26 replies to this topic

[Brucey Mc]

If you are like me and find the terrain painter in Unity a pain in the you know what when painting large areas I have been messing about with editing the Splatmaps externally and then re-importing into Unity.  Forget the actual example textures used and it needs a fair bit of touching up but I think generally it works pretty well.  There is a little bit of normalisation around some of the objects (the bright bits) but these can be cleared up with a little gentle brushing.

[Kablammo11]

Quite ingenious, Brucey. But CF Shapes (green, fairways, semi, roughs etc) are a separate thing entirely and will not be governed by splatmaps. So in the long run, splatmapping will only be viable for the more distant terrain shapes. And it's still useful, of course, for quickly laying out a golf course and allocating textures to the topography for later upgrades.

I still recommend you try and get comfortable with the terrain painter. It can be quite irritating at times, but that's mostly becaue it takes a little bit of getting used to - and it most certainly will give you a lot more control over many texture details once you've made your peace with it. The textures look very nice, btw.

 

We all could easily think of a dozen or more features of Unity where we'd wish that there were a quick fix, an elegant and effortless way to gets things done more easily… alas, there isn't. Sometimes some Unity stuff simply ...sucks. Don't fight it, don't shirk from it neither, and it will get a lot easier a lot quicker. Just a matter of mileage.  

[Brucey Mc]

Thanks K, I know that the CF shapes are different but I am just trying to create a base terrain that can be used as a marker for my shapes.  All good fun.  I have used the terrain painter and I am ok with it in general but trying to create larger areas I prefer this way of doing it.

[Brucey Mc]

After all that hard work, I'm off for a short break.  Might just make it a busman's holiday...... 

[shimonko]

I've being doing the same Bruce and agree, it's the large areas which the real time-saving comes in, such as painting in all the rough.

I would have loved to do this for all my terrain painting but couldn't replicate Unity's ability to automatically do the normalization (to avoid the bright spots). Well I could for the first three terrain textures but I wanted to mix up to 8 (two splat maps).

[Brucey Mc]

I'm using 4 splatmaps so far, the trick is to have one photoshop (or equivalent file) with each terrain a different colour,  then select each layer by colour and make as a red layer, blue layer or green layer for each splatmap file.  I use a texture that will only be used for touching up for the alpha layer as I couldn't work out how to use the alpha layer properly.  All unused parts for each splatmap in the alpha layer should be black.  Sorry I don't think I have explained that very well.

[Brucey Mc]

One more shot......

But this seemed the natural place to start.......

[garynorman]

I haven't really looked into splat maps at all, but what you are saying is that you can basically draw the course in say photoshop and then transfer it, with textures, into Unity??

 

If so that sounds promising...

[Brucey Mc]

Yes gary, pretty much 😄

[garynorman]

Good news then!!

 

Is that Turnberry you're starting work on there??  I hope so as it's a great course   I have a basic lighthouse model if you want it...  Let me know

[Brucey Mc]

Going there on Sunday for a couple of days and a trigger happy camera 😄

[shimonko]

That was my hope Gary but Unity demands the combined 'strength' of all textures at any one spot on the terrain add up to 1. So if you paint with the texture in slot 4 in Unity, Unity will automatically scale down the strengths of any other texture(s) at that same spot so the 1 total is not exceeded.

 

Photoshop can do this fine if you've only got 3 textures (just paint with either red, green or blue), but the 4th texture is represented by an alpha channel in Photoshop and painting this won't reduce the intensities in the other three textures channels to keep the result at 1.

 

For textures 5-8 in Unity which I had, you'll need a second Photoshop file and painting in it (of course) won't reduce the intensities of textures 1-4. A custom Photoshop plugin I suspect would be the only efficient way. 

 

But as Bruce says, there is still merit doing this for some things, like painting larger areas and touching up the overblown areas.

[Brucey Mc]

I created four splatmaps from a photoshop file and had them in unity as shown in the pictures in about twenty minutes so it can be a great time saver

[Mike Jones]

I'm just wondering, what are the perceived advantages of doing it this way instead of inside Unity?

 

Let's think of it this way, think of your unity course/scene as the canvas. So you're painting a picture and normally you use your brush on the canvas and you can immediately see how things blend with everything else and interact with the lighting, slopes etc. What you're doing seems to be more akin to painting your picture on another canvas and then 'porting' it over onto your real canvas.  

 

Am I missing something or is that counter intuitive and unproductive? Is this more about your planning stage rather than any final artistic work?

[IanD]

It's different...

 

I'm wondering if it is that familiarity with everything photoshop... ie icons, relative position of them and grouping of everything familiar. Whereas with Unity, there are simple once you become accustomed to them. Maybe it's not that at all...

[Brucey Mc]

If I was starting a design from scratch I would use unity but when you know where things are initially it's a quick way of putting an image down of where things are. There are a few ways to skin a cat

[Kablammo11]

My way would be to draw a Photoshop layout, see image below, and then import it as a terrain-sized texture.

I'm not advocating this as a better way to prep a project, just describing how I do get my cats naked. Ah, those sweet, purring, pink-fleshed, bald little critters…. hm… they look so stupid.

 

[shimonko]

I'm just wondering, what are the perceived advantages of doing it this way instead of inside Unity?
 
Let's think of it this way, think of your unity course/scene as the canvas. So you're painting a picture and normally you use your brush on the canvas and you can immediately see how things blend with everything else and interact with the lighting, slopes etc. What you're doing seems to be more akin to painting your picture on another canvas and then 'porting' it over onto your real canvas.  
 
Am I missing something or is that counter intuitive and unproductive? Is this more about your planning stage rather than any final artistic work?

Not planning or final for me, rather just a shortcut when I would find myself otherwise painting large areas in Unity to match the overhead imagery. Most of that of course will be eliminated with CF's splines, but still I use my own terrain shaders for the rough and importing splat maps here saves a lot of boring blocking out.

[highfade]

Splatmaps? Never heard of it.  Thank you very much, another thing I know nothing about.

[Brucey Mc]

Highfade, select your terrain in your assets folder and you can drill down into your splat maps. These are generated by unity