Poser

Following 11, 303 Followers and 49 Favorites

This is the FIXED version of the previous PP2014 MT5 (with Atan_Approximation compound node) here http://www.sharecg.com/v/88194/gallery/11/Poser/MT5- …

more »

This is the FIXED version of the previous PP2014 MT5 (with Atan_Approximation compound node) here http://www.sharecg.com/v/88194/gallery/11/Poser/MT5-Atan2_Approx_With_Error

Details of the fix are as posted here -https://forum.smithmicro.com/topic/1918/pp2014-material-room-math_functions-nodes-no-atan-asin-acos/36

The atan(X) approximation is as found here http://nghiaho.com/?p=997 which is valid for X between -1 and +1 and returns angles in the range -π/4 to +π/4 radians (-45° to +45°).  Max error seems to be just under 0.1° (compared to my spreadsheet atan() function).The approximation is:

atan(X) ≈ (π/4 * X) - (X * (abs(X)-1) * (0.2447 + (0.0663 * abs(X))))

 

The ZIP file contains just the 'Atan_Approx1_FIX1' MT5 and PNG icon in the following subfolder:

Runtime\Libraries\Materials\3DCheapskate\Generic\Test

The big MaterialRoom image shows the network with test nodes before I collapsed the relevant bits into a compoundnode « Less