Class: Scratch built
Animated Textures: No
Allow Reflections: No
Allow Truck Reflections: No
Download: <a href="http://mtm2.com/~trucks/dl.cgi?dl=876" target="_self">184</a>k
Class: Scratch built
Animated Textures: No
Allow Reflections: No
Allow Truck Reflections: No
Download: <a href="http://mtm2.com/~trucks/dl.cgi?dl=884" target="_self">187</a>k
A few preliminaries. First up, for all intents and purposes, <i>The Skelaton</i> and <i>Pretty in Blue</i> are the same truck but with different color textures. As such, we'll treat them together for the sake of this review. Secondly, both vehicles are a frame without body which makes it tough, tho not impossible, to know where to begin with a review. Thirdly, we have no idea what Metalicagod had in mind when building this, and we had an equally hard time knowing what to make of it ourselves (ie, repaint, custom, scratch built; and what level of experience we should pin him down to). Lastly, with all this in mind, we decided not to impose many preconceived ideas about truck making onto his submission and opted to take an open approach to its assessment which we'll try to base on its strengths (and there are many) and its weaknesses (and there are a few) so that hopefully anybody who reads this will come away the wiser for it.
As for the truck, we'll stay away from conceptual matters, tho it's tempting to guess or speculate why no body was fitted over this frame (was it recognition of an already high vertex count, or just a cunning way of getting into the advanced category without getting bogged down with building an entire truck body? ;o) Instead, we'll try to stick to the basics of truck making: art (including texture use and design), modeling (including the truck and peripherals - wheels and axels), and the trk file (including handling and fine tuning). Make no mistake, we don't mean to sell this short, not at all, because even this naked-framemobile is over 2450 verts so there's obviously been some good concerted effort put into it.
The paint scheme on The Skelaton is interesting... being a mixture of white, fleshy sort of tones and streaks of red, applied all over the vehicle (including the driver) to make it look like a skeletal structure that has had the flesh ripped off it. Hence the (misspelt) name. Disappointingly, the truck body alone uses a whopping ten 256x256 texture files - disappointing because practically all of the textures would be just as good at 64x64 size (indeed the bone texture looks like it's been blown up from that size), and/or could have been combined together into as few as one texture. As it is it's terribly wasteful, especially so when you consider the nature of the truck (stripped, minimalist look; certainly no painted logos or the like). There are whole texture files with nothing but one colour in them here - and they're 256x256 too.
Pretty in Blue is essentially exactly the same (minor differences to the driver's seat, etc; two fewer verts), except that the bare-bone textures have been repainted in sky-blue tones - less gory to look at but thematically not as clever as the skeleton idea. We have to wonder whether at some point it was intended for a body to go on these frames, or whether it was designed like this from the start.
Metalicagod claims in his readme that this is all his own work and we can find no reason to doubt that assertion. However, it seems apparent from the high vertex count, the incredible detail in many portions of its construction, and some design strategies that it would be a good bet to say this was made using professional 3D software (eg. 3DSM, Lightwave, Maya, or the like). This is not a problem, per se, but these programs lend themselves to practices and habits that are not really suitable to good running vehicles for mtm2.
The problem in each of these cases is not so much one thing or another, but a combination of them all, and which stem from approaches in design. For the purpose of illustration, let's use the shock assemblies as an example. Metalicagod created each shock cylinder-piston assembly using four parts: an inner cylinder, an outer cylinder, and two end caps (one for top and one for the bottom). Three of the parts use twelve points to make up the circular outer area, and the inner cylinder (presumably because it's hardly seen) uses six points. The inner cylinder is fitted inside the outer cylinder and extends beyond it. The end caps are then fitted over both cylinders, top and bottom, to complete the assembly of a reasonably good looking shock absorber.
<center><img src="http://cownap.com/~mtmg/contests/expo2003/pics/reviews/skelatonshocks2.gif" width="548" height="432"></center>
Looks great, but bogs down the truck in excessive vertices and increases problems in the game. This can be avoided by taking a couple simple steps. (1) Since the inner cylinder is barely seen, we needn't use two when one can do the job. The choice is to lengthen the outer cylinder and put ends on it, or fatten the inner cylinder to the size of the outer one. Then discard the one you don't use. (2) The end caps make use of two face creation strategies. The first is to create faces all the way across the outer surface. Then the, in this case, inner cylinder extends out through it so it looks like a snug and proper fit. The other is a ring of twelve faces that are sized to fit the gap exactly between the outer rim of the cap and the cylinder wall, thus forming a dust boot affair. The problem here is this second method uses twelve extra vertices and six extra, unnecessary faces. The solution is to delete this second group of faces and replace it with the same arrangement as the first - faces that stretch across the entire surface and allow the cylinder to penetrate through it. So, what we end up with looks something like this.
<center><img src="http://cownap.com/~mtmg/contests/expo2003/pics/reviews/skelatonshocks3.gif" width="548" height="432"></center>
If you're still following, let's recap: By eliminating a redundant cylinder, we reduce the shock by twelve vertices. By redesigning the dust boot, we eliminate another twelve vertices. In all, we've saved twenty-four vertices. Big deal, you say? Well, now it's time to do some math. Each wheel uses four shock assemblies. By reducing twenty-four vertices on each cylinder, we save ninety-six per wheel. Each truck has four wheels, so by reducing each wheel by ninety-six vertices, we have now dropped the vertex count for the entire truck by three hundred and eighty-four vertices, which just happens to be nearly double the vert count of a stock truck. That's no small potatoes, and we haven't even reduced the number of points used for the circumference of the shocks. If we reduce each shock from twelve to eight points (which is still high), we drop the truck another two hundred and fifty-six vertices for a total of six hundred and forty down from the truck as its currently made. Six hundred plus verts and we've saved more vertices than are used on some complete fiberglass trucks - and we haven't touched anything except the shocks, nor have we lost significant detail. This streamlining of the shocks could conceivably allow the addition of a fairly complex body without increasing the overall vertex count. Big results from some pretty small steps.
This is definitely the hazard of using professional 3D software to create mtm models. In those programs, you can use as many points as you wish (forty-eight for shocks would not be uncommon), then you start a render, walk away from your computer and come back in a couple hours to a beautifully drawn scene. In mtm, on the other hand, the game bucks at high vert and face counts, and if you can get it to run at all, then frame rates are down the drain. Therefore, increasing the detail of shock models even a little can cause a big increase to the total vert count of the model, and dramatically effect how it performs in game. And take this same approach to the engine, the steering wheel, the frame, the driver, and you can convert a vertex laden vehicle and turn it into a sought after and highly praised and highly usable truck for mtm2.
The readme claims Metalicagod made "everything", and it seems this even extends to the axles and wheels, which have been painted to match the colours applied to the truck (ok, frame), though unfortunately there are visible horizontal texture seams on the tread face. The wheels on this are a little unusual, as they are centred around the inside/back of the wheel hub rather than the exact centre of the model. That is, they are offset horizontally so that for steering they pivot about the point where the wheel joins the axle, rather than the central axis of the wheel. This makes it appear more realistic, but it has the drawbacks that when the truck leaves tire tracks, they are not aligned with the wheels. Also, the circumference of the axles is too small and the shocks do not come all the way down to connect with them.
Overall, these two are a great scratch-build attempt that is let down by some very poor texturing decisions, excessive vertex detail, not enough attention to handling, and a vague feeling of incompleteness. Metalicagod has shown himself to be an excellent modeler who now needs to gear his work more specifically to the game.