@Tristan: I studied computergrafics for 5 years. I'm making 3D art now since about half a year fulltime, but I had some experience before that. Its hard to focus on one thing, it took me half a year to understand most of the vegetation creation pipelines. For speeding up your workflow maybe spend a bit time with the megascans library. Making 3D vegetation starts from going outside for photoscanns to profiling your assets. Start with one thing and master this. @Maxime: The difference between my technique and Z-passing on distant objects is quiet the same. (- the higher vertex count) I would start using this at about 10-15m+. In this inner radius you are using (mostly high) cascaded shadows, the less the shader complexety in this areas, the less the shader instructions. When I started this project, the polycount was a bit to high. Now I found the best balance between a "lowpoly" mesh and the less possible overdraw. The conclusion of this technique is easily using a slightly higher vertex count on the mesh for reducing the quad overdraw and shader complexity. In matters visual quality a "high poly" plant will allways look better than a blade of grass on a plane.
Is this not like gear VR or anything else
Daniel McKay showed how he creates amazing realistic materials with Quixel Suite.
Hi all. First up, my name is Daniel McKay, and I am currently employed as everything from a Motion Graphics Designer and 3D Generalist, to Concept and Production Designer at a post-production company named TOYBOX. I have a background in design, having studied for 4 years and gaining an Honours degree in Computer Graphic Design. My first experience with 3D was learning Electric Image Universe and a tiny bit of Maya. On graduating, I did a mix of jobs in the television industry, but nothing of great mention. I decided to make the move to London around 2008, and the opportunities that arose on various freelance projects (as well as working alongside some amazing talent!) really pushed me to further myself and my skills. Some of the highlights were: a brief stint with Territory Studio working on Killzone 3 cinematics, working with the talented folk at SPOV, and an extended ‘permalance’ gig working with the fine folk at Flock. Throughout all these jobs, I dabbled in anything 3D, using Maya and Zbrush, as well as motion graphics bits and pieces, predominantly in After Effects. Since 2013, I have been based back in New Zealand (Auckland) and am currently interested in learning Unreal Engine and furthering my knowledge in real-time technology. If anyone’s interested, most of my latest works end up here, as well as more frequent updates on stuff/experiments there.
As anyone in the industry knows, the internet is a bottomless pit of resources and assets. Coupled with this, we always seem to be up against tight deadlines and the requirement to get that idea floating around our heads out onto the screen to see if we’re hitting the mark brief and concept wise. In my role at work, I wear many hats, with 3D Generalist and Concept Designer being just some of them. You have to be able to adapt to the brief, learn that the first idea won’t always be the right one, and be ready to have an idea that you spent hours on might get binned within a few seconds of a review. It’s for this reason that I’m always looking for short-cuts to get those ideas out and in the best quality possible. The use of the Quixel Suite has allowed me to generate some great-looking textures extremely quickly, and with little fuss, so that I can up the quality of my final concept with a small time commitment. Not to mention that by making textures seamless, where possible, you’re always adding to a library of assets that you can keep re-using for future projects. Win-win!
In this industry, I don’t think anything is ‘impossible’ for an artist to do. Research, time, practice, and commitment might make these things ‘seem’ impossible, but ultimately, if you put in the work, you’ll eventually learn the workflows and processes involved, and have those skills to apply to your future projects.
I chose this particular workflow as I’ve worked with the Quixel Suite in the past and have been extremely happy with the results I have gotten with seemingly little effort. It made sense, therefore, that I would see what else they had to offer in their extensive toolkit.
When looking for the texture, you generally have a pretty good idea of the sort of thing you are looking for. Whether it be a wood surface, dirt, concrete, fabric, skin, etc, you have some criteria you’re hoping to meet… Is the wood old enough looking? Will this concrete be broken up into slabs? Will the ground be made up of mud, stones, gravel? Etc, etc. I generally begin with a Google search and will start with the most basic and direct thing I can think of – For example ‘wood seamless texture’. Again, time is usually of the essence, so if I can find something that has been tiled for me already, it can save additional time for the important stuff down the line. Resolution can be a problem of sorts, but generally, during the search process, I’ll filter the results to at least be more than 1024 x 1024.
The resulting images will be varied, and with practice, you learn to get a sense of which images will work better than others… is there a strong contrast between lights and darks, making sure there are no shadows/vignettes on the image, are there any blemishes in the image that look good in one tile, but will be a giveaway if it gets tiled 5 times? Etc…
When I first started this project, I began by following a bunch of Quixel tutorials online. Whatever I could find, there’s generally something in there that will make my life that little bit easier somewhere down the line. Like I said, I’ve been using Quixel for a while now, but I always get that sense that I could be doing something easier than the current approach I know. It was during this process that I would watch tutorials and follow along myself, trying to figure out the quickest and most efficient ways of doing something.
For my first attempts at using the NDO texture generation approach, I started with a brick material that actually comes standard in DDO. It was a great way to start with a texture that I knew would be able to give me what I wanted detail-wise and to learn how to generate the required textures from NDO.
Once I got a handle on how to generate a decent result from an image, I started to explore the internet to see what I could achieve using nothing but stock imagery.
It’s probably worth noting here that the imagery I download from the internet is used only for concept purposes. When a concept is approved and we move into production, these textures would either be purchased, sourced from a resource allowing us to use them in a commercial production, or generated from scratch and attempting to replicate the concept as close as possible.
I guess the best way to explain this would be to do a mini tutorial of sorts. Here we go…
In this instance, I’m going to look for a pebble texture. I found this image using the search method mentioned earlier. It has a 3151 x 3151 pixel resolution, which is plenty of detail to work with. I usually work at 2k (or 4k in some instances) and then output the size required. For this example, I have reduced the size to 2048 x 2048. Also, you can see there are no shadows, there is a nice contrast between light and dark, and it is already tiled.
Sometimes, an image can say it is seamless, but there could be a blemish on the texture or frame discoloration that can cause obvious seams on the edges. I usually do a quick test in Photoshop doing the following:
Under Filter > Other > Offset… you can offset the x and y position of the texture and see if it is actually seamless, or not. I usually half the resolution in x and y so you can see it clearly.
In this case, I did half with a 1024 pixel offset in x and y. If you have the ‘preview’ box checked, you can see it update automatically, and check for anything that might need tweaking. This one looks fine!
With the image loaded in Photoshop, I start Quixel and access NDO. I use the drop-down menu and select the Photo-Normal Presets. You can also generate Normals from an image using manual features, but again, I’m just looking to gain a bit of extra detail out of an image for concepting purposes.
These presets cover a wide range of possibilities. You might find you try a few different ones to give you a good headstart, but in this case, I’m looking to produce some rough stones. Make sure ‘Is Tileable’ is checked, and you can select the ‘Active Doc’ button to get the ball rolling.
NDO does a pretty neat job of generating some nice normals straight off the bat, but you have a range of sliders you can toy with to generate the look you are hoping for. I use NDO in conjunction with 3do so I can get a rough idea of just how the normal is going to look.
Once you’re happy with your Normal map, flatten it and save it off.
Now that you have your Normal map, you can use this to generate a bunch of other textures you might feel necessary to help achieve the look you’re going for. With the Normal map flattened and open, select the ‘Map Converter’ option in the drop-down menu of NDO.
Make sure the ‘NORMAL’ button is pressed on the left-hand side, then select the map you’re wanting to generate. Again, you will be given a range of sliders to adjust the effects for each map. I recommend using the 3do window to get a better idea of what you’re going to end up with. I created a Spec and a Height Map from the Normal map to provide me with the following:
So now I have an Albedo, Normal, Height, and a Spec map. The Spec map can be used in a number of ways using levels/curves in Photoshop, as well as inverting the map, to create additional maps such as Glossiness/Roughness. Again, I’m not looking for production-ready textures with this process… rather, achieving an idea that will help sell my concept.
Here are some renders from Marmoset Toolbag showing the outcome of these maps plugged into their necessary channels:
I think I covered this briefly in the previous mini tutorial, but just to recap… Quixel has the beauty of seeing live updates in 3do as you adjust the sliders for your Height Map. Agreed, Height maps can be somewhat tricky to figure out, and there’s usually a bit of back and forth to tweak the textures to get them to your liking. It can also depend on how you intend on using the final map, where my weapons of choice for such things are usually Marmoset (for an asset, prop, or singular detail concept) or Unreal (for more environment or set based concepts). Height maps rely on the poly count, so again, while super dense models will allow for more detail, this comes at a cost of performance. Concept-related usage is fine, however, as you’re usually looking for 1 – 2 key still frames to work up.
Using materials for games
It’s possible that such materials could end up in the final production, but ultimately I think you’d want to give them more love than the quick preset techniques I used here. As I mentioned earlier, these materials are more aimed at concepts than the final, polished piece, but I’ve been blown away by how good some of the results have come out.
These maps are great for tiled areas, where you will potentially end up doing paintovers over the top anyway, during the concept process. They ultimately just add a nice bit of depth to the image… maybe it allows the capturing of a bit of light or something, that helps capture the mood of the image you’re creating. You can see from the examples I rendered from Marmoset how much more they pop when they have been enhanced.
For smaller objects, I would tend to use a rough UV layout and then use Ddo to generate a specific set of textures. This way, you can vary up the materials on one object a lot easier, taking into account the varying normals, roughness, and gloss of different surfaces. This NDO technique is a lot better suited to single substance style map generation.
As with any jump into learning a new bit of software or workflow, start small. Following tutorials and getting your head around the simplest of features will all be beneficial to you later on. Quixel offers a bunch of free tutorials to follow along with and learn every step of the process from start to finish on several projects.
The only way to get faster in my experience is to ‘do more and fail faster’. In this case, you start to narrow down potentially successful textures a lot faster by looking for some of the key criteria I mentioned earlier… contrast, shadow, seamless edging, etc. You’ll also learn some key presets that become your go-to first attempts. And also, how to use the textures most efficiently once you finally have them in your package of choice.
Again, the more you do this, the bigger the library of assets you can begin to collect. There are many resources out there these days that offer substance, material and texture libraries, but these can still take time to break down and work out how to apply them to your project. The beauty of having your own arsenal of assets is that (hopefully) you can remember the ins and outs of how they work!
Again, if anyone’s interested, I post my completed projects here.
I post regular updates and tests to my Instagram account.
And, if anyone has any specific questions, or feedback on how I can do my stuff better, let me know on my email.