What are you working on? V4 (HTML ISN'T PROGRAMMING) - Developers

[QUOTE=Robber;18152341]It won't make a difference. I could crop the high resolution image and would get the lower resolution image with the same quality and aliasing.[/QUOTE] If you display the images at the same resolution as they were taken at it does make a difference.

[QUOTE=jA_cOp;18151946]You do get it perfectly fine, he just has a hobby of writing relatively simple things in an overly complicated tone. Don't let it confuse you.[/QUOTE] It's the way I think. When I drive a car, I don't think "this pedal on the right makes me go faster and the one on the left makes it go slower," I think "this pedal on the right modulates the fuel intake of the engine, which I will assume is directly proportional to the power output of the engine."

[QUOTE=Jawalt;18152365]If you display the images at the same resolution as they were taken at it does make a difference.[/QUOTE] You got it the wrong way around. If I display it at 1024x768 it doesn't make a difference. When I display it at 1680x1050 it does make a difference. [IMG]http://i230.photobucket.com/albums/ee15/FPRobber/Unigine/noaa.png[/IMG][IMG]http://i230.photobucket.com/albums/ee15/FPRobber/Unigine/aa.png[/IMG] This was taking at 1680x1050, so you can't complain about that.

[QUOTE=Jawalt;18151743][IMG]http://i132.photobucket.com/albums/q17/Mattayu/sample2.png[/IMG] Approx 17x17 resized to 50x50 [IMG]http://i132.photobucket.com/albums/q17/Mattayu/sample1.png[/IMG] A line at native 50x50 Both of these lines are identical in angle and drawing style, just one is of a lesser 'resolution'. Now tell me that resolution doesn't have shit to do with AA.[/QUOTE] you're a moron. do you even know how sampling works?

[QUOTE=Robber;18152491]You got it the wrong way around. If I display it at 1024x768 it doesn't make a difference. When I display it at 1680x1050 it does make a difference. [IMG]http://i230.photobucket.com/albums/ee15/FPRobber/Unigine/noaa.png[/IMG][IMG]http://i230.photobucket.com/albums/ee15/FPRobber/Unigine/aa.png[/IMG] This was taking at 1680x1050, so you can't complain about that.[/QUOTE] In screenshots, aliasing can be a problem, but when you are saving the universe and there are explosions all around and you're screen is shaking all the time you won't notice any jagged lines.

[QUOTE=xxxkiller;18153503]In screenshots, aliasing can be a problem, but when you are saving the universe and there are explosions all around and you're screen is shaking all the time you won't notice any jagged lines.[/QUOTE] Lines which have a gradial gradient such as distant terrain is noticeable. I play with 4x AA in newer games since my 8800GT provides it with no performance hit with its dedicated AA processing and it's not distinguishable between things like 16x when you are in a game. In source games you may as well wack it up to 16x anyway.

As far as I know, there are no monitors in wide circulation that have a high enough pixel density to eliminate the need for anti aliasing. Anti aliasing is the simulation of a higher resolution by blending colors based on how much of a pixel is covered by each object's edge. In fact, some anti aliasing is done by rendering at a higher a resolution and then shrinking it down.

I've always been curious about aa. What are the fastest methods? How do they work?

[QUOTE=xxxkiller;18153503]In screenshots, aliasing can be a problem, but when you are saving the universe and there are explosions all around and you're screen is shaking all the time you won't notice any jagged lines.[/QUOTE] I do.

[QUOTE=SupahVee;18153771]I've always been curious about aa. What are the fastest methods? How do they work?[/QUOTE] The most straightforward way to do antialiasing is to super-sample. This means essentially rendering the entire image at a higher resolution, and then scaling it down to native size, the final value of a pixel being the average of 4 or so pixels. But this isn't the way it's actually done. That's too computationally expensive. (notice how I just said it is the straightforward way) Antialiasing is done [I]only[/I] for the polygon edges as these areas are the places most likely to find high-frequency content. If you know how the fragment shader works and barycentric coordinates, then you would know that one of the 3 coordinate components is between 0 and PIXEL_SIZE for fragments along the edge of a polygon. (PIXEL_SIZE is another variable that varies proportional with barycentric coordinates. It represents the size of a pixel in clipping space, and also the distance in clipping space between 2 samples) The opacity of the generated fragment is inversely proportional to how close the smallest coordinate is to 0.

You took the words right out of my mouth.

Actually, the probably even more straightforward way is to run the whole image through a low-pass filter. It does kill the aliasing, but it also makes the image look like blurry shit. [editline]05:03PM[/editline] [QUOTE=ryandaniels;18153747]As far as I know, there are no monitors in wide circulation that have a high enough pixel density to eliminate the need for anti aliasing. [/QUOTE] Actually, as long as monitors are still using pixels (discrete samples of the virtual image) then they will not be enough to completely get rid of aliasing. In fact, short of low-pass filtering, there is no way at all to completely get rid of aliasing. We can only combat it, or make the aliased region smaller. The reasoning behind this is with 3d graphics (or any vector graphics for that matter) have an [I]infinite[/I] spacial frequency along the edge of a line or polygon. (One sample pixel is on the face of the polygon, and the adjacent sample falls on the background.) No amount of resolution or pixel density (sample frequency) is going to change this simple fact. In order to completely be without aliasing, you are going to super-sample at [I]twice times infinity[/I] resolution. Clearly impossible.

Today I wrote a BrainFuck interpreter in JavaScript. But there's a difference! It plays out slowly so you can watch its progress and watch the memory registers change. [url=http://jakemadethis.com/d/brainfuck.html][b]See it here BRAINFUCK[/b][/url] I might change the interface a bit and add different speed options. Give me some creative brainfuck scripts to add in the examples please. Edit: I've only tested in Firefox 3.5

[QUOTE=Vampired;18155064]Today I wrote a BrainFuck interpreter in JavaScript. But there's a difference! It plays out slowly so you can watch its progress and watch the memory registers change. [url=http://jakemadethis.com/d/brainfuck.html][b]See it here BRAINFUCK[/b][/url] I might change the interface a bit and add different speed options. Give me some creative brainfuck scripts to add in the examples please.[/QUOTE] That looks pretty awesome.

[url]http://99-bottles-of-beer.net/language-brainfuck-101.html[/url] That will take a while given the slow mo.

Feeling nostalgic, working on an inventory database program in BASIC, keeping the program itself as small as possible. Nearly done, with only 50 lines of code.

[QUOTE=Jallen;18155283][url]http://99-bottles-of-beer.net/language-brainfuck-101.html[/url] That will take a while given the slow mo.[/QUOTE] Its finally started outputting after 15 minutes of execution. Its at 95 bottles, and decreasing slowly. EDIT: One hour and counting, and still 28 bottles to go.

[QUOTE=Cathbadh;18154354]Actually, the probably even more straightforward way is to run the whole image through a low-pass filter. It does kill the aliasing, but it also makes the image look like blurry shit. [editline]05:03PM[/editline] Actually, as long as monitors are still using pixels (discrete samples of the virtual image) then they will not be enough to completely get rid of aliasing. In fact, short of low-pass filtering, there is no way at all to completely get rid of aliasing. We can only combat it, or make the aliased region smaller. The reasoning behind this is with 3d graphics (or any vector graphics for that matter) have an [I]infinite[/I] spacial frequency along the edge of a line or polygon. (One sample pixel is on the face of the polygon, and the adjacent sample falls on the background.) No amount of resolution or pixel density (sample frequency) is going to change this simple fact. In order to completely be without aliasing, you are going to super-sample at [I]twice times infinity[/I] resolution. Clearly impossible.[/QUOTE] True, but I was talking about perceptible aliasing. At a certain pixel density, our eyes wouldn't be able to notice the aliasing. Although obviously it's much more feasible to just use anti-aliasing. Also, distance aliasing would probably still look terrible even on a screen with a very high pixel density.

Shit, I should start going with my Arduino again. My last bit of code failed hard, so I should probably try recreating it step by step.

So, let's say I'm making commands for said video layer on my engine. Would I just have stuf like renderQuad(vertex1, vertex2, vertex3, vertex4) renderTriangle(vertex1, vertex2, vertex3) or is there some hidden code I'm missing out on?

[QUOTE=jmazouri;18139061]Isn't that the same :|[/QUOTE] No, the cel shading one has outlines and stepped light values, Borderlands is normal shading with outlines.

Remember how I had to start over with my build system? Well tonight (after a half week off because of borderlands), we finally got very basic binaries compiling, and linking. Currently only supports the C language but it will be trivial to setup the C++ compiler settings. Still need to put in file checking (i.e. has the file already been compiled/changed in anyway), but the backend for that is already setup, and it's really just a simple if statement to check if we need to compile it :) So excited, because I'm so close to 0.1. :buddy:

[url]http://winsaucestudios.com/forum/index.php?action=mgallery;sa=item;id=236[/url] Sorry about the laggy, shitty video, there's no decent free screen recorders for macs. But anyway :D It's just a set of entities really, and one map brush.

[QUOTE='-[ Fizzadar ]-;18165138'][url]http://winsaucestudios.com/forum/index.php?action=mgallery;sa=item;id=236[/url] Sorry about the laggy, shitty video, there's no decent free screen recorders for macs. But anyway :D It's just a set of entities really, and one map brush.[/QUOTE] can you use a format other than piece of shit quicktime?

[QUOTE=SteveUK;18165675]can you use a format other than piece of shit quicktime?[/QUOTE] Not with the piece of shit recorder I'm using D:

ffmpeg2theora

[img]http://imgkk.com/i/GhgSva.png[/img]

What is this?

vBulletin skin.

[QUOTE=compwhiziitothemax;18167534]vBulletin skin.[/QUOTE] And that's relevant to this thread because?