I haven't posted a still shot for some time.
This is a workup of the Malaspina glacier in Southeast Alaska from USGS DEMs, and a NASA satellite photo of the same area as color (image map geotif).
As you can see, the sat photo gets a bit blurry close up. It does look great from altitude though. I would like to get some shrub populations in the darker area on the moraine.
So, I have a question: Is there a way to use color separations for distribution masking of population? I copied the sat photo and made a grayscale tif of just the darker areas with all others transparent.
I've been trying to put it in as a mask but it just doesn't do anything. I am wondering if I have to go through some blue node to extract the scalar level of gray to affect a mask.
Way over my head in this.
Cheers,
Russ
If it's grayscale it should represent the amount of distribution you want in each area for 1 or more populations (each would have the same distribution unless you use a more advanced approach to split up the spectrum). So assuming that's the case, white would be "maximum number of objects given all other restrictions on the population", and black would be no objects. I assume you know that, just making sure. So then the question is how are you attempting to use it as a distribution mask?
- Oshyan
You could try adding a few color adjusts under the image, and adjusting each so their subtraction would yield the grey area you need. But I would take the easier way to open the image in PS and draw the area you want populated in another layer, and save that as greyscale mask, with (of course) same location and dimensions. Or if you want more masks for several pops, use an RGB image and put the masks in the channels, and separate in TG again (blue to scalar, green to scalar, red to scalar).
Thanks for the input.
I have tried to use the grayscale tif image that I made from the dark colors to mask a surface shader, not much joy there either. It textures nothing as a mask, and everything when invert mask is checked—no variability in distribution like a painted shader does so well. That's really what I am trying to do—just have it plugged into the network like a painted shader.
The darn tiff loses its georeferencing when I copy it. I'm having to brush up on Qgis to re-georeference it.
Every day is a lesson :P
r
You could also do a topdown render from a second camera above the area you want popped, keep it there, paint in PS in your (low quality if needed) render and project the resulting mask through the same camera. That's very easy.
yes, a painted shader is my fallback option. I just wanted to find a way to use data from actual satellite images for populations or shaders.
Makes me wonder if there is a way to copy/paste a raster image into a painted shader. That would be cool!. I think I'm just missing something that's all.
Thanks again for the tips Ulco.