Author Topic: New user with a few questions  (Read 45 times)

I just purchased B2M the other day in hopes that I can create some textures or materials from photos that I will eventually take myself. I have been playing around though with images found from the web and I have a few questions.

1. If the image is 1024x1024 and I change B2M to 4096x4096 the perceived quality isn't really that much different. I suppose that is because my input image is of much smaller resolution. To counter that, can I increase the tiling amount by 4 to upscale the source image to 4096x4096? Since we can force random tiling I don't see the issue, however the quality would not be the same as a 4k source image correct?

2. I understand we want to remove as much shadowing as possible from the image so the renderer we choose to use(UE4, Unity, etc.) can use its own lighting to create that data itself, so with that being said are there any reasons to even work with the AO settings? To me I would just choose not to export that and let the game engine handle that itself. When would I want to keep the prebaked AO or tweak the settings? I think I'm getting lost on why we remove that, and then add it back in.

3. The height maps seem a little tricky to work with. For example when working with an image of small pebbles or dirt, when I try to tweak the height map it just turns out looking like a million small pin pricks across the surface, especially if the texture is mostly dirt cavities with grass patches(like someone shoveled out a section here and there) I cant seem to get the height balanced properly. To that note should I not expect to have each pebble showing individually on the height map, should it be toned down quite a bit to only give shallow depth to the pebbles? Larger rocks are easier since they take up more space on the height map they can be tweaked a little bit better.

So with that being said is there anything I can improve on workflow wise to help me generate higher quality images? I am more than ok importing the images into substance designer for further work, I just need to get an idea of some of the tools that might be helpful there that B2M doesn't provide.

The training materials are extremely thin when it comes to B2M. While the program is pretty straight forward I would love some more in depth tutorials taking a source image all the way into a game engine to show the final result.

Last Edit: May 19, 2017, 03:40:25 pm