Re: All Graphics Cards related queries here.
What are these AF...AA like 4AA or 2AA...
are both these two important while gaming....
what will the difference if if dont use above two(off)
more over is graphics improved if i play @1400 * 900 coz i play @ 1024*768
Well here you go:
Anti - Aliasing (AA):
Is a technique used to remove the distortion from images. All images are rendered by creating polygon shapes (primitives), and then joining them together. Now obviously a polygon does not have 'round' edges, so we will see jagged edges for non angular shapes. (Example a pipe or wheel). The graphic rendering system sees objects which will have edges of this manner, and apply an AA scheme to those portions of the image. These are basically again primitives which have a different level of opacity then the original image. These two are then joined at run time (the original smooth edges polygon+the AA supporting primitive) to create a non-jagged soft edge.
So if you set 4xAA as the render method it is actually rendering 4 primitives per pixel on your screen at run time...! It is extremely easy to correlate why the AA level immediately take a toll on the FPS.
A method(s) used by GPUs is Full Scene Anti Aliasing (FSAA) or Custom Filter Anti Aliasing (CFAA). You can say FSAA does the work in a full batch per frame. It takes the frame and renders it 2x or 4x times the normal resolution, and then down samples before it is pushed to the screen. So at 4x FSAA each pixel would have 16 super sample pixels. CFAA is a technology that ATI cards are using.
Anistropic Filtering (AF):
Is used to enhance the quality of 'textures' in images. It is used to show greater details at angles which are distant or oblique to the viewer. For example a farm house really far away. This is with respect to the camera and not actually the polygon. This takes up a huge amount of memory bandwidth. Lower resolution textures are used for objects that are further away from the camers (read player), to show the affect of distance.
To quicken this process a logic of 'mipmapping' is used. For example if a window is rendered using 256x256 pixels. Now you want to show this window at a distance, what does one do. It makes the window smaller in size --- but why use 256x256 pixels again..? So a shrunk version with a high quality filter is used. Any many such textures are 'stored' at run time, and re used over and over again. But this involves a lot of calculations and is resourceful on the GPU.
Now another issue arises. With so many mipmaps, when they are joined together, the 'joins' will not look nice. You will get uneven surfaces as distance is varied. We will see this as artifacts. This is corrected by using Trilinear filtering, in this method each mipmap is filtered along x-y axis but also with respect to the mipmap next to it. So as the angle is changed the AF will start to vary.
You have a good GPU. Try to run games, with these setting as a variable, you will start to see a difference. To notice AA, look closely at edges (no linear). And to notice AF look at distant objects. Yes these are important. They are the
eye-candy drivers, and vary the performance of the game.
Resolution:
1. 1400 * 900 = 1260000 pixels x FPS
2. 1024*768 = 786432 pixels x FPS
By using (1) you are getting ~ 40% more pixels on the screen. Of course you will see better images, but the GPU has to do more work, so you might see a performance hit.
I totally agree with asigh. It is better to go with the rating given by manufacturer than to push the PSU to the limits.
you get a 450W PSU, you make sure that you draw 450W or less power from it.
You have a rig that takes 550W, then you better get a PSU that is RATED to give 550W power.
Could not have said it better.
