Doom 3 Requirements.

[BenSkywalker]

AF is anisotrophic filtering. It allows textures to be rendered more clearly and is especially noticable at long distances and at sharp angles.

Better to remain silent and thought a fool then to speak and prove it.

You have described a LOD bias adjustment, not anisotropic filtering.

 

[Cat]

Edit: running for the hills.

 

[Nebor]

Originally posted by: BenSkywalker

AF is anisotrophic filtering. It allows textures to be rendered more clearly and is especially noticable at long distances and at sharp angles.

Better to remain silent and thought a fool then to speak and prove it.

You have described a LOD bias adjustment, not anisotropic filtering.

?A method of filtering textures using a nonsquare area. With this method, a pixel may encompass information from many texture elements (called texels) in one direction and fewer in another. It yields sharp textures on objects that slant away from the viewing plane, such as a road running into the distance.

:roll:

It's kinda cute that you thought you were witty though. Now run along back to ATOT where you can be ridiculed mercilessly.

 

[BD2003]

Originally posted by: BenSkywalker

AF is anisotrophic filtering. It allows textures to be rendered more clearly and is especially noticable at long distances and at sharp angles.

Better to remain silent and thought a fool then to speak and prove it.

You have described a LOD bias adjustment, not anisotropic filtering.

LOL, actually he was perfectly right in his defintion. You are completely and utterly wrong.

The only thing a LOD bias will do is change the distance at which one mipmap is switched for the next. Raising it will make textures sharper *in some ways*, especially in screenshots. But it will also make a shimmering effect, because theres too many texels for the number of pixels. AF is an improvement, whereas a improper LOD bias is almost always a downgrade.

Nice try though. :laugh:

 

[clicknext]

Originally posted by: BenSkywalker

AF is anisotrophic filtering. It allows textures to be rendered more clearly and is especially noticable at long distances and at sharp angles.

Better to remain silent and thought a fool then to speak and prove it.

You have described a LOD bias adjustment, not anisotropic filtering.

lol, he described what visual effect AF has. I'm guessing you don't use AF normally.

 

[Nebor]

You'd think he'd been here long enough to know better.

 

[BFG10K]

Of course for you, it has to be 1600x1200 @ 8xAA 32xAF and 210fps to be playable.

Not "for me". Resolution sharpens the scene but it can't help the artifacts caused by bilinear/trilinear, artifacts most commonly manifested as blurry filtering.

Defending one of the forum's least popular members is not a good way to go.

I most certainly I didn't see anyone defending you.

Of course I know what it does.

Actually it appears you don't, otherwise you would've never made your original nonsensical argument to begin with.

LOL, actually he was perfectly right in his defintion. You are completely and utterly wrong.

While Nebor's generic definition was reasonable, I believe Ben is referring to AF having the requirement of increasing texture sharpness without increasing texture aliasing. He's also probably going to mention a non-square sampling pattern that takes more samples in the direction of the stretch and its angle.

If you boys want to argue with him then you'd better know what you're doing because he's certainly no "n00b".

 

[Nebor]

quote:

--------------------------------------------------------------------------------
Defending one of the forum's least popular members is not a good way to go.
--------------------------------------------------------------------------------


I most certainly I didn't see anyone defending you.

That's funny, coming from you. If we could get Alkemyst to say it, that would kill.

See, it's not funny or witty when you say it, because you're mean to people. I'm nice, and try to help people. I don't storm into their threads saying, "That's UNPLAYABLE!!! Get a CRT!!! Unacceptable!" Be a little nicer and maybe you'll find some good folks to back you up some day.

 

[BD2003]

LOL, actually he was perfectly right in his defintion. You are completely and utterly wrong.

While Nebor's generic definition was reasonable, I believe Ben is referring to AF having the requirement of increasing texture sharpness without increasing texture aliasing. He's also probably going to mention a non-square sampling pattern that takes more samples in the direction of the stretch and its angle.

If you boys want to argue with him then you'd better know what you're doing because he's certainly no "n00b".[/quote]

And where exactly do you get the information to say what you believe he's referring to? Because your description of it certainly sounds fine, but no where do I see ben say anything of the sort.

Anyways, I've dealt with him before, and I a have been long convinced that he's a retard who has no idea what he's talking about. I remember quite some time back that he actually thought the nforce's soundstorm was doing amazing things no sound card could possibly do. Any attempt to prove how absolutely and utterly wrong he was (which seems to be the case quite often) was met with total ignorance. He seems to be capable of nothing else but telling OTHER people they have no idea what theyre talking about, when in reality he appears to know nothing himself.

I dont know what he has an elite for, but if its for stuff like this, it should be immediately revoked.

 

[BenSkywalker]

Nebor

A method of filtering textures using a nonsquare area. With this method, a pixel may encompass information from many texture elements (called texels) in one direction and fewer in another. It yields sharp textures on objects that slant away from the viewing plane, such as a road running into the distance.

The definition above is somewhat passable, although it should read may yield, not yields in terms of sharp textures as the way it is worded is improper in its implication that it will yield sharper textures which is certainly not a given. Although, by pure definitive standards anisotropic filtering only implies non square- all the rest is a byproduct of adjustments to the calculations afforded by the modified sampling pattern(barring an incorrect selection of LOD bias).

That definition however, is something totally different then what you prior post stated which was a LOD bias adjustment- increasing detail on textures while mentioning nothing at all about the modifed sampling pattern which is requisite of anisotropic filtering.

Now run along back to ATOT where you can be ridiculed mercilessly.

I think I maybe have half a dozen posts in OT in the last six months, Vid forum is where I almost always am.

I'm nice, and try to help people. I don't storm into their threads saying, "That's UNPLAYABLE!!! Get a CRT!!! Unacceptable!"

So you think giving people lousy advice is being nice, while BFG being honest is mean? He tries to help people by giving them accurate information, not misinformed or dishonest drivel.

BD2003-

The only thing a LOD bias will do is change the distance at which one mipmap is switched for the next.

And his quote was-

It allows textures to be rendered more clearly and is especially noticable at long distances and at sharp angles.

I think if you try real hard something may start to click for you.

But it will also make a shimmering effect, because theres too many texels for the number of pixels.

It makes a shimmering effect because of a lack of proper sampling for the frequency that the image is being displayed at. I assume by your tone that you must be an expert on signal theory and how it relates to texture sampling in a 3D environment, why word things like a moron? Don't hold back now, please expand on what it is you think constitutes proper anisotropic filtering and what sort of shortcomings current implementations have and why those shortcoming have been made in relation to performance not to mention their resultant impact on transistor budgets on current GPUs.

And where exactly do you get the information to say what you believe he's referring to? Because your description of it certainly sounds fine, but no where do I see ben say anything of the sort.

These forums have archives, look in to the Vid card forum on this topic and use my name as a cross reference. BFG has seen my go off on enough people over the years because of their half witted, at best, comments on texture filtering to be quite aware of how it is I will respond. Anisotropic filtering implementations in particular is something I'm rather quite obsessed with and have been for years, BFG is also quite familiar with that.

I remember quite some time back that he actually thought the nforce's soundstorm was doing amazing things no sound card could possibly do.

I quoted nVidia's tech docs in that discussion including the amount of latency DD encoding took along with the amount of samples they used to generate the effect. All you had to do was go to nVidia's site and check it out for yourself.

I dont know what he has an elite for, but if its for stuff like this, it should be immediately revoked.

Because someone makes a moronic statement and insults a long time acquaitance of mine and I call them on it? Please. You want to debate the topic at hand feel free to do so.

 

[newParadigm]

how do u think my rig will do?

U proly dont no tho.


AthlonXP 2500m @ 2.415Ghz (230x10.5)
512mbs Micron DDR3700
GFX5600ultra
Seagate 80gbx2 RAID0

Thanks,
newParadime

 

[clicknext]

Originally posted by: newParadime
how do u think my rig will do?

U proly dont no tho.


AthlonXP 2500m @ 2.415Ghz (230x10.5)
512mbs Micron DDR3700
GFX5600ultra
Seagate 80gbx2 RAID0

Thanks,
newParadime

It will do o-k-a-y! Heh, other than that... wait until it comes out or wait for the demo.

 

[BD2003]

The definition above is somewhat passable, although it should read may yield, not yields in terms of sharp textures as the way it is worded is improper in its implication that it will yield sharper textures which is certainly not a given. Although, by pure definitive standards anisotropic filtering only implies non square- all the rest is a byproduct of adjustments to the calculations afforded by the modified sampling pattern(barring an incorrect selection of LOD bias).

That definition however, is something totally different then what you prior post stated which was a LOD bias adjustment- increasing detail on textures while mentioning nothing at all about the modifed sampling pattern which is requisite of anisotropic filtering.

I did not realize that a pragmatic functional definition of something wasnt enough to satify most people around here. Apparently we need an all out technical definition or else we're entirely wrong. :roll:

Just because he didnt specify the sampling pattern doesnt mean that he doesn't know thats how it works. Even if he didnt, it barely even matters because his defintion is perfectly fine for 99% of the people who arent so anal as to chastise someone for not specifying each and every single nuance of the definition.

People turn on anisotropic filtering BECAUSE it sharpens textures. In fact, I challenge you to find a real world situation where turning on AF will blur or otherwise make image quality worse.

The only thing a LOD bias will do is change the distance at which one mipmap is switched for the next.

And his quote was-

It allows textures to be rendered more clearly and is especially noticable at long distances and at sharp angles.

I think if you try real hard something may start to click for you.

Thats a brilliant deduction there sherlock. Just because an LOD bias change and AF are both noticible at long distances and sharp angles do not mean that anyone is mistaken as to which is which. LOD Bias is an ugly hack to sharpen textures. AF is a more proper way to do it. Some basic logic wouldnt hurt you know.

It makes a shimmering effect because of a lack of proper sampling for the frequency that the image is being displayed at. I assume by your tone that you must be an expert on signal theory and how it relates to texture sampling in a 3D environment, why word things like a moron? Don't hold back now, please expand on what it is you think constitutes proper anisotropic filtering and what sort of shortcomings current implementations have and why those shortcoming have been made in relation to performance not to mention their resultant impact on transistor budgets on current GPUs.

Again, the same BS. People dont come here for highly technical definitions that would be utterly useless to them. They want to know what things do, and what theyre for, and how they could be helped. You are doing nothing of the sort. I dont have the time nor inclination to seek out detailed specifications of the algorithms behind every 3d technology.

And as far as tone goes, you should keep in mind who was the one to imply that a certain someone was a fool.

These forums have archives, look in to the Vid card forum on this topic and use my name as a cross reference. BFG has seen my go off on enough people over the years because of their half witted, at best, comments on texture filtering to be quite aware of how it is I will respond. Anisotropic filtering implementations in particular is something I'm rather quite obsessed with and have been for years, BFG is also quite familiar with that.

Then perhaps maybe you should stick to the highly technical or video card forums where someone might find such obtuse knowledge of anisotropic filterings, which would be more than useful in an actual discussion of the theory behind it. This thread is certainly not that place.

I quoted nVidia's tech docs in that discussion including the amount of latency DD encoding took along with the amount of samples they used to generate the effect. All you had to do was go to nVidia's site and check it out for yourself.

And you're still wrong about it, and it had very little to do with the latency of DD encoding. We could dig that all up again, but in the end I had to leave that discussion because you were failing to listen to reason.

Because someone makes a moronic statement and insults a long time acquaitance of mine and I call them on it? Please. You want to debate the topic at hand feel free to do so.

There was nothing moronic about it, and while you may be pretty satisfied with yourself for calling someone out for not specifying the actual algorithm used, in the end, you're still the one who looks the fool.

 

[BFG10K]

Alkemyst

Should I know who that is?

I'm nice,

Would you like a cookie?

and try to help people.

I've helped hundreds if not thousands of people in my time here.

Be a little nicer and maybe you'll find some good folks to back you up some day.

That comment is beyond ludicrous, especially when factoring in that it's coming from you.

 

[BFG10K]

And where exactly do you get the information to say what you believe he's referring to?

Based on what Nebor didn't say.

Because your description of it certainly sounds fine, but no where do I see ben say anything of the sort.

Well you'd have to ask what Ben what he meant, but based on my past history with him I have a reasonable idea of how he thinks. Regardless, this whole thing started with Nebor's original inept comment when he responded to mine.

Anyways, I've dealt with him before, and I a have been long convinced that he's a retard who has no idea what he's talking about.

I'm astonished that you'd call Ben clueless and then defend Nebor, especially since Nebor's first moronic argument was followed by a loose definition followed by something which he probably pulled from a website when questioned.

I remember quite some time back that he actually thought the nforce's soundstorm was doing amazing things no sound card could possibly do.

Probably 5.1 Dolby Digital Out but again you'd have to ask him.

 

[BenSkywalker]

did not realize that a pragmatic functional definition of something wasnt enough to satify most people around here.

So your definition of a heart attack is shooting pain in the left side of the body? It wasn't a passable definition of what anisotropic is or anything remotely close to it- it was an effect of what it allows.

Just because he didnt specify the sampling pattern

That is the ONLY thing anisotropic filtering implies, the sampling pattern.

People turn on anisotropic filtering BECAUSE it sharpens textures. In fact, I challenge you to find a real world situation where turning on AF will blur or otherwise make image quality worse.

I said it didn't always sharpen textures which is point of fact. Tron 2.0(to answer your next question). You need to have the additional data in which to take samples from or to generate noise for the impact of the LOD bias selection to make an impact on the data. Real basic stuff here, and quite self obvious.

Thats a brilliant deduction there sherlock. Just because an LOD bias change and AF are both noticible at long distances and sharp angles do not mean that anyone is mistaken as to which is which.

And if someone were properly aware of the difference they wouldn't describe a LOD bias change as anisotropic filtering completely ignoring the one factor that is absolutely required for it to be dubbed anisotropic.

LOD Bias is an ugly hack to sharpen textures. AF is a more proper way to do it. Some basic logic wouldnt hurt you know.

But defending someone who defines anisotropic filtering by the modified LOD bias selection is truly logical.

And as far as tone goes, you should keep in mind who was the one to imply that a certain someone was a fool.

Nebor implied it about BFG, I recall quite clearly.

Then perhaps maybe you should stick to the highly technical or video card forums where someone might find such obtuse knowledge of anisotropic filterings, which would be more than useful in an actual discussion of the theory behind it. This thread is certainly not that place.

I don't understand why I should tell people the wrong thing just because it's easier. Even at its most basic level the description given of AF was wrong.

And you're still wrong about it, and it had very little to do with the latency of DD encoding.

You can argue that point with XB developers or nVidia engineers, I quoted the specs.

There was nothing moronic about it,

Of course for you, it has to be 1600x1200 @ 8xAA 32xAF and 210fps to be playable.

Really?

 

[BD2003]

First off, thats a terrible analogy. But I dont need to be a cardiologist to know that sharp pain in left side = heart attack = go to hospital.

Listen, when someone asks what AA does or is for, you dont go into a tirade on sampling patterns, you tell them that it smoothes the edges and gets rid of the jaggies. Or is that not good enough?

I dont care how technical you want to or have to be, for 99.999999% of the people in the world, AF is something they use to sharpen textures, particulary in the distance or sharp angles, and couldnt care less about the algorithm itself, or anything to do with sampling patterns. Therefore AF = Sharper textures. That is not a giant step of logic. Wtf is so hard to understand about that?

I dont know if chastising people for not giving a ten page defintion of a term gets you off, but its pretty retarded.

And for the record, Im not defending Nebor, and I couldnt care less who the hell he is. He even sounds a bit like a dick. But I have to speak out when the terminology nazi's come along on their high horses and start crucifying people who give a defininition that is barely even "loose".

 

[Nebor]

I don't really see the point in all this arguing. It's not going anywhere. It is spanning pages though.

 

[apoppin]

Originally posted by: Nebor
I don't really see the point in all this arguing. It's not going anywhere. It is spanning pages though.

i don't have popcorn but i am eating peanuts and enjoying the show (since i can't sleep)

PLUS, i am getting more data on AF than i EVER cared to know (although i had a reasonable grasp b4)

carry on :roll:

 

[Nebor]

Originally posted by: apoppin

Originally posted by: Nebor
I don't really see the point in all this arguing. It's not going anywhere. It is spanning pages though.

i don't have popcorn but i am eating peanuts and enjoying the show (since i can't sleep)

PLUS, i am getting more data on AF than i EVER cared to know (although i had a reasonable grasp b4)

carry on :roll:

Yeah, but keep in mind who this 'data' is coming from. Maybe we can get Ben and BFG to fight each other. I imagine them as one of those cartoon balls of dust with arms and feet and stuff kicking out occasionally as the ball of dust rolls around.

 

[Budman]

Originally posted by: Nebor
I don't really see the point in all this arguing. It's not going anywhere. It is spanning pages though.

what you said made me think of this Picture.

 

[BenSkywalker]

But I dont need to be a cardiologist to know that sharp pain in left side = heart attack = go to hospital.

That's because over the course of our lives we don't have a bunch of people stating that it is commonly something else. If everytime someone asked what is AF and they were given the proper response we wouldn't need to be bothered with half witted/ p!ss poor/ inaccurate explenations being apologized for.

Listen, when someone asks what AA does or is for, you dont go into a tirade on sampling patterns, you tell them that it smoothes the edges and gets rid of the jaggies. Or is that not good enough?

This is a good example of why I don't hand out half witted explenations. You just explained one benefit of AA that is the only benefit of one type of AA. AA is for reducing aliasing, edge aliasing is not the only type and there are other methods used to deal with aliasing including AF. Texture aliasing is at least as bothersome as edge aliasing for a lot of people, and anisotropic filtering's ability to significantly reduce texture aliasing is a major benefit that the 'LOD bias' type explenation completely ignores. Saying AA 'gets rid of the jaggies' influences people to think that is what AA does and that's all it is for, why give someone that impression? We end up with a bunch of people that think they know what they are talking about when they don't have a freakin clue. Does it take all that much more effort to type out a few more sentences to explain it at least in a passable fashion? If the person is interested enough about to ask why short change them with misleading or false information?

I dont care how technical you want to or have to be, for 99.999999% of the people in the world, AF is something they use to sharpen textures, particulary in the distance or sharp angles, and couldnt care less about the algorithm itself, or anything to do with sampling patterns.

If they don't care enough to learn what it does, then why should they answer someone who asks about it?

I dont know if chastising people for not giving a ten page defintion of a term gets you off, but its pretty retarded.

Being honest and accurate is retarded, but handing out inaccurate information and misleading people is the right thing to do? Not in my world.

But I have to speak out when the terminology nazi's come along on their high horses and start crucifying people who give a defininition that is barely even "loose".

I'm not the one spreading misinformation to keep the populace in the dark, that would be your end of the discussion.

Nebor-

I don't really see the point in all this arguing.

Making it clear that you don't know what you're talking about, and that has been accomplished quite soundly.

It's not going anywhere.

It has run its course, and hopefully now you understand AF a bit better and won't be so apt to hand out misinformation.

Yeah, but keep in mind who this 'data' is coming from.

Go ahead and research it for yourself, although you already have the proper explenations here.

Maybe we can get Ben and BFG to fight each other.

BFG and I have argued more about 3D related topics then any two people I'm aware of for certain, it's just on such incredibly basic elements there is nothing to argue about. These are hard facts we are talking about here(much as the handicap LCDs have on gaming is). You can argue degrees if you would like, you can argue about implementations and their end impact- the basic core of what these things do is not arguable by any two people who understand them.

 

[VirtualLarry]

Originally posted by: Nebor
Yeah, but keep in mind who this 'data' is coming from. Maybe we can get Ben and BFG to fight each other. I imagine them as one of those cartoon balls of dust with arms and feet and stuff kicking out occasionally as the ball of dust rolls around.

Or like those super-speed aerial battles in DragonBallZ?

(Not really a fan, but that was the first image that came to mind.)

Personally, I like technical accuracy about things. It drives me extremely nuts, whenever I see "PCI/AGP lock" used as a term to describe async bus frequencies. In fact, the AT 915/925X OC article right here, used terminology that was almost entirely wrong, and the initial meanings that were conjured up by terms like "float", turned out to be the exact opposite of what the writer was actually trying to state.

Most AT articles are top-notch, I was actually somewhat disappointed by that one, mostly due to the misleading terminology used. Otherwise, it was still interesting on a technical level.

 

[Fuchs]

I like bananas

 

[amdguy]

i like pinky, he's so cute