[WIP] Prop practice. Ripping to shreds with critique very welcome!
null
Thanks for looking, and thanks for helping me with your critical thoughts!
I hear that the industry needs environment artists, so I'm teaching myself environment, prop, and lighting design. I can't afford much, so what isn't a free texture/graphic is hand-sculpted and painted in Blender and Photoshop.
I would list exactly what I'm looking for critique on, but honestly at this point I'm interested in anything you have to say.
Rendered in Unity 2017 using the post-processing stack, a glass material, and some wood textures from the asset store. Image captured via screenshot. Snowglobe uses 11,306 tris and 4 textures, card uses 2 and 1. I'm not sure what other information to give, so I'll answer any questions as soon as I see them.
Again, thank you so very, very much for taking the time to help me improve!
(I'm calling the composition "Dad Loves You". Semi-autobiographical in that pops wasn't a big rig driver, but I feel that the symbology is stronger then a barrel of oil.)
I hear that the industry needs environment artists, so I'm teaching myself environment, prop, and lighting design. I can't afford much, so what isn't a free texture/graphic is hand-sculpted and painted in Blender and Photoshop.
I would list exactly what I'm looking for critique on, but honestly at this point I'm interested in anything you have to say.
Rendered in Unity 2017 using the post-processing stack, a glass material, and some wood textures from the asset store. Image captured via screenshot. Snowglobe uses 11,306 tris and 4 textures, card uses 2 and 1. I'm not sure what other information to give, so I'll answer any questions as soon as I see them.
Again, thank you so very, very much for taking the time to help me improve!
(I'm calling the composition "Dad Loves You". Semi-autobiographical in that pops wasn't a big rig driver, but I feel that the symbology is stronger then a barrel of oil.)
Replies
Obviously, as a beginner, there's going to be a lot to critique. But it can be hard to find a specific area to improve in - when all of your skills are around a relatively similar bar, nothing stands out as being particularly bad or particularly good compared to the rest, and thus it's difficult to decide what to focus on.
Your modeling, textures and lighting are all rather simple. This isn't to say they're *bad*, they're just simple. To illustrate this, take the example of the snow globe's model. There just isn't a ton to model there, especially if you do a game piece. You've got a dome, a stand, some (usually rather simplistic) elements inside, and that's roughly it. You've done a decent job of representing them here.
Now, how do we take this further? By practicing making something more complex. And this is my main critique here, as well as what I believe to be the fastest way one can improve. Pick a specific area and choose to do something more with it. It could be to make a really detailed truck in your snow globe. It could be to try to figure out more advanced texture painting, to add wear and roughness to your textures to give them a photorealistic look, or to go in a completely opposite direction and add stylized details to them. Or it could be to start a different model altogether - maybe to practice your sub-d modeling on something complex (like a fully detailed wheel of a life-size truck, or a plastic game controller, or basically any sort of object you feel will challenge you while being something you could also feasibly model).
For games, there's a couple things that are really important to learn. Aside from your traditional modeling skills, being able to create a high and low poly, and properly bake from one onto another is something really important to get good at. If you're planning to texture your models as well, some experience with PBR texturing (and Substance/Quixel) will be really useful, although in theory you can do all your texturing in PS the old-school way (I just find it's a lot harder to get into).
Porting stuff into a game engine is very important and I encourage you to keep doing it (it shows employers you're actually making your stuff with games in mind), and creating little scenes or vignettes for your work helps a lot with presentation and is another thing you should keep doing with anything you want to put in your portfolio.
Lastly, if you post a technical breakdown of your scene (wireframes, UVs, texture sheets), I'll be happy to give you feedback on these. Technical skill is equally important as artistic skill for 3D artists, and is another area that is often overlooked.
You can also get free student licenses for a lot of the paid programs. Maya/Max are definitely worth it (you can get either for free very easily) and I'm fairly sure you can get free student licenses for Substance programs as well (though you might have to provide a student ID).
In my opinion there's honestly no "power" difference between the free options (Blender and such) and paid programs. You can make good art in any of them all the same. However, there's way more learning resources targeting paid programs, which is why student licenses are so valuable and useful. You can learn game art using Blender, PS, and nothing else, but it won't be near as fast. (There might actually be some decent tutorials on Blender, but Maya and Max still have a greater variety, and the same is true for programs such as Substance/Quixel, Zbrush, and so on).
Wireframes from Blender - http://imgur.com/a/IT5ci
Textures - http://imgur.com/a/pqXSx
As far as a challenging thing to model... how about this thing? Cauldron shrine pic from shutterstock. The Exalted tabletop RPG has me on somewhat of an Eastern kick, and a foo dog is a bit too advanced for me right now :P
In terms of getting paid programs and tutorials, I'll sit down and give Autodesk's 3D tools/Substance Painter and Designer a serious look on my next two-day weekend. Which tutorials specifically would you recommend? I can swing maybe one a week at most at this point.
You'd be throwing in yourself for a challenge in making that cauldron. There's some pieces, like the faces and dragons, that'd most likely only look right if sculpted. There's a lot of other ornate pieces that you could polymodel (with subdivisions), but would require a significant amount of work. If you're curious how to polymodel ornamental pieces, though, Hugo Beyer has a good tutorial on the process visible here: https://www.artstation.com/artwork/B3kVD
You likely won't get to reproduce every detail on that shrine right, because doing all of it will require the use of a lot of techniques (the sculpted parts being the biggest pain), but you could try to get as much of it done as you could, which would still be a pretty good exercise. I would personally recommend leaving out the complex animal shapes and faces and focusing on getting some ornamental detail done with subd modeling, though if you're feeling like giving sculpting a shot, don't let me stop you (I'd probably try it on something less complex and with more reference available first, though).
There's basically three situations you could end up in when trying to challenge yourself with a new modeling project. You can stumble upon something you know how to do well, something you think you might know how to do somewhat, and something you have no idea how to do at all. Each of these situations should be approached differently. If you already have experience in what you're doing, you're gonna want to look at the details and how to make them better. Add more geometry to your models, make it cleaner, more accurate, work on more detailed or lifelike textures, etc. If you sorta know what to do, or don't know what to do at all, then you're gonna have to clearly sort out the problem you're going to face (in this case it could be the ornaments, for example), then find a way to approach it. It could be by looking at how other people have addressed this problem before, looking for tutorials on the subject (very helpful when starting out, rarer and rarer as you get more advanced), or trying your own experiments and asking others for feedback (can be very slow if you don't know what you're doing, but being able to take initiative is also an important skill and will often be needed as your skills progress. I wouldn't stress that stuff as much when starting out because most simpler 3D problems have already been answered by someone).
Your general progression is up to you, but I would recommend picking a tutorial on a subject you find interesting, following along, then once you're done, doing a model of your own that applies something you've learned through the tutorial. That way you'll be able to put your skills in practice (which will help you understand them rather than simply memorize them).
For recommendations, the one tutorial I'll always recommend to beginners is Chamferzone's Max/Substance Painter grenade tutorial. It follows a complete and up-to-date workflow for modeling, baking and texturing current-gen assets, step-by-step, in full detail, while simultaneously being relatively concise. Best of all, it's also free.