Archive for March, 2012

Warming up.

In art, like in physical exercise and other activities, warming up is often recommended.  Whether in a formal art class, or in text references, there is almost always some attempt to get the students to get in the habit of skating the page with scribbles.  This helps get a tactile sense of pressure, resistance, and speed of the substrate as marks are being laid down.  Mike Mattesi, in his books and video lectures (which you can find links to in my reference tab above) also has a variation on this, where the purpose isn’t just to get a feel for the surface, but also as a quick reminder to the brain to be conscious of how curvature works with speed, and how the sharper the curve, the slower the speed.  He uses the analogy of a car driving around a track, speeding up in the straight sections and slowing down to make the bends.

The result of that kind of exercise usually looks something like this:

All of this is extremely helpful, but I’ve been adding to my ‘warm-up’ routine as of late, because seriously, who ever warms up with just one stretch?  But there has to be a purpose behind the warm-up and in this case, it’s about putting the mind in a place where it sees form and structure in addition to line.  Taking Mr. Mattesi’s lead a bit, we draw two curved and forceful arcs, such as per his rule, that one of the forces leads directionally into the apex of the next, forming a rhythm.  The key here is that I’m not looking to draw anything real.  I’m just placing lines at this point.

Just two lines, right?  Well, because of what we know of force from the references mentioned, the next step is to see form, and one of the keys to generating form is overlap, so let’s throw an example of that into the mix.

Now you can almost start to see the shape of something dimensional forming in the mind.   Now keep in mind that though I’m not specifically planning any of these marks, I am still trying to be aware of space.  If I put the marks too close together, seeing form and volume develop might be more difficult.

In any case, let’s go ahead and add a couple of straight lines to add some structure and contrast to the curves.

Mind you the order I’m doing this in is not gospel.  Straights followed by overlap works just as well, but in the very beginning it helps to use the two curved lines in some kind of forceful arrangement.  We can go further by adding some marks indicating volume:

Not my best endeavor, but that’s what warm ups are for, to work out the kinks and set your mind working right.  I went ahead and did a little more to this one and added tone for effect, but not that you have to, unless you want to practice using tone to indicate depth at the same time, which is great, but as I’ve been told and instructed myself, it’s probably best to learn depth through line first, and tone later.

Anything you come up with here may end up looking rather surreal, but at the same time you may find yourself inadvertently drawing part of a figure.  It’s whatever comes into your head on the spur of the moment.  These are quick exercises, and doing them has helped me look for and see these same qualities in the representational work I go on to do after the warm-up.

That’s pretty much it.  Sorry for the delay in updates the last couple of weeks, but I hoped to make it up a bit with a double post this week.

Until next time.

Advertisements

, , , , , , , ,

Leave a comment

Blend over Blend

Ever since I first started tackling Maya, and the subject of blend shape deformers, I wondered if there was a way around the most common way of using them; namely, the duplication of the original shape and the push-pull tweaking of the original shape .  I’m not going to complain that said method is ineffective or inefficient, because if you’re using it to do what it’s most commonly tasked with, that of creating facial expressions on characters, then it’s certainly the best way to go about using them.
However, what if you wanted to have one shape completely change into another specific shape?  Well, if you can visualize comparatively what the parts of those shape are that will become the parts of the new shape, then what I’ve learned is that indeed, you can do this.

The trick is, how to build two sets of objects side by side not only with the same number of vertices, but so that the vertices are ordered identically.

Fortunately, I covered a way to do this several weeks ago.

What follows is how to take that knowledge and applying it across several parts intended to be merged into one object without losing that ordering of vertices, or the intended effect of either blend-shape, independently able to be controlled.

To that end, I’ve created for this demonstration, what will end up being a single object with two blend shapes attached.  For now though, that object takes the form of two separate objects one on top of the other, and their corresponding two blend-shape targets, all created independently of one another (not duplicated and modified) but tested to ensure that they have the same number and order of verts as the first using the methods already outlined previously.

From here, we simply create two blend shape nodes, one for the two top halves and one for the two bottom halves.

The rest of the process is relatively straightforward.  After merging the objects, the surface can be made seamless again without re-ordering the verts by simply selecting and deleting one row and one row only of polygons from around the circumference of the morphing object.  The reason this is, is because eliminating one row of faces along the seam line only eliminates the faces while leaving the verts and their ordering in place, and it’s the vertices that matter most when doing this.  Now all that is left is to either use the bridge tool, or the append polygon tool to make the surface continuous.  The final result is a morphing object that has two independently created targets applied to it, and with independent control, as you can see by the varied rate of morphing in the following clip:

Now, this method is not without it’s drawbacks.  Primary among them is the fact that we are still left with multiple blend shape nodes, which bogs things down the more you have of them.  Have not yet figured out whether or not it can ever be done so that both blend shape controls occupy the same node by modifying the steps shown here, but if I do, I’ll be sure to make an update about it.

Leave a comment