finally Ears

Finally finished an approximation of the ear. By far the most difficult modeling task so far. Definitely not a great thing for nurbs.

This represents my 5th attempt and is a rough approximation of an ear. The problem areas included a huge buildup of isoparms along the front edge of the ear causing smoothing problems, the challenge was to keep the isoparm count down while not loosing detail. I was able to get the surface density down to 13 x 9 spans. The other problem area involved capturing that middle flap of cartilage that starts from the front edge of the ear and sweeps rearwards. While I can't speak for polys or subD's, nurbs has a tendancy to want to flow in a uniform direction, but that flap of cartilage interrupts the circular flow around the ear by curling in towards the middle. That was a pure sculpting challenge and I've come to realize nurbs are a total bitch for something as complex as an ear.

In a previous iteration, I had sculpted a decent ear, even matching the right number of spans to interface with the head patches. The problem was, the ear isoparms did not line up to the head isoparms. They were so off in position that it was impossible to align them without fucking up the ear.

My first attempt came from a tutorial which involved duplicating the edge surface isoparms that defined the ear opening into curves. Those curves were then attached and closed. The resulting curve was then duplicated numerous times to form the ring profiles of the ear. Once lofted, the surface was then properly sculpted, detached and stitched to the head patches. This didn't work because positioning the profile curves with any accuracy was impossible, but that was to be expected, and the sculpting process introduced a lot of smoothing problems, not to mention the final ear isoparms did not align evenly with the head isoparms. The other problem involved the final stitch process, the ear was nowhere near properly aligned to the head patches and the stitching became problematic.

Someone else suggested I build the ear profiles out from the head isoparms and then do a radial loft. This was the worst solution as the resulting loft not only introduced so many ring isoparms, the isoparm flow looked like a wavy mass of uncontrollable spaghetti.

The method that eventually worked for me was to build out tangentially aligned 1 degree curves from each head isoparm, lofting them and then closing the lofted surface. I then strategically started inserting ring isoparms and pulling the cv's out slowly sculpting the ear. I was careful to keep the isoparm count down and made sure the difficult sculpting areas were dealt with first so that they would not become problems later on when there were a greater number of isoparms to deal with. Finally when finished sculpting, I detached the ear surfaces and was able to easily stitch them to the head patches.

The next step is to take the ear into mudbox or zbrush for more accurate deep grooving.

Another tweak on the nose to properly shape the nostril and this head should be done.

No matter what pains I took to accurately sketch the character in orthographic, it still contained monstrously glaring flaws in accuracy. I now realize the modeler must also have an intimate knowledge of anatomy. The concept sketch is exactly that, a 2 dimensional abstraction. At best a roughly measured guideline for the modeler who then must translate the spirit of that sketch into 3 dimensions. Therein lies the true skill of the modeler and what sets a good one apart from a mediocre ones.