Hi Rachael, It sounds like you are running into an issue with data type in your Band Math expression. If your input data and values in the expression are integer data type (versus floating point values), then your expression will be calculated using integer math, which won't allow the output to be anything other than integers. Which is going to end up looking like nonsense. In this case, you really want to do the calculation using all floating point math, so that you can get floating point (decimal) values in the output. The easiest way to indicate that in your Band Math expression is to change it so that the integers in your equation are all floating point values, by adding .0 to the end of them. So it will look like this: (2.5*((B2-B1)/(B2+6.0*B1-7.5*B3+1.0))) Does that give you the values you expect? When manually entering wavelengths in the ENVI header, I think it is best to enter the center of the wavelength range for the band. ENVI doesn't really care what you use. It will interpret and report the pixel values as being for whatever specific wavelength you supply. Mainly this is used when viewing pixel spectra, because you need to know which wavelength to assign each pixel value to. With Landsat data, you don't have measurements at 531 nm and then again at 570 nm. Instead you have one band (the visible green band) that covers that whole range, from about 520 nm to about 600 nm. So the equation for the PRI isn't meaningful in that situation. Regards, Peg Exelis VIS

Hi, I believe the help is incorrect where it says, "Values range from -1 to 1." I would expect that range with the normalized difference indices (x-y)/(x+y), but not with the EVI equation. The ranges that you see should be correct. We may need to add some details to the help on how to interpret the data values of EVI.