Siegfried Act 1

I’ll be photographing a production of Siegfried Act I later this month, so I did some reading up on it. As much as I enjoy opera, there’s plenty of them that I know nothing about.

Anyway, there’s a pivotal scene in which the main character shows off the sword he just forged by slicing an anvil in half.

That’s a pretty powerful mental image, and I fixated on it to the point where i had not choice but to try and model it.

I’m still super n00bish at blender, and especially the texturing component. I intentionally kept it illustrative, rather than go for realism. I also kept the scene kind of sparsely theatrical to reference the opera production that inspired it.


I plan on doing a more interesting version with some particle ember effects in the background. But I’d love some feedback on work thus far?

Thanks!

More drama in this one:


The second one is a lot better! The anvil looks so much more 3D and grungey with the lighting. I think the composition may be improved by adding some background elements that lead back into the focal point (the anvil leads to the sword and then away into the corner. There’s nothing to lead the eye back!). The line that points toward the corner should NOT do that, but if you’re a photographer you probably know more about composition than me.

I think that simplifying the image is smart, but you definitely shouldn’t leave anything flat. Grab some textures from CGtextures.com and start adding grain and bump to those flat surfaces! The floor is a stage floor, so it would love some wood-grain textures. The anvil wants some rust. The black block wants some sculpting.

Anyways, the lighting is good. The sword looks good, and it looks really sharp, too. The texture on the cut part of the anvil is great.

Hmm. If you aren’t interested in photo-realism, I think it’d be good to experiment with Freestyle, and post-processing in the compositor and GIMP, if you haven’t already.

I have a little advice to give about making textures if you want.

thanks so much for the feedback!

The anvil’s textures (the dull lumps and the grain of the sliced face) i generated in photoshop with difference clouds and motion-blurred noise.

It’s the application of textures that i’m working towards getting better at. Specifically the layering of multiple textures. As you mentioned, some patches of rust and dusty diffuse spots would be great, but how do I layer that on top of the existing textured material?

It sounds like you’re pretty comfortable in Photoshop. That’s good, because most texturing isn’t even done in Blender. It’s taken me a while to learn about texturing, but there are some great references and resources online.

I don’t know how much you know, so I’m going to write about everything, starting with adding layers on top of existing materials. Go ahead and skip the parts you already know.

First, hook up the texture you are trying to layer into an emission shader so that you can look at just the texture. Then, mix the textures-I always use an alpha map of some sort, or plug a noise texture node into a colorramp and use that as a factor to mix textures with a Mix RGB.



Once you are satisfied with the layering, use that same colorramp as a factor to mix the other textures (as I have done in the example) and plug’em into the shader just like normal. You’ve made two specmaps into one, and two colormaps into one. You can do this with normal maps, too. And you can use any kind of texture instead of the noise at the beginning- marble, musgrave, voronoi, you can even use image textures if you set them to non-color-data.

Some nodes to know and use often are: colorramp, color curves, mixRGB, invert, math (this one is so important). It’s all about tweaking the values till everything looks right.

Getting the Textures:
CGtextures.com has great high-resolution textures, 10MB a month for free, with a few limitations. It’s really awesome, even if you can’t afford any more than a free subscription.
3Dtotal has some nice free textures, too. Make sure to read the license information for both sites.

Making the Maps:
http://www.blenderguru.com/tutorials/the-secrets-of-realistic-texturing/#.VDnamvldV8E this tutorial is awesome because it goes over the 5 basic texture types and then gives you a good solid node setup. BUT: Don’t worry about the program. You don’t need Crazybump or Knald or anything like that (But they are awesome, I wish I had one).

So if you find a nice woodgrain texture for the floor or something, then open it up in Photoshop, and then make several copies of that layer (I use GIMP so some of the steps may not follow). To tile the image if you want to do that, offset it by 50% on the x and y axis (in GIMP this is ctrl-shift-O, took me FOREVER to find that), then basically just use the clone tool intelligently. Here are a couple of videos of a pro doing it in PS: (and a video of him alpha-masking a texture that I haven’t seen yet. Awesome!) https://www.youtube.com/channel/UCjT4kEPRJvt8AnxtC8ejfCg/videos.

I make spec maps by playing with color curves, brightness/contrast, and merging several layers together using darken/lighten only, multiply/divide, and overlay. I often make changes color-by color (for example, an image may be of painted metal, the metal part will be more reflective than the green paint, so darken the green channel and lift the other channels) and I always desaturate it, usually before most of the other steps. Specmaps should be black and white. Just look at the picture, think “what part is shiny” and find a way to isolate that part as white and darken the other parts. Of course, the contrast shouldn’t be too high, you should never have areas of full black or white in any image, but should make good use of every value range. So instead of pushing the contrast with the brightness/contrast, sometimes it’s best to use curves. Also, when you save the map, it’s usually best to use a lossless file-type that can hold an alpha channel (like png), and you should save spec maps and occlusion maps in grayscale mode to save on filesize (hi-res textures in color can be 25MB each).

Occlusion maps are made the same way as spec maps but they have more white, basically. I usually make the occlusion map first, copy it, and use it as a multiply over my specmap.

Displacement maps, are hard, I still haven’t figured those out.

Normals maps are easy. For you, there should be an Nvidia Normal Map plugin in PS, but I had to search for a copycat plugin in GIMP.

So anyways, I hope this isn’t too much, I hope it helps. There are a lot of tutorials out there about node setups, and this is stuff I’ve learned from watching a lot of them.
Anyways, I love the idea of doing an Opera piece. Opera doesn’t get nearly enough recognition. It’s awesome, I too wish i knew more about it. I only know a few of the names, like Wagner, and Barber of Seville. I should watch one of those one of these days. Good luck!

I agree with JosephBburg, texturing is fundamental, you have to improve the sword, it seems not a 3d model, just paper

Yeah! Did a good job with those textures. And the grass, too. I haven’t tried grass in a while because I’m too lazy to UV unwrap the grass models I did a thousand years ago. I wanna see some of that in this scene!