Texture is not enough for an object

So, I’ve downloaded the texture from blenderkit add-on. And now, concrete texture isn’t covering my plane fully and my plane is separated on small parts with this material. I’ve tried to apply scale of the plane, but nothing helps. Also, I’ve upscaled the material in texture coordinate, but this doesn’t work too and quality of material is felt down. Can someone solve this issue?

Hi, it looks like your mapping is not connected.

For a surface this large, you need a seamless texture that can be repeated across the model. This one doesn’t look like it can repeat, it hasn’t been prepared for it.

No, it connected. I made the screenshot in such a way that the link between the mapping and texture is not seen.

1 Like

Thank you! Your idea made me think about using procedural textures. They are endless and, of course, they can be upscaled or down scaled without missing of quality.

1 Like

Using a procedural texture is fine, just be aware of the pros and cons of each texture type.

  • Image textures take space in memory, because they store each pixel. However, they are very fast to render because there is nothing to be calculated, the renderer only needs to look up the existing data.

  • Procedural textures are made from pure maths, there is almost no data that needs to be stored in memory. However, this comes at the cost of render speed, because the maths have to be calculated on the fly for every pixel of the texture that’s visible on screen.

The performance impact of a procedural texture depends on its complexity (detail levels, number of textures used) and on which texture types you are using (noise, voronoi, gradient, etc.).

For some materials, the slowdown can become considerable. If this happens, you might consider using a repeating image texture instead.

There is also using procedural methods to look up and randomize image textures, as well as using complex/expensive generator setups (or external app to generate seamless textures) to pregenerate the image texture to be used with the first method. Using a single 4k/8k image and manually derive other values from that is an easy way to create details that are too complex to try to setup manually through nodes.

My projects typically contain a 4k and an 8k wooden texture for the detail level, and from those two I can create all the wood variations I need. No it’s not always “perfect”, but it has always got the job done without complaints. Obviously the approach doesn’t fit everything, but it’s a very valuable tool for the toolbox.

1 Like