Spectrum: Interactive Media & Online Developer News

4 April 2005
Reported, written and edited by David Duberman
For editorial/subscription inquiries, send mailto:spectrum1@broadviewnet.net
Search the Spectrum archives at http://www.3dlinks.com/spectrum


Editor's note: Three big reviews in this week's Spectrum: a 3D program, a book, and a game. I hope you enjoy them. We'll be back with news next week. - David Duberman


Today's Headlines (details below)

--Software Review: modo by Luxology
--Book Review: chris crawford on interactive storytelling
--Game Review: God of War from SCEA

About Spectrum



modo Review
by David Duberman

Things have changed since the '90s. In those days, it seemed as though a new 3D graphics program came out every other week or so. But lately, innovation in 3D software has slowed not so much to a crawl as to a virtual state of suspended animation. Certainly, the few publishers that survive continually churn out new versions of their existing programs, but they're based on fairly old and creaky foundations. Thus it comes as a breath of fresh air that Luxology, a company founded recently by former Lightwave 3D stalwarts Allen Hastings, Stuart Ferguson, and Brad Peebler, has embarked on a course of world domination, or at least domination of the 3D market, with a bold new line of 3D software. The first program in Luxology's campaign is modo, which, as its name suggests, is a modeling program with chutzpah out the wazoo.

Whenever you see a modo demo, the first thing the demonstrator does is start dragging around UI elements. Ferguson's previous 3D modeling program, the Modeler module of Lightwave 3D, was hardly customizable at all. In modo, however, he's gone to the opposite extreme. Not that the program isn't very usable out of the box, mind you, but once you start using it intensively, you might want to rearrange things to suit your working style.

modo's interface takes advantage of the malleability inherent in the medium of software more than most. You can instantly change any window/panel into any other, move things around within the interface or create new overlaying windows, create new rows and columns for panels and windows, drag borders to resize UI elements, etc., etc. And you can save, load, and delete custom UI layouts. You can't delete the three default layouts provided with modo, which is a good thing, because it's easy to completely screw up the UI with just a bit of messing around. But power often comes at a price: I managed to crash modo several times while rearranging the UI. Also, the layout-recall feature is a bit quirky: After playing with the program for a while, now whenever I load the permanent "modo Default" layout, the view comes in sideways; I have to load it a second time to get it to look right.

Customization goes even further: You can open a Command History list and use it to apply a keyboard shortcut to any action, such as a specific transform. There's also an interesting Tool Pipe feature that I didn't quite get a firm grasp on, but at least I learned that it's how you create a primitive and then change its creation parameters. Speaking of which, as far as I can tell, modo has no such thing as a modifier stack (as in 3ds max), which seems a serious deficiency. This doesn't let you go back and change modification parameters (such as a bend operation) after the fact. modo is highly scriptable, but I didn't have time to investigate that aspect of it.

Viewport navigation is not too intuitive: Alt+drag to rotate, Ctrl+Alt+drag to zoom, and Shift+Alt+drag to pan. It's much easier for me to remember how to do these operations with 3ds max using the different mouse buttons and wheel, but that could be because I'm more used to the latter program. On the other hand, the fact that modo is also available for the Mac, which is still hobbled by its one-button mouse, could be at least partly to blame for some of these arcane key combinations. And let's face it: No matter how easy it is to customize a program (and in modo, it could be easier), most users are going to use it as is.

Selection in modo

modo's selection tools are quite powerful. By default, selection works uses a "paint" mode, where you LMB-drag over entities to select, or RMB/MMB drag to select a region (lasso/rectangle/circle/ellipse). You can click buttons or use the spacebar to change selection modes, but the easiest way to set a mode is simply to right-click an element in the model. The program remembers different selection sets of vertices, edges, and polys, but by holding down a qualifier key when you change modes, you can convert a selection to a different type. Even better, there's an option to let you convert a contiguous poly selection to an outline of edges.

Double-clicking has a special selection function in modo; for example, double-click an edge to select the loop to which it belongs, or double-click a poly to select all contiguous polys. You can use the arrow keys to increase and decrease selection areas, and even move selections over the surface of a mesh object. For example, say you want to select an edge loop, but it's not readily accessible. Just double-click a nearby edge to select its loop, and then press the arrow keys until the loop you want is selected. Even better, say you want to select every other poly around an object's circumference. Just select two alternating polys, and the use the More selection command (up arrow key) repeatedly until you get ever other one. More recognizes the nature and order of the current selection and increases it accordingly; pretty clever! All selection commands are available from what modo calls a "popover" menu, essentially a contextual menu bar that appears at the mouse cursor when you press Alt+spacebar.

Selection can be a bit unorthodox. For instance, in wireframe view, dragging a region with the middle mouse button selects only front-facing polys, while using the right mouse button selects front- and back-facing polys. But in shaded view, it's the opposite.


modo's modeling functionality comprises an embarrassment of riches, of which I'll barely be able to scratch the surface here. To start with, it offers the standard set of primitives: box, sphere, cone, torus, etc. You create these in the workspace by various combinations of clicks and drags, or you can instantly create a basic example of the current primitive by Ctrl-clicking the tool button. Once you've added one, you can distribute identical copies by MMB-clicking anywhere in the viewport. A nonstandard primitive is the Tube, which lets you create a snake-like shape with multiple clicks. Essentially, you're extruding a circle along a spline that you define on the fly.

Another way to create geometry is by starting with an outline (i.e., a spline) and then extruding or lathing it. You can create standard Bezier splines or simpler curves vertex by vertex, or the latter simply by sketching. You can then use Duplicate commands such as Lathe, and Curve Extrude/Clone, which use a background curve to extrude or clone polygons or another curve. This, incidentally, is one of those commands that I just couldn't figure out; the results didn't even come close to my expectations. I had better luck with other Duplicate commands, such as interactive Array and Radial Array tools. There's also a Bridge tool that works almost exactly like the one in 3ds max.

Once you have your basic form, modo gives you a multitude of editing tools, starting with the standard move, rotate, and scale for objects, vertices, edges, and polygons, plus a specialized squash-and-stretch function. And the powerful Falloff feature lets you apply transforms and other operations with widely varying results over the geometry surface depending on the position and shape of the falloff envelope, which can be linear, cylindrical, or any of several other shapes. Incidentally, with no sub-objects selected, modo considers all geometry on active layers to be a single object.

The ability to perform transforms around different types of centers is important, and modo offers a wealth of choices in this regard. You can set as a transform center the origin, and arbitrary point or element (such as a particular polygon, or even a point in space), the local coordinates of each selected element, and several others.

Other modeling operations include bevel, subdivision surface (SDS) features, and a host of deformation tools such as twist and bend. There's also Push, which moves elements in or out along their normals; Sculpt, which lets you work the geometry like clay, and taper and shear tools. A couple of particularly nice and intuitive tools are Soft Move and Soft Drag, which let you set a radial falloff and then move parts of objects while affecting their neighbors to a lesser degree.

One of my favorite modo features is the implementation of layers. You can create a new layer with a single click in the Mesh List panel, and then cut/copy/paste geometry between layers. You can activate any combination of layers by highlighting them in the mesh list, and then perform any operations on them as if they were all in the same layer. You can also set a layer to Background status, so that it's visible but can't be acted upon. This is very intuitive and easy to use.

An important aspect of modeling entails using background images in the viewport to serve as templates. modo supports a wealth of options for this purpose. You can load different images into memory and then switch among them easily in the different viewports. You can adjust size and position for background images interactively in each viewport, and rotate the image using a spinner. You can also set contrast, brightness, and resolution values separately in each viewport. For black-and-white images, the Invert function can be useful, and you can even set the image to be an overlay, seeing through to geometry behind it thanks to a Transparency setting.


modo's contextual help is a no-brainer: Press F1, and the mouse cursor turns into a question mark. Then, when you click any GUI element, a relevant HTML reference page opens in your Web browser, which then comes to the front. Likewise, pressing F1 causes a question mark to precede every menu entry, and then clicking the entry usually, but not always, opens an appropriate help page. This works on a one-time basis; if you want to know about several items, you need to press F1 several times.

The help page accessed via F1 is usually specific to the item you clicked, but in some cases, such as a form (equivalent to a rollout in 3ds max), instead of a description of the parameter you clicked, you just get a less-helpful generic help page that just tells you what a form is. Often this page has a link to a helpful QuickTime movie about the feature, but in most cases when I clicked the link the movie opened twice--once on the page and again in a new window--forcing me to stop one instance in order to be able to understand the narration. What's more, if I closed the new window--the intuitive thing to do--the other movie kept playing, but without sound. Kind of frustrating.

The HTML reference isn't the most helpful; despite its online nature, surprisingly there's no search capability. If you go to the home page, which you can do only from the Help menu (there are no cross-reference links on the reference pages), all you get is an incomplete, poorly organized index-cum-table-of-contents. You can try to find items in the index pane with the standard Web browser search, but when I searched for "pipe" (as in "tool pipe"), an important concept in modo, it came up empty. The documentation writing takes a mostly professional tone, but is occasionally a bit too self-congratulatory with language like, "That is sweet," or, even more embarrassingly, "Super bon bon!" (in tutorials).


modo is very sexy: It's a shiny, new, next-generation 3D modeling program. The designers (mainly Stuart Ferguson) have learned from all the modeling applications that have come before, and created a piece of software that both old hands and novices can benefit from using. At the same time, the program design seems to revel in its eccentricity. Getting used to modo's idiosyncratic working methods takes a fair amount of concentration and practice, and if you stop using it for a while, you're likely to face a steep learning curve again, despite your prior experience.

Frankly, I find the program to be relatively unintuitive; in fact, working through the user guide, which combines overview material, tutorials, and procedures (focused mini-tutorials), typically as soon as I tried something different I started getting error messages, no response, or wildly unexpected results (including regular crashes). Often, while using modo, I feel truly at sea, and I've been using 3D software for well over a decade. For example, while following a tutorial on lathing a wine bottle shape from a Bezier curve, I wanted to adjust the number of sides (i.e., lathe steps), but was unable to figure out how. The parameter still appeared in the UI, but adjusting it no longer affected the bottle.

The bottom line is that modo needs to come with lots more tutorials, and they need to go into much more detail than do the current ones. That said, I know it's a very capable program; I've seen Luxology employees do amazing things with it; it's just that I can't figure out how they do it.

I really am impressed with modo, or at least I will be once they improve the documentation and nuke some of the nastier bugs, but it's still only a modeling program. Of course, it's just the first blow in Luxology's campaign to conquer the world of 3D content creation. Once Allen Hastings throws his animation software (demonstrated briefly but memorably at the 2004 Siggraph) into the mix, the company will truly be a contender; I'm watching with bated breath.



Book review: chris crawford on interactive storytelling
by David Duberman

At the recent GDC, and at just about every one before that, one of the hot topics has been how to make storytelling (aka narrative) work better within the context of games. This is as it should be, because currently, for the most part, story is essentially a distant adjunct to games. In most games, you get a little non-interactive story, then you run around fighting, then you get a little more story, fight some more, and so on, ad infinitum/nauseum. A game I've been playing lately, The Lord of the Rings, The Third Age, is a perfect example of this. It's fun, but the relationship between the two parts is utterly contrived. There's got to be a better way, but game developers have yet to discover it.

Enter Chris Crawford, famous game developer and founder of the CGDC, who has spent the last 13 years-plus thinking hard about this problem. Starting in 1992, after creating such critically acclaimed titles as Eastern Front and Balance of Power, Crawford gave up game creation to devise an interactive storytelling system, called Erasmatron. He named it after Desiderius Erasmus, the 15th-century writer/satirist, of whom Crawford is a devotee.

In pursuit of this goal, Crawford recently wrote a book, "chris crawford on interactive storytelling," published this year by New Riders Games. True to its title, the book starts out by defining storytelling, then interactivity, and concludes the introductory section by discussing interactive storytelling. In the first chapter, Crawford explains and illustrates how the story is a linear medium by which we transmit pattern-based knowledge to "our pattern-recognizing mental modules." It's also here that he gives us the first of the various important lessons in the book: Stories are complex structures that must meet many hard-to-specify requirements; stories are about people; stories are not about puzzles or spectacle. It's in these latter lessons that Crawford starts to bring out his curmudgeonly side, to wit: He really doesn't think much of modern games; in fact, he hates them. He condemns as them emotionally crippled; they're not so much antisocial as asocial; and game characters are to drama as inflatable dolls are to sex. So far, so Crawford; he doesn't hide his antipathies.

Crawford begins the next chapter by attacking interactivity, or rather a description of interactivity in a recent book (Pause and Effect: The Art of Interactive Narrative by Mark S. Meadows, also published by New Riders). After deriding Meadows's statements as "vague … misleading … drivel," he gives his own definition: A cyclic process between two or more active agents in which each agent alternatively listens, thinks, and speaks. He gives reasonable justifications for this, and follows up with requirements for interactive artists: second-person insight, which he acknowledges is quite difficult, and discipline, in terms of the ability to quash your own ego, because "you must face each player individually."

Later in this chapter, Crawford, presents, in extra-large boldface type, "the most important lesson" in the book: Interactivity depends on the choices available to the user. There now, you don't have to buy the book; aren't you glad you read this review? Seriously, though, Crawford validates this statement admirably, and you'll have to at least take the book out of the library (assuming they've bought it) to see how. Crawford closes the introductory section of the book with a chapter on interactive storytelling, in which he begins to get down to the practical aspects of the topic at hand, using Star Wars as an illustration of the issues that need to be dealt with.

There's much more to the book than I have room or time to tell you about here, but I should at least mention what Crawford believes to be the greatest obstacle to interactive storytelling. It's verb thinking, which is a nonintuitive mindset that developers must get into, as opposed to the noun thinking that is our wont. It's services vs. goods, wave vs. particle, logic vs. truth, and so on. According to Crawford, "software sucks" because good interactivity requires verb thinking, which most people just don't get. He goes on to offer practical suggestions on how to learn verb thinking, using examples such as the writing of Homer. Crawford is perhaps the most literate member of the interactive community, and he uses his facility in good stead here.

The remainder of the book offers strategies for interactive storytelling such as those based on data and language, and describes necessary core technologies such as personality models and drama managers. Crawford is as adept at programming as he is literate, so these sections tend to be somewhat technical, but they should be relatively accessible even to readers who possess only a nodding acquaintance with software development. He closes with descriptions of historical research into interactive storytelling, as well as early attempts such as text adventures. As always, it's interesting to read Crawford's takes on these predecessors/alternatives to his current work.

In his introduction, Crawford states, "This book reeks of the crotchetiness that is my alliterative middle name," and that's no lie. If you're put off by confrontational writing, this book probably isn't for you. But if you don't shy away from controversy and are willing to expose yourself to original, intelligent thought about the true potential of digital technology for advancing the art of storytelling, and consequently the collective maturity level of the human race, I recommend "chris crawford on interactive storytelling" highly. A mind like Crawford's is an amazing thing to observe, and in this book he lets us in on exactly what's been going on in his for the last decade-plus. It's well worth the price of admission.



God of War Review
by David Duberman

God of War rawks, as in, "Dude, you rawk!!!". The new action-adventure title came out of nowhere to become one of the highest-rated PS2 games in quite some time, and I have to concur, for the most part. In the game, which is based on Greek mythology, your third-person character is Kratos, a brutal, Spartan warrior who's tasked by the gods of Olympus with protecting Athens from an attack by Ares. Essentially, his quest involves whaling the tar out of countless enemies and solving some fairly clever puzzles. After each stage, Kratos encounters one of the Greek gods who grants him a new power and then lets him progress.

The action in GoW often seems incessant. There're some platform and jumping stunts, but most of the time, you're fighting … on the ground, on the walls, hanging from ropes, etc. You start out with a single weapon that consists of a pair of swords on chains, so you can throw them for a sort of ranged attack. Various combos are available right off the bat, and as you progress you gain more. And you need to learn them; you won't get far in this game just by repeatedly pressing the standard attack button. You're rewarded for using combos by faster recovery of the Rage of the Gods meter; when it's full, you can launch a brief high-power attack period. It's tempting to use it as soon as it fills up, but after you play a while you realize it's best kept in reserve for times of extreme need.

Similarly, there's no inventory, health potions, etc.; just single-use treasure chests placed at strategic locations. Some of these give you items that eventually let you increase your maximum health or mana, while others let you fill the power-up meter, used to increase your various fighting skills. And some chests let you restore health or mana. Again, your impulse is to use these as soon as you find them, but that's not always the wisest course; usually you can return later to use one when you really need to. Interestingly, it takes a little time to open a chest, so doing so in the midst of a heated battle can be risky.

Some boss battles are winnable only by succeeding in a mini-game that appears after the battle has gone on for a while. In these you must press a specific sequence of buttons with split-second timing. The sequence is randomly generated, so your memorization skills won't help you there. But what's most remarkable about these is the scale of the bosses; they're truly Brobdingnagian, and the graphics and animation keep getting better as you progress through the game. Likewise, the levels use some extraordinary special effects we haven't seen in other PS2 titles. This isn't one of those games that shoots its technical wad at the outset.

The puzzles are well thought out and nicely integrated with the rest of the game. Far be it from me to spoil any surprises, but I will warn you to be very observant. Interestingly, in the included "Making of GoW" video, the developers talk about how they needed to emphasize some puzzle keys during the fine-tuning process, but they're still not always too obvious.

If I had to describe GoW with a single word, it would be "spectacular." If you believe most of the reviews of God of War that have been published by other outlets, the game is one of the best ever for PlayStation2. And since it's published by Sony, that's the only platform you'll ever see it on, unless they make a PSP version. So should you buy it? About halfway through it gets pretty difficult, so I'd answer with a qualified "yes." If you're not a hardcore gamer, I'd suggest a rental tryout first; otherwise you might not get your money's worth, unless you don't mind playing on Easy mode. In fact, if you lose a battle several times in a row, the game offers to switch you to Easy mode. SCEA's Santa Monica Studio and designer David Jaffe are to credited with creating an enormously entertaining game, albeit one that's conceivably a bit too intense for casual players.




About Spectrum

Spectrum is an independent news service published every Monday for the interactive media professional community by Motion Blur Media. Spectrum covers the tools and technologies used to create interactive multimedia applications and infrastructure for business, education, and entertainment; and the interactive media industry scene. We love to receive interactive media/online-development tools and end product for review.

Send your interactive multimedia business, product, people, event, or technology news by email only to: spectrum1@broadviewnet.net.

If you contact companies or organizations mentioned here, please tell them you saw the news in Spectrum. Thanks.

Please send address changes (with old and new addresses), subscribe and unsubscribe requests etc. to the above address. If you use the Reply function, please do _not_ echo an entire issue of Spectrum with your message.

Publisher's note: We are now accepting limited advertising. If you'd like to offer your company's products or services to Spectrum's elite audience of Internet and multimedia professionals, send an email query to mailto: spectrum1@broadviewnet.net. - David Duberman


Spectrum founding editor: Doug Millison (www.Online-Journalist.com)


©Copyright 2005 Motion Blur Media. All rights reserved. No reproduction in any for-profit or revenue-generating venue in any form without written permission from the publisher.