Need help with creating maps for the Displace modifier

[Anonymous]

Hi, I'm looking for some info on how best to use the max precision of my data as a map to displace geometry. I have raw data that ranges from 0-1 with 16-bit precision in a huge array. My intent was to create an image where each pixel in the image corresponds to a data point in that array.

At this point I'm unsure of how to store the values in the image itself. Should I be storing 16-bit floats in a single color channel? Should I store 3 copies of the 16-bit floats in each channel of a RGB image? Should I be storing ints 0-65535 instead of floats like .017368 etc?

So yeah, I'm really just not sure how to go about creating the appropriate image to use as a map. I should mention that I'm currently creating PNG files, though I suppose I could create tiffs. I'd prefer PNG simply because I already coded for creating PNGs.

Thanks!

[leeminardi]

I do not have much experience with this particular question but here's my 2 cents. I would lean to use an integer database rather than floating point. The general rule is to use integers to count and floating point to measure. With your application you have a discrete number of steps (colors) so using integers is more appropriate. In addition, at some point you will need to convert a number to RGB values.  This is more readily done with integers where you do not need to be concerned with roundoff error.     It doesn't sound like an alpha channel needs to be considered so an integer format ( +/- 32767) should be sufficient rather than a long integer (+/- 2 billion). 

You might find a better and more informed response to your question by posting it on the programming forum for 3ds Max.