Fixed Vertical Scale, POA Training Ground Entrance, & Hogwarts Torch Sconces

Okay, I might have overreacted a little bit. The vertical scale issue didn’t take that much work to remedy; I forgot how relatively little I’ve modeled so far. A 5% increase in height got everything back on track, although I had to be careful with a few shapes that needed to stay perfectly circular or perfectly square. Here’s how the results look:

…yeah, pretty much the same as before. Hey, it’s only 5%. But this small fix will help ensure that all the different structures line up correctly as I continue to build.

Hippogriff-eyed viewers might notice the addition of the so-called training grounds entrance, just to the right of the DADA tower. I say “so-called” because it stopped being an actual entrance after Chamber of Secrets. Originally, this small cubical building connected to the suspension bridge on one side and the training grounds on the other. But then the suspension bridge moved in Azkaban, turning this building into a dead end that could only be accessed from inside the training grounds. It never changed back. Still, this seems to be the name that the fan communities are using, so I’ll roll with it. (Except for in the internal layer organization of the model, where I stubbornly refer to it as “small cubical building.”)

While you’re taking a peek at this modest addition to the model, you might also notice the torches burning on either side of the archway. I’d been meaning to model the black wrought iron sconces seen in numerous exterior shots, and this was the perfect chance. As always, I started without texturing:

As I worked on this, I noticed that the bowl of the sconce was always empty in daytime shots, but always full of visible firewood (?) when the fires are lit at night. I’m taking a similar approach. Here’s a classic day-for-night shot of the torch by Sirius’s cell atop the dark tower, created by significantly reducing the brightness of the daytime background HDRI:

I’m rather proud of the simple little fire material I created, too. It’s a volumetric shader applied to a roughly conical shape, with black firewood shapes down in the base. The 3D texture that simulates the characteristic shapes of the flames is tied to the world coordinates, not the local coordinates of the individual fire object, which means I can duplicate the exact same torch and fire objects but still get a unique pattern of flames on each one – very important in areas where multiple torches are in close proximity. Then, when switching back to daytime renders, all I have to do is rewire a couple of nodes to hide all the firewood and flames across the entire model.

The one thing I didn’t like about this render was that some areas of the stone bricks are just too smooth, though. So I went back in and beefed up the normal maps that simulate the rough, bumpy surface texture. If you look closely, you can see a bit of it in this last render of the training grounds entrance, though I’ll probably keep tweaking it as I continue. (Also, I forgot to un-hide the lake surface below for this render, so you’re just seeing the photographic bottom of the HDRI background. Whoops.)

One last thing I keep forgetting to mention: If some of these buildings in this model look like they start too low or are too tall, that’s just because the surrounding terrain will cover most of the base areas. Because the terrain is so uneven, there are lower areas that get exposed here and there, and my model already includes those lower areas. Once I’ve got the ground and the hills and the ravines in there, it’ll look a lot more…Hogwartsy.