's my opinion about why i know better about the game than pros

its not harder to spot the enemy at all.

I can't see the guy when it's moved to the left, thank God they removed the detail of that wall to make the guy more visible.

completely missing the point

if you retards had your way every map would look like cp_orange and be only dev textured for visibility. you need to be put into labor camps, plain and simple

pros most certainly do not understand anything about anything for the most part it's just that bilinear is the lowest appearing option and they think it'll lead to the best performance

That's why the settings have to be standardized. The spirit of the competition is not about squeezing every minute advantage playing 4:3 resolution with the game looking like shit, it's about skill being the main factor all else being equal.

I mean, at this point might as well reduce Texture Quality as well. Since it will reduce the amount of detail too. (I am not saying I am against or in favor of doing these things. I haven't tried. It makes sense to me though.)

yep, and I imagine most of them do use “low” for that setting as well

Except it is placebo. If you can't see the enemy, you need better glasses.

clueless

Have you ever thought that YOU are the clueless one? If you had tried using your eyes, you'd see how much easier it is. You have no once but yourself to blame for your incompetence.

you're either ragebaiting or too dumb to be engaged with in conversation if you are unable to understand that in a competitive setting, being able to spot an enemy faster because of well-known perceptual phenomena in the way our brains decode visual information in the presence of more detailed (high-frequency) backgrounds, is a tangible advantage and, by definition, the opposite of "placebo". again: clueless

You know how to spot an enemy faster? Using your fucking eyes. Again, YOU are the clueless one. "VAL" tells me all I need to know about you.

ragebait, low quality

🥱💤💤💤

Your example is priceless why I don’t use 16x like all the posts the last months tells us too. “Just use 16x the performance degradation is negligible”… okay but maybe I don’t want every tiny rock and leaf on ancient to be fully sharp and in focus or even standing out 🤣😅🙄

It's funny how this example has the CT being so fkn painfully blatant against the background regardless of the sharpness of a couple oblique textures surrounding him that this will make 0 difference in-game unless you have bra1n dmg

Pro mostly have no idea how to setup their computer No need to explain how they optimize their game or windows Most of them don’t and just focus on the game with the vibe

pew pew? :o

I'm not going to argue that billinear with high textures on dynamic player models isn't easier to see. I am going to argue that there is simply no reason for the optional advantage to exist when 16x anistropic, in regards to performance, is ubiquitous on modern hardware.

fuck that cs niggger account. All it posts is sensationalist bullshit.I blocked them long ago.

We need some kind of central lore for these kids man, they keep discovering "mysteries" That are lessons we learned decades ago.

Pros don’t really give a fuck besides guys like Ropz

Best option is to spot enemies easily , is to use cheats like Wallhack , no need to play around with settings , follow me for more tips .

I can bet money only the super nerdy pros even consider something like this. 99% just put shit on low for performance and fuck the rest. Also the clutter example is nitpicky. In some areas more detail may reduce clutter.

explain HOW more detail could reduce clutter. Give me a single example. Go ahead. Literally an oxymoron, it's so wild. And I said they have an INTUITIVE understanding. It very well could be that "it feels better" is as deep as they model it mentally. It doesn't matter, the advantage is there.

I am pretty sure donk is playing on random settings and still owning. Most pros know nothing about settings or meta. Only when they are consistently disadvantage they think to adjust something like the Nvidia reflex thing they played with it off for so long and SSG553

The difference is so fucking minor as to be pointless.

I would say a bigger advantage would be to play 16:9 instead of 4:3, yet most pros play 4:3. The advantage here seems negligible but i hear you.

No they don’t, most pros hire these “optimizers” who don’t know jack shit about pc gaming and make a living changing schizo placebo settings that don’t have any real effect. A blurry non filtered texture is a much bigger distraction than a uniform texture that our brains ignore.

literal bullshit and if you're implying that I'm an optimizer you're dumber than a rock, I already posted ACTUAL RESEARCH as if it was needed, I can't believe how many people fail to understand such an obvious perceptual point that if you add more detail around a target it stands out less, I am seriously worried for how even such an obvious thing (so much so that we use words like "stand out" to describe this perceptual phenomenon and you can't fucking do 2+2 and figure out that less information around a target makes it stand out, and yes information==high frequency features, aka edges and sharp gradients of colors) still eludes so many of you, and even more worried about how people like you actually think the opposite is true. It's SHOCKING

adding details is the point of using 16x xd to get better nade line ups and having more details to contrast player models off of is better, than a washed out muddy visual mess for player models and textures. The problem would be if the higher quality would also affect player models, then it would increase clutter, because high quality player models wouldn't contrast over the background, as both would be to detailed to instantly at a glance see changes.

Son 😭😭

Bro can we just use nearest I Want the #crunch on those pixels bro

Input lag is different between Bilinear and 16x

Exactly. 👍 Its not big difference, but might win some fights. Not pro, thinking as a graphics engineer.