Oblivion with HDR (High Dynamic Range) :: 100% Ubuntu Linux + WINE 1.1.18 + NVidia 180.11 Drivers

Yes, another oblivion video, but this time with high dynamic range support.. I got this to work according to directions given here at the WINE application database at WINE HQ: Kasperi Kuuskoski's comment http://appdb.winehq.org/objectManager.php?sClass=version&iId=7870 This one has more talk about getting HDR to work http://appdb.winehq.org/objectManager.php?sClass=version&iId=7506 Note, I'm using the 180.11 version of the NVidia drivers for linux, and I'm using Wine 1.1.18 in this demo. What is HDR? Normally images store integer values for red, green and blue, 0 to 255.. But an HDR image stores RGB values are floating point numbers eg. 1.036364E6 which means the number 1036364, anytime a calculator has to represent a large number it uses this notation, but inside a computer there is a data type called a floating point number that keeps track of the location of the decimal place, thus the name "floating point" and the number. There are various ways to represent a floating point, like using the binary representation of a number and specifying the location of the decimal place.. Or using two integers to represent the whole and fractional parts, and the decimal place. A fractional binary number produces a representation like ".1011" which is 1/2 + 0/4 + 1/8 + 1/16 . Whereas a whole binary representation looks like 1011. which is 8 + 2 + 1 . This is why in some programs, like blender, where if you scale down an object, and scale it back up the precision is lost, because binary numbers are not very good at representing fractions, they are approximate. So why binary and not decimal?, because the computer works best in binary computations than in decimal, also it depends on the programmer, what they want to do.. Anyhow, the purpose of HDR is to permit you to represent values outside the normal 0 to 255 intensity values of Red, Green and Blue, that was used in 24-bit images (8 bit = 256, R + G + B bits = 8 + 8 + 8 = 24). In HDR, the RGB components are given a 32-bit floating point number, which CAN BE (not is) broken up into 24 bits binary plus 8 bits for exponent (base 2 or base 10). The idea is to represent ultra dark and ultra light values.. Such images can be used in two places, in lighting the realtime game, and in pre-rendering stationary objects (more precisely) with radiosity. There is also methods that the two can be combined such that the realtime component can make use of stuff deduced in the pre-rendering. Anyhow, it works, and it makes things look more realistic..