Had some great ideas with this, as I generally do. I had a model of a rolling shutter door I'd made back when I interned for Piranha Bar last summer.
I've been looking into learning more rigging so I thought this might be a good start. A rolling shutter door, it just rolls up, shouldn't be to hard to path that, right? Turns out I'm not a great rigger. I initially attempted to use the Wire deformer as a means for pathing the pieces of the rolling shutter door. I had looked at a few tutorials but I couldn't find one that used 2017, my current version of Maya. Which I've learned is a big problem with Maya because the tools are CONSTANTLY changing. The wire tool in 2016 seemed to work differently in 2017 and sadly I couldn't figure out the gaps and it simply refused to anything useful for me. I definitely wasn't getting the same results, and Autodesk themselves aren't always the most reliable source of useful information. Maybe the wire tool wasn't the best course of action but I really think I just need a bit more education in rigging. The result above is actually something I managed to create using a MASH network. A handy little system that's been integrated into Maya and actually has some really fun features in Maya 2018. It got me this far using a curve node and a distribute node but I couldn't seem to really solve the issue with the weird sparse distribution and I had a little 'foot' at the tail end of the shutter door that I couldn't find a way of timing properly with the rest of the shutter door. Speaking of timing, MASH doesn't seem to be a tool that likes the idea of rigging. It's fun for simulation and stuff but for some reason it can only be controlled by the time (I guess because it's a simulation tool) and I couldn't really find a workaround. I know a lot about Maya and it's various in-engine relationships but evidently not enough to experiment. What I really need I suppose is Brian Horgan's bhAnimPath tool.
0 Comments
Well, I sure had some bright ideas for this one. The world was at my fingertips, I could achieve ANYTHING. Wow, what a mighty fine idea, what a project this one was gonna turn out to be. So I was thinking about 'high to low poly' workflow ideas, after making the spaceship you can see on my main page. I was wondering what other practical applications such a workflow could have and got me thinking about environments. Usually big hefty spaces with a lot of detail but also encompassing a huge space. So I made a plane in Maya, gave it 1 million subdivisions, my computer struggled a bit with that. I did a quick google for an alpha/depth map of a mountain, got some satellite footage someone had turned into a black and white image and plugged into the 'stamp' for the sculpting tool in Maya. Using that I slowly pulled some detail into the 1 million poly plane. Like, reeeaallly slowly. Eventually I had something that I thought looked like a pretty decent mountain range. You can see the absurdity of the mesh in the very top image up there. Once I had that done, I brought the crazy high poly mesh into Zbrush, made a duplicate of it, and Zremeshed it, giving me an almost exact replica of my moutnain range but with only 9 thousand polygons. Much more manageable. I unwrapped both in UVs and attempted for hours to get the UVs to sit in the same UV space. After hours of faie attempts and computer crashes, I finally gave up and decided to see what would happen if I just brought the fucking thing into Substance Painter. I pulled in the low poly mountain and asked Substance to bake the normals and AO using the high poly mountain as a foundation for that. Miraculously it worked, or so I thought. The maps affected the low poly model and gave me what LOOKED like the high poly model as shown above. Perfect. I slapped on some procedural 'mossy stone' material onto the mesh, played around with that until I was happy and then did a "quick" render because I was pretty happy with hwo things had turned out. You can see how beautifully it all came together up above. After all my slaving away and troubleshooting, I was oh so pleased to see my work finally coming to fruition. One thing left; bring it all into Maya. I exported the few maps I had, opened up the scene in Maya and applied all the maps through an Arnold 5 aistandardsurface material. I looked upon the fruits of my labour and I .... what... how? no!..... Fuck.
Robots, robots, robots. Do you think I'd ever get sick of working with robots? Well I guess we'll have to keep waiting to find out. For now however, I cam across some really cool robot concepts that I wanted to try modelling. I've featured one of the concept pages above but there are more you can view on Artstation. They're by an artist named George Brad, you can view his work here: https://www.artstation.com/artwork/4YvVk
There are a whole bunch of cool concepts there that I'd like to get around to trying. Especially the one with the cloth hanging off of it, or the one with no lower abdomen. There's a lot of really interesting shapes there. This orb encased one kind of caught my attention though. Initially when I started modelling I didn't feel like sculpting a human head like in the concept but after seeing how my gammy replacement that I made up out of my head turned out, I realised that I probably will sculpt a simple head because the dumb one I went and made just really brings down the concept as a whole. THIS IS WHY YOU WORK FROM REFERENCE, PEOPLE! See what happens when you diverge from the original concept? And I did have to do that out of necessity for some areas, when you look closely at the concept it stops really making sense and it's just like "What? that's that's just a line, it doesn't go anywhere" so I have had to use a little initiative to make sure the model makes sense. And I want it to make sense (including the floating arms, leave me alone, it's magnets) so I've built a piston I plan on sitting into position wherever a piston would be needed. I also plan on rigging this robot, I think it would be really cool when rigged and I could set it in a more natural position quite easily then. It would be nice to get some more rigging practice in but I maaaay have to wait until I can upgrade to a better workstation. Currently I'm working off of my laptop and it does not quite have the power for rigging. Minesweeper, maybe, but riggng? Well it's getting pretty slow. And I need a new computer anyway to get back to that sweet, sweet, photogrammetry. Mmmmm... photogrammetry... Anyone who knows me well enough is probably damn sick of looking at this damn octopus shape. I've used it for so many things for so long now, this window probably being the most complicated. I made the base shape for that octopus so I could imprint it on a tankard, you probably saw it on my main page. Funny thing is, the original Octopus mesh, is a total mess. I modelled it ages ago when I first started to learn modelling and it's actually reeeaaallly basic.
But that hasn't stopped me from bringing it on to do great things. I'm no lighter, I prefer sticking to my modelling guns, but we dip in and out of other areas, and some part of me seems obsessed with volumetrics, and GLASS. I keep trying to play around with glass in various 3D applications and they ALL HANDLE GLASS DIFFERENTLY. Seriously, I've struggled so much with getting the way Substance Painter looks at glass to align with how Maya looks at glass. At first glance it looks like I got things working in the pictures above, but the textures in Substance Painter were, well, actual textures. Above it all just seems to look like shaders, there was dirt and stuff on the glass but I don't know where it's gone. Alas, back to the drawing board. I'll stick to my modelling until I foolishly attempt to make something out of glass again. Oh well, you gotta learn somehow. Can you tell what all of that is up above there? Those are photographs. Hundreds and hundreds of photographs. Of rocks, of grass, of snow, recreated in 3D. If you can see it, you can build it. The wonders of photogrammetry. I really love photogrammetry and the uses it provides for 3D work. I still need to practice it a lot more and learn how to really utilise the textures and geometry I get out of the tools to create something of actual substance.
The above images are just some random pieces of nature I decided to recreate for practice, and to figure out the workflow for photogrammetry. I have a pretty good handle on how to get the geometry into Maya, I just now need to figure out how to best use these results to go further with workflow. Anything I want, if it exists in the real world, I can recreate in my computer. This was an idea I had back when year 2 of college was starting and I had a lot of time on my hands to play around with my own ideas and projects. I wanted to model an airplane and paint that shark decal on the side like a spitfire. I could have made a dinky little DR1 WWI plane but I thought a Boeing B-29 Superfortress was more interesting. Big, beefy, bombing planes from WWII.
However, I never got around to finishing it. College got busy and I never had time to get back to it, which was a shame as the model was going pretty well. I never saw the use in getting back to it, I had more down to earth modelling to do, but it was pretty good modelling practice at least. I still want to paint a shark on the side of something though, maybe I'll make a shotgun or something, that would be cool. This is the first ever scene I put together when learning 3D modelling. None of the models are UV'd, the topology probably has loads of problems, a lot of the shapes are super basic and boring, it's crazy how much you can learn in two years. Seriously, the most complicated model in here is really tiny and hidden by poor layout of shapes, try find it.
Someone pointed out the layout and the models are very 'angular' but I suspect they may have been finding a subtle way to call me a square. (FYI; I'm not a square) |