Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

74 Excellent

1 Follower

About TWahl

  • Rank
    Advanced Member

Recent Profile Visitors

1,816 profile views
  1. I will definitely get back to you to discuss this, Josh - this was actually something I was thinking about even if finishing this media pack up will be quite a while away from now. I don't really know for sure how long "a while" will be but I definitely don't mind taking the necessary amount of time to create something portfolio-quality that will be available for the community in some form or another. I can't exactly stake a price at the moment though so this would definitely be something I would consider in the future when I have something a lot more fleshed out. As of right now I will probably just be working on remodeling my old texture maps and baking them with my current computer before I can get a hold of substance painter/nDo for PBR texturing. And that will come after getting and building a beast of a computer that won't choke and die when using that software (my first priority). I'll be keeping in touch and providing updates here when I can. Yeah it is really funny. The conventional approach I have seen everywhere is to use 3ds max to make shapes with turbosmooth, etc. but I always found this to be so finicky and it takes so much more time (and sometimes frustration) to do complex shapes with nice transitions. I like doing the CAD-based approach since I can always roll back features in a feature tree if I don't like something, and pattern designs really easily. There are some tricky aspects to it (have to pay attention to certain settings upon export) but overall I just like it so much better and I don't have to download 40 different scripts to create specific shapes that I won't be able to easily reverse later on. You might be interested in reading this article about the artist who did the weapons for Doom 4: https://80.lv/articles/building-guns-for-doom/ There were also some other artists who worked on the same game that used Fusion 360 for some of the environment props. I would check that out, maybe do a google search. Overall the program is really good for industrial models that have clearly defined shapes or object lines. Some people hate on using Fusion 360 for game art but I think it is because they have not had much experience with CAD software. I have many years of combined experience with Solidworks, Siemens NX, and AutoCAD that I have acquired through engineering school and also just doing personal projects on my own, not game-art related but more "I need to mock up a metal plate that will fit with X Y hole/thread pattern for this machine/bracket/etc".. The programs that work best for someone might not work the same for another person, so it really depends on the background of the artist and what you are modeling. I for instance would not use Fusion for modeling something organic, or with too many sweeping or curved surfaces (which oftentimes they don't even technically have any real mathematical geometry). There are many factors at play here, really. Zbrush and Max would probably be better for something like that, or you could use a combination of programs to achieve a good result. I am grateful to have that engineering experience though, because in a way I try to factor it into my designs to keep them semi-realistic - even if it might be science-fiction. Still really wanting to give Zbrush 2019 a try when I have the opportunity though - looking at the latest video demonstrating their new features. Glad to know, lol. I just like going into detail with what I do so other people might be able to gain something from my current methodology - even it it might be slightly unconventional.
  2. Yeah it was more or less a quick comparison I wanted to do just to judge how crisp the "major edge lines" would be at 4K compared to just 1K... What I am really interested in though is when I create diffuse maps, the chipped paint and other small, secondary details that make the texture more realistic will look a lot better. If you look more closely at the 2nd picture, the chipped paint looks a lot like mud on the wall - not very good in my opinion. Another advantage is that if I were to sell this media at a native 4K resolution (or higher, potentially), the end user can scale the textures to meet their performance or graphical needs. I might look into getting a trial of nDo just to screw around with some of these textures, but again I am currently limited in terms of my integrated graphics card (I will need to assess the requirements of the program carefully). I know I keep saying that but it is a huge obstacle in my progress still. But for now, experimentation is fine with me since I have other life priorities that I must satisfy before I begin to focus on this project a little more. Once I get settled down with good employment I will probably just start a new blog or thread to chronicle my updated progress.
  3. Quick material in leadwerks (with my old map, which thankfully still loaded!) Below, the wall is 1K in resolution while the floor is in 4K: And the difference in detail in comparison to the even older 512 textures: Overall a much better looking result in my opinion. I'll apply this to some of the other textures that I have made and see if I can make a new level when I have time.
  4. Baked maps from the 3D CAD geometry in Xnormal, here is a snippet where I have rather quickly composited the cavity, bump, and occlusion maps together to see some edge detail and shadowing. I will make a material inside of leadwerks in a bit, see how it looks under lighting, and post a comparison (before and after) with respect to the same texture I made two years ago so you can all see the difference. The resolution here is 2048x2048 pixels, but in the final version of this texture I would be duplicating these images so they were in a 2x2 array (for a total resolution of 4096x4096 pixels) before painting diffuse maps. You may notice that towards the corners that the "rounds" look a little bit polygonal. This was due to the fact that I was making use of Fusion 360's .FBX export option instead of exporting the highpoly model file as an .STL (as in, for 3D printing). At Fusion's current state, the program does not allow for manipulating .FBX mesh resolution upon export, so I just made do with it for now as a proof of concept. Ideally, I would have used the .STL exporter and upped the mesh resolution to its maximum (for optimum quality), but my laptop can only handle so much in terms of polygon counts and the assocated files took too long to import into/crashed other modeling programs because they were so large. Yes I am still using my same old "poor mans" setup with 8GB ram... and it really sucks. More later. Sorry if I am being a little bit too technical or if some of what I am saying doesn't entirely make sense - feel free to ask me for any clarification.
  5. After 3 hours-ish of work I was able to make this model based off my old map: I was able to achieve a much higher level of 3D detail also, as seen here: I will bake normal, height, and occlusion texture maps at 4K resolutions and put them into leadwerks as materials to see how they turn out. (Edited for clarity)
  6. The beginnings of my experimentation (currently just establishing sketch lines from which geometry can be extruded and further refined)" I am starting with a more basic floor texture that I can mirror in the X and Y directions after I am finished. This will save some time and is how a fair amount of my original textures were actually concieved.
  7. Hi Just updating since it has been a long time. About four weeks ago I finally graduated school with my bachelor of science in mechanical engineering. Currently job hunting, but I am looking to get back into this project. I will be revisiting my old files over the coming months (in my downtime) and playing around with some hard surface modeling techniques using CAD software (like I have mentioned in the past). I will be using Fusion 360 for this purpose and will be loading my old vector-generated heightmaps into F360 and using them as a guide for modeling. With this method I can achieve a much higher level of detail and not have to worry about artifacts associated with image resolutions. Thanks again guys. It might be a long time until I update again with any sort of new designs, but be assured that I want it to happen (even if it takes a few months).
  8. That will definitely be a plus, yes. But the main reason why I will need a better PC is the be able to run programs like substance, zbrush with many subdivisions, etc. That will allow me to produce much better and finer content in quite literally half the time (and with half the hassle!) The new LE5 renderer is looking good so far though Josh! Right now seems like a good time to wait to buy PC parts as well, since hardware seems to be fluctuating in price quite a bit at this time.
  9. Quite a bit going on in my life atm, so I have not been at this project in a long time. I currently still do not have enough saved up to buy a new computer to start pushing better graphical work (might not be that way for a while). My current setup at this time is a Dell laptop with an integrated chipset (intel graphics 4400 - not exactly a good gpu). Leadwerks does not really like to run on it... Hopefully I can get a new rig so I can get back to what I love doing! I will also be starting summer classes soon (will be finishing up university with a mechanical engineering degree in May 2019). Finished my final exams for the spring semester a week ago after tons of studying and a few gut wrenching all nighters - I went to the bars in my college town for a few straight days after to unwind a bit though . School comes first before anything else. In the meantime I have been doing a lot of research on PBR materials, texture creation techniques, color theory, and more. I have also been watching the LE5 updates closely, mainly I am waiting to see what the graphical capabilities will be in the new version of the engine (I am aiming to create some very realistic game environments). I really like where the engine is going and I hope it is going to be a game-changer for LE Hope all is going well with everyone!
  10. Fooled around a bit more and actually made something
  11. Would it be better if I started a community blog for this project? Or should I keep it here? Been Thinkin
  12. Also xnormal is freeware. You can download it online - it is widely used by many talented cg artists
  13. In this case I checked the box under the baking options in Xnormal to render out a displacement map from the high and low poly models I sent into the program. Same for the normal maps and stuff. It then renders it for me while I wait. Xnormal has options to bake other maps too like bent normal maps, curvature maps, convexity maps, even derivative maps as well but 90% of the time they are of no use for most applications.
  14. Did a little experiment today and successfully baked a solidworks model onto a plane (took me little to no time to make this funky triangular widget). I likely will not be using solidworks for any of this at all, however I just did this to see what would actually happen. Surprisingly the bakes came out great (the images are a little fuzzy because of the watermarking software I used). I baked these at 512x1024 in Xnormal without a lot of anti aliasing but I was more concerned mainly with looking for any glaring baking artifacts after importing/exporting solidworks .prt files. Screencaps from the solidworks window: The final bakes after re-exporting the .prt file (at the highest possible resolution) to a triangulated .obj. The model file was around a million triangles or so: . I am thinking about remodeling most of these textures and rebaking them all to be a higher resolution, to then be put into substance painter in the future (nice 4k maps with PBR as well as legacy normal/gloss compatibility). 4K to me gives a lot more room for detail than 1k and I will use my previous maps I made as guides for remodeling (should not take too long once things get going). This process can be used on game objects as well as textures (high to low poly retopology). On another note, finals just ended for the semester so I am on winter break now at uni! So glad to have that over with... Can't wait to get my hands on Fusion 360 in the future as the process is similar to solidworks although it is a lot more flexible and less clunky (in the case of creating hard surface models, not engineering parts or drawings). have a good one!
  15. Thanks! Yeah school has been the main priority for me since the semester started back in late august. I actually have a gumroad account @mdgunn so I will check those tutorials out. I will probably be downloading the free trial for F360 onto my laptop to get a feel for the software and how it works. MOI3d I have also heard of but I like sticking with autodesk products as they seem to have the most solid support.
  • Create New...