Posting this a day late because I came home exhausted yesterday after a huge list of changes to my plans and woke up at 2am. I wont go into much detail besides mentioning that I am currently writing up a proposal for graduate studies at chico and I have about a week from tomorrow to finish it and turn it in. On top of that I will be in SF for GDC 3 days after that. So very glad I decided to postpone my throne room until after GDC.
Last week my portfolio finally went up, after a year of WIP versions. Honestly I am still not quite happy with the current one, but from here on out I will be doing small edits to the layout instead of taking it down and putting it back up. I want the ability to apply for a job as soon as the posting goes up, and that means having a well maintained and populated portfolio.
On top of that my business cards came in a couple days ago:
As for work, Osman and I started wondering about creating a somewhat perfect wood graph in substance. Wood is one of the more challenging things, in my opinion, to pull off procedurally. I’ve only got about an hour or two worth of work into is so far but I am liking the results. Osman has been working on his own version and whenever we get to a point that we like we have been sharing info. Most of the resources we have found online have been either lacking in quality or lacking in information.
I have some potential solutions for making the knots better, along with making the grain more believable and wood like. I just need to get back in and work on it.
I’ve also been working on some more tile, this time focusing on flooring. Inspired by the amazingly well cleaned floors found all across the Chico campus.
To add onto these tiling textures I have been working with substance in another way this week. The first of witch is creating a brush for substance painter that creates welds. Right now they’re a little too clean around the edges and I want to add in some exposed values giving you more freedom to control what types of welds you want to do, ranging from clean to messy. But it is a start.
The other one I haven’t quite taken into painter yet because I am working out some kinks but it is a paint drips substance effect made for painter or even just for some masked images. I have been planning on playing with PopcornFX so I might try to just create it all in substance alone, or just base it off of a current effect like leaks, but I like the idea of being able to use it in substance without painter involved.
This weekends focus will be finishing the throne and its’ surrounding area’s high poly & low poly models. I spent a large majority of last week getting ahead in my classwork so that I would have a week or two to work so I don’t really have much to show. The main improvement I have is that I’ve finally settled on my main design rules for the environment so moving forward here should me much quicker as most of my delays have been creative blocks. On top of that I am no longer sick so I have no more blocks there.
Another thing I worked on this week in my free time was taking this leaf generator network and putting my own touches and improvements on it:
My goal when I am finished is to have a pretty decent leaf node network for generating them and then taking that and applying it to a forest floor like tiling texture. I am still about halfway through that so I will be posting it in a later blog.
I guess it is pretty much habitual now that I will be making a 2nd post either Sunday or Monday along with my polycount post so expect that. This time I will be posting the updated version of my website along with the blog so there is that to look forward too as well.
Unfortunately didn’t get much done this week for modeling due to putting in a lot of work on my new portfolio site, resume and dealing with start of the semester type things. But I have the weekend to commit to modeling and will be posting an update with the final blockout sometime Sunday, which will be the same day that I update my polycount thread. Also my new site will be up sometime in the coming week, after that I will be updating my business card and ordering those for GDC. My current ones are fine but I have learned some good tips for making a card specifically for events such as GDC. One of those is adding a headshot of myself and a section for notes on the back of the card, that way my card feels more personal and people can write any note about my they want when they take my info. Another issue with my old card is that the model I threw onto it was my just finished character model from last semester, the problem is that I am not a character artist, and my card understandably confused people. Good thing I wasn’t looking for jobs yet I guess.
As for modeling: I am playing with a lot of basic forms tonight and will most likely be doing so for the next few hours, I just need to get this post out there as an update. I feel like my last environment’s main flaw was a lack of a solid blockout. The current blockout I have now is about the same level of detail, if not more, than what I spent on the old environment. Because of that, last time I fell into a horrible trap where I was half way done with the project, and didn’t like where it was at, or have any clue how to fix it by adding more. I had to basically cut my losses because of deadline and do the best I could with what I had. This time around I am working on an even tighter deadline, but I have many more hours per day to work on the project so I think over all time will even out.
I am really enjoying where this environment is headed though. It will also be the first environment that I don’t really base off of someones concept and then adjust slightly. This one will be mine from the ground up so my blockout is sort of concepting things as well.
From here on out my plans are to finish the blockout (this weekend), make high poly meshes (either in zBrush or Maya depending on what is needed for the mesh) for the base environment (this week and next week), create the lowpoly and bake out, take things into nDo as needed(this week and next week), then Substance Designer or Painter and possibly photoshop if needed (this week and next week), throwing them into unreal and testing lighting (this week and next week). Then onto taking the individual small props from blockout through that same pipeline (next week and the week after).
Part of my goal this semester is to have a get set of Tiling Textures for my portfolio. Usually when I look into what studios are looking for when it comes to environment artists they all mention tiling textures. What a better way to start that series than with some tiles?
For this first one I purely used substance designer. Future ones I am going to incorporate sculpting and possibly even some hand painted textures.
Besides this I wasn’t able to get too much done because of the Global Game Jam and it messing up my sleeping schedule. But I did start a really basic blockout of my Throne Room scene. I am still playing with the overall layout and forms that I want the rooms in so it isn’t very impressive to look at right now. Either today or tomorrow I will be posting everything to polycount and linking that on this blog.
I’ve discovered that Maya 2015 can have an issue with batch rendering contours for things like wireframes. When I went out to find a fix I couldn’t find much at the time. So instead I poked around until I figured out a quick solution.
If this solution doesn’t work for you I suggest exporting your file out as an .obj and reimporting it back into a fresh maya scene.
Below you will find a video tutorial with the fix, followed by a quick image of how to setup your shader group for those a bit more knowledgeable.
I did some tests with TURTLE vs xNormal the other day. As much as I want to like TURTLE it just doesn’t hold up when it comes to the quality of xNormals AO bake. The normal maps are pretty similar so that there is pretty much no difference. But the AO has numerous issues, ranging from the max color value not being represented where it should be, to it ruining a tiling sculpt because it seems to have a very slight gradient color change from left to right giving you a seam between tiles.
Where TURTLE excels is in bake time. About 2 minutes to bake a 2k normal and AO map is pretty impressive, and if I were making something that tiled would be pretty acceptable. I could always go in and play with levels and such in photoshop to fix the max color value of the AO, but fixing a subtle color change is more work than I want to do for a subpar bake.
xNormal on the other hand took close to half an hour, but the bake was excellent.
Another method, the one I ended up going with, was to bake my information in Substance Designer. I ended up getting a similar quality bake in the AO but with different details emphasised (the normal map wasn’t half bad either but I prefer the one created by xNormal or even Turtle for what I needed). I then took my bake from SD and multiplied it over the one from xNormal. This allowed the quality of both The bake times were similar to that of xNormal, so I wouldn’t suggest this method for quick work since it is close to an hour.
At this point you have a good base for your texture work. I will be updating some of that on Wednesday most likely. I need to go in and compile screenshots.
For the longest time I have focused entirely on environmental and prop modeling. Here and there I have made animals and characters, I understand Edgeflow for bodies and faces, as well as the other complexities of being a 3D Character Artist, so the workflow isn’t new to me, but I have never created anything Reel qualit. Most of the characters I have modeled and textured were used to teach myself something or to keep my skills at organic modeling up to date. But I have always wanted to focus on characters. Like my previous post about how excited I am about working with a PBR workflow for a project this semester this post is to express similar excitement for my ability to work entirely with character art for a class.
Finlay (my name for him) is a concept created by Alexei Samokhin in 2011. I took the character as is and created some orthos out of him and at the same time created a backstory/personality for him to help flesh out the turntable for my reel. The most immediate thought about this guy was that he was an oafish sidekick of a hero. Lovable, reliable and protective, but a bit clumsy while at the same time being a strong fighter.
The next step of the process, after establishing the character and creating the orthos, was to decide how they would be textured. For this I gathered some references and threw them together into a ref-sheet that I can keep open while I work.
Once that was all done I got to modeling. I approached this model from a more traditional workflow. First I created a base mesh in maya, focusing on topology, with extra attention paid to the facial area.Once that was done I separated the mesh into parts that would be skin, and parts that would eventually become clothing. At this point I got a tiny bit ahead of myself and decided to see what the final textures might look like in a preview state. I separated the arms and face and threw them into Mudbox. Here I sculpted some details and did some quick suggestion of muscle groups on the arms and neck. The character himself has a bit of a soft look to him instead of rigid well defined muscles so this worked out very well.At this point I tried throwing it into substance painter, but having recently updated my drivers this was impossible due to technical issues (it is working great now!). Instead I baked out my maps using TURTLE in maya and threw it back into mudbox and started using the paint tools there.
One thing that helps out a bit was using the blur tool to fix any weird issues with the normal map in areas like the tips of the fingers, that don’t bake very well when going to a much lower polycount.
Right now all the textures are in a temporary state and will remain so until the rest of the character is modeled, sculpted, and baked out for painting.
As a bonus: While I was sculpting the head I found these great reference images on Polycount showing off the planes of the face.
Or at least that is the way that PBR makes me feel. As a 3D artist working in games PBR is something I have been devoting a large chunk of my time to. Sadly I have had no projects focusing on using a PBR workflow since most of my work in the last few years has been for Unity and focused towards mobile.
This semester I am finally able to create some PBR work, across multiple projects. I will be working in UE4, and using Maya, Substance Designer/Painter, Photoshop, and Quixel’s new suite.
I also plan on going a bit more stylized with the look, partially because I love stylized texture work, but mostly because I feel this is an area that has yet to be fully explored with PBR and would love to work through it. When most people think of this workflow they think of realism, in fact there was a lot of people wondering if stylized/hand painted looks would even be all that possible.
Both of these are great resources for starting out with PBR and gaining an understanding. Once you have your footing however I suggest you take a look over at marmoset’s website since they they two great write-ups on the subject.