Stages and Previews

I'm sure I've extolled the virtues of stages in Dragon Age modding previously, as they are an incredible useful way to save a lot of work. Using these, it is possible to have the toolset automatically generate poses, gestures and camera cuts for your conversations.  Another excellent tool is being able to select an NPC line, then go to the edit menu and select "Convert Line to Cutscene", which enables you to make short cutscenes that blend in seamlessly with the dialogue as it is happening.

When using stages, it is useful to be able to preview the dialogue without going into the game.  So select the root node of your conversation and click on the cinematics tab, and select the specific stage blueprint you wish to use. You also need to type out its (unique) tag when you're using multiple copies of the same stage for different conversations within the same area. Finally, fill in the relevant positions for the various actors in your conversation. Note that your choices for both the cinematic and preview tab (which we'll be using next) will propagate to all children nodes, meaning you can set up the entire conversation by just editing the root node. However, you can change the stage (and preview) for specific nodes under the root node should you so desire.

Once this is done, you can then set up the preview tab so that you can configure a stage to view the conversation inside the toolset.  However, this part is a little bit buggy in that it doesn't deal with updates to the stage or its location within its parent area.  If you're using a generic stage and trying to position it such that you have particular items in the background, this can be a significant problem. The trick is that using "First Match" doesn't update the stage location for your preview, as it seems to remember the location the stage was in when you first selected it.  It also doesn't update the numerical value of the stage's location in the drop down menu.

As such, if you adjust your stage's location in the area, while still in the area editor with your stage selected, highlight its location parameter, hit Ctrl-C to copy it to the clipboard, and remember its angle value (to the first decimal point should be all the granularity you need).  Then go back into the preview tab of your conversation, Ctrl-V paste the location over the existing location, and then manually edit the angle.  You should also select the appearance of your various actors in the preview screen, particularly if your dialogue involves anyone not of "standard" human/elf height.

With all of that done, you can now right click on lines in your dialogue tree and hit "Preview Line".  You'll also want to select the safe-frame button (the square below the playback buttons) so you can see what it will actually look like in the game itself.  And no, I have no idea why that isn't set as the default option, as I can't see any time when you wouldn't want it to be on. Even in an ideal world where I have multiple 24" monitors on which I can float various windows, I'd still want to see the framing of the shot.

I should quickly cover the "auto-generation" I mentioned earlier.  Again, we want to go back to the root node. But this time you want to right-click on it. The three options you're interested in are: "Generate Poses for Children", "Generate Gestures for Children", and "Process Cameras for Children". Voila! Instant semi-cinematic dialogue.  Unfortunately, the system isn't perfect, and on some lines you'll have NPCs gesturing (sometimes wildly) when they aren't talking, and you might not like some of the camera shots... but it still saves a lot work you'd otherwise have to do manually.

One thing I'm still confused about with stages is the use of "functional shots" as described in the wiki. I have no idea how these are implemented, despite the information presented there. If anyone knows anything more about these, or has used them, I'd love to hear from you.

Know When To Stop

When you're doing modding, it's essential to keep in mind when to stop fiddling with something. This can particularly be the case when it comes to level design.  The temptation to keep tweaking landscaping, texturing, decoration, lighting, and all the other elements that make up a level is extremely high.  As such, I've forced myself to consider the level that I've been working on complete.

It may have some minor tweaks done to it once I start working on the actual area itself, populating it with creatures, items and various interactive items, but for the moment it is "finished".  As I am somewhat of a perfectionist, making this decision to stop working on something is a difficult one. There comes a time when the extra effort of tasks like finessing the exact location of vegetation becomes more an exercise in obsession rather than useful design.

Unfortunately, I don't believe I have any hard and fast rules to determine when it is time to stop. It is going to vary from person to person, and potentially from task to task.  Face morphs in particular are something I find very time consuming, especially when I'm aiming for a particular type of appearance.  It probably doesn't help that I don't seem to be very good at doing facial morphs. I have been envious of the skills of people in the "prettiest character competition" thread on more than a few occasions.

But, as a general rule, if you find yourself going over the same ground for a few hours and making repeated tweaks on those final touches to make something "perfect", or second-guessing your work, then it's probably time to stop.

Anyway, it's time for a small look at the completed level in-game:

Given the number of screenshots I've posted of this area, it feels as though a short description is in order. The following is part of the codex entry on the area:

Patren's Gorge
A rocky scree bisecting the North and South of Patren's Wood, Patren's Gorge is a particularly dangerous area high within the Frostback Mountains. Named for the Ferelden explorer Patren, who in 4:89 Black, made an attempt to explore and map as much of the Frostback Mountains as possible. The previous Orlesian invasion of Ferelden prompted the Arl of Redcliffe to send Patren to explore other possible routes across the Frostback Mountains.

Types of Choice

In my recent articles about replayability, I have been talking a lot about choice in games. I have been focusing primarily on RPGs, but that's because when it comes to giving players the ability to make a choice that will affect a game world, RPGs are pretty much the king of genres at the current point in time. Even so, the ability to make big choices that will affect the outcome of a game are still a relatively new thing. Note that here I'm focusing on choices that have consequences rather than purely "cosmetic" dialogue decisions that have no real effect on events.

There is still a long way to go before we have game worlds that truly react in a seemingly organic fashion to our actions, at least in a single player experience. Regardless, designers should carefully consider the types of choices they present to their players. By using multiple decision types, the designer can develop the player's roleplaying experience in different ways and offer a more believable gaming world. Here are five key types of choice that are presented in RPGs:

Good or Evil
The classic dichotomy of Good versus Evil, Light Side vs Dark Side, Law vs Chaos, etc, etc. Two diametrically opposite schools of thought, and the player must pick one. These are generally straightforward for characters/players, and help provide a basic disposition for the character.

The Blind Choice
With very little information, the player is forced to make a decision one way or another. This forces the player to go with their gut instinct rather than attempting to overthink or "meta-game" a decision. These can be very effective in getting a player "into their character's head" if used well, but also pose the danger of frustrating the player by confusing them through lack of information. They are also less effective in subsequent playthroughs, as the player has additional information to guide their choice the second time around.

Pick One Benefit
The player is presented with a number of different options, and picking one will provide a different positive effect depending on the decision. This is typically a means for a player to further their character's development, or further their personal relationship with a character. This could be as simple as choosing to support one party member over others in a verbal argument, or making the decision of where to allocate resources to rebuild a castle.

No Good Option
The opposite of the above. Some of the best choices that I feel have been presented in video games are when I have had to pick between a lesser of two evils, or make a choice where neither choice is desirable. BioWare and Obsidian are the only two developers I can really think of off the top of my head who have presented choices like these. I won't give specific examples because of the spoilers involved, but if you've experienced them, I'm sure you'll know what I mean.

Grey Morality
Possibly a superset of "No Good Option", a "grey" choice is when there is no clear good or evil decision to be made. Either options has its benefits and drawbacks, and arguments could be made for either side. These can be very hard to craft for a player, and will often cause large divisions in player communities - case in point is Loghain in Dragon Age, where people both staunchly oppose and support him with their own reasoned arguments. Some would argue that it is not a "grey" decision at all - on both sides - which simply serves to demonstrate that it is.


If you have any other categories of choice you find interesting or effective, please comment and let me know!

Cliffs = more usable space

In my continuing look at level design and texturing, I've got two more screenshots to demonstrate how to improve area layout. When designing an area, the usable space is something that is very important. This is the area that is able to be used by the player or other characters for movement, or for other interactive objects. In general, you want to maximise the usable area within a level.

In exterior levels, hills and mounds serve multiple purposes. Firstly, they limit the player's movement, forcing them along the set paths that you the level designer want them to travel upon. Making the player travel through a level in a particular path allows you to deliver information in a set sequence, which is highly useful for plot development. They also serve to reduce visibility. Yes, this is a good thing, because it prevents the player from seeing too far ahead, (never underestimate the power of the reveal) and it also reduces the polygon count on screen, thereby improving performance.


However, looking at the above picture, there is a distinct problem (aside from the muddy texturing). The problem is that the amount of usable space is a very low percentage of the available space because of the rocky hill/mound in the main part of the shot. It segregates the walkable areas nicely, and does a good job of blocking of the players view, but it occupies an awful lot of room. It also looks a little funny from a landscape point of view, such mounds aren't really all that common, at least not quite so thin and evenly peaked. The solution is to simply turn it into a cliff, like so...


Even ignoring the obvious texturing improvements, the area is now a lot more visually appealing and significantly more practical in terms of usability. The open space provides a good vantage point from the top, while still maintaining a somewhat claustrophobic feel in the tight canyon at the foot of the cliff. The amount of space has increased dramatically, allowing this to be a reasonable area for a fight or a cutscene. There's also increased room for vegetation, which is a very important part of the overall appearance of a level.

The only drawback is the extra effort required to make the cliffs look organic, as lots of straight lines do not exist in nature, and tessellation is required on the cliff faces to increase the vertex density so it is possible to introduce more variance into its shape. In this case, the payoff was more than worth the additional work required.

Blending: Take 2

So after my article on texture blending, and the two comments (I'd love more opinions on all of these posts... really!) I received, I thought I'd post a direct comparison of texture blending on my level.

The screenshot that I posted last time was like this:


But I've since reworked this section of the level as a small trial to determine whether the stronger texturing look is better. As such, here is a second comparison shot:


So which do you prefer? And why?

Reviews

Unfortunately I haven't been able to do much modding today. However, in order to get another update (yes, I'm trying to do them more regularly) I'm doing to do two reviews.

The first thing I'm going to discuss is Return to Korcari Wilds, which I mentioned in passing a few days ago. Now, given it's been out for a month, and already had more downloads, just as many endorsements and twice as many comments as Alley of Murders in one quarter of the time, I was expecting something pretty special. Now, I'll happily admit that I'm jealous of its statistics, but I'm more disappointed by the fact that it doesn't deliver.

Put simply, Return to Korcari Wilds is a 'dungeon' romp. It's a short and linear adventure where you kill a whole bunch of darkspawn. You're expected to encounter the three sets of NPCs in a set order - and in fact, the third one doesn't even appear until you've talked to the other two. There's no real roleplaying to speak of, and the dialogue is filled with spelling and grammatical errors. Yes, I understand that the creators speak Spanish as their first language, but there is barely a sentence that doesn't contain an error.


The combat of the mod seems like it would be very challenging for the most part. I played with a fairly high level warden and got through the majority of the fights without too much hassle, but I imagine it would be nigh on impossible for someone attempting it early in the game. An abundance of boss monsters, often with two in a fight, means that it would be very challenging.

There are a few cutscenes, notably featuring music from Prince of Persia: Warrior Within, which are reasonably well done. There are a number of pieces of custom loot, which are balanced apart from a few overpowered items; a two-handed sword with +30 spirit damage is the worst offender here. There are two companions, one of whom has a nice side quest to kill a Dragon, but again the quality control is lacking. See the mystical floating mountains of Ferelden...


However, all this said, Return to Korcari Wilds isn't a bad mod. For a straightforward combat-fest, it's fine, but the overall lack of polish made it trying for me. Given how well it has been received, I was expecting to be impressed by an adventure that integrated smoothly into the main campaign, but I was ultimately unsatisfied. This isn't to say I don't appreciate the effort that's gone into creating it, and I'd happily encourage the modders to improve upon their efforts. However, I suggest you play it expecting just to kill a bunch of enemies. With that expectation, you'll probably be happy.

The second thing I'd like to review is Warden's Fall, a short Dragon Age machinima. It's apparently going to be a series of videos telling the story of the Warden Kristoff, who seemingly takes his duty of killing darkspawn very seriously. The production values on this work are extremely high, and the people involved deserve credit for all the hard work they've put in. The only real complaint I have is that the directing during action suffers a bit from Michael-Bay-ism, or rapid-fire shots that reminded me of the second Bourne movie. However, I only comment on that because I can't find any other really constructive criticism at this point in time. The music, voice acting, cutting and sequences are well crafted and put together nicely. Seriously, go watch it if you haven't already.

I'd love to get some of the animators doing cutscenes for The Shattered War...

PS Please, give me your opinions on the texturing post I made yesterday, and vote in my poll!

Texture Blending

Despite my recent posts regarding how I don't replay games, a series of events led me to find Matrix Cubed. This game is a nostalgic favourite for me. Looking at it now, the graphics are bad, roleplaying choice near non-existent, combat simplistic... but I can't stop help feeling it is fun.


When you've grown up with something and played it dozens of times over (I played it so much that I hit a bug that prevented warriors from levelling up due to not being able to allocate a weapon specialisation point), there's something about it that means you overlook its many flaws. If I could determine what causes the feeling that keeps you playing a game even though you can see it is somewhat shallow... well, I imagine I'd be selling that secret to gaming companies around the world.

Anyway, I haven't been entirely wasting my time playing Matrix Cubed. I've actually been doing quite a lot of work on an area for The Shattered War. However, looking at it in comparison to BioWare's original levels, I think I've perhaps done a little too much texture blending. What do I mean? Let's do a quick comparison.


The above screenshot is from a portion of my level. This level is very rocky, and as such I've used quite a number of textures in order to stop it from looking too monotonous. BioWare's level artists are typically very strong in their use of textures, whereas I've done a lot of blending. That's not to say that BioWare's artists don't blend, as they carry out blending very well, but they are willing to use a large section of the same texture at 100% and then blend small patches of a related texture. See the reference screenshot below:


Looking at my level now, I wonder whether I could or should have used less blending. I think the level might have been to boring if I hadn't tried to use that many textures, but at the same time, I realise it is something I may need to keep in check in future. As it is said, less can be more.

I guess I'm curious to hear from people as to which they like more? Do you like the complexity of a landscape with a lot of textures, or do you feel it gets far too cluttered and you'd prefer a more simplified approach to texturing?