Detonator 52.14 Test & 52.10 Re-Test

October 8, 2003 / by Leonidas / page 3 of 7

   Direct3D Filter Quality  (cont.)

This is the reason for our second mistake we made in the original article about the 52.10 Detonator. Because of the correctly working application setting, the Detonators 52.10 and 52.14 are offering two different filter qualities from nVidia for the first time. On the one hand, there are the applications settings as shown above, and on the over hand, the settings by the means of the control panel, as shown below (CP = per Control Panel):

52.10/14 Direct3D 0xAF (CP)

52.10/14 Direct3D 2xAF (CP)

52.10/14 Direct3D 4xAF (CP)

52.10/14 Direct3D 8xAF (CP)

Setting the filter quality by the means of the control panel, we can see the known "optimizations" of the Detonator 51.75, here in comparison with the exemplary filter quality of the Detonator 45.23 (MouseOver effect using Javascript, alternative: click opens both screenshots):

The trilinear texture filter is still replaced by a faked trilinear filter and the texture stages 1 till 7 are filtered with a 2x anisotropic filter at the best. Nevertheless, there is one more difference between the two detonators 52.10 and 51.75 (MouseOver effect using Javascript, alternative: click opens both screenshots):

Other than the Detonator 51.75, the Detonator 52.10 and 52.14 differ in so far as that the faked trilinear filter is also used for the texture stage 0 (which usually is the basis texture). That’s why both new Detonators are more "optimized" than the 51.75!