Ok, now as to this castles thing. I suppose we’ve got enough interest to actually start something, so.

Unless everyone decides on a better name, we’ll fold this under the AFK_ umbrella and call it AFK_Castles. Goal is to give more population to the castles and expand the interiors a little bit. Quests and anything major we can get to later.

Thinking division of labor thusly:

Dwip – Lead, AI, scripting, probably dialogue, general purpose whatever.
Sigurd – architecture and cluttering.
Hanaisse – Upper Class clothes retextures.
Conner – Enthusiasm and ideas.

And we can all chip in to develop the various NPCs and whatnot.

Figure the best way to do this is for me to keep a master ESP file, where I’ll deal with adding everything. If Sigurd’s going to all the architecture stuff, he can keep a seperate mini-esp with just the buildings and any custom statics in it, and I’ll Gecko merge it in from time to time. If the rest of you are just doing NPC facegen and such, probably the best way to do that is to have you do mini-esps where you create an NPC, do the face up, add basic and non-custom equipment as necessary, then pass it along to me so I can add it in.

Would like to keep things as relatively vanilla as possible, so would like to keep away from the custom hairs and such, with the exception of the clothing stuff, since there’s no way we can get by on the same 5 outfits. This can be negotiated, but I’m going to maintain a veto on aesthetical choices.

As to the NPCs themselves, let’s say 4 per castle for Anvil, Bravil, Bruma, Cheydinhal, Chorrol, and Leyawiin. Unless we come to a consensus otherwise, I’d say no Skingrad to keep with the nature of the Count and the place. That’s 24 NPCs. Either in addition to that or as part of that, we should have about 4 “roaming” NPCs who spend time at least 2 courts, just to mix things up a bit.

At the least, each NPC should have a unique greeting, with a bit about who they are. Unique responses to the various town topics are also good. Also, it’s worth noting that only the Breton, Elven, and Imperial voice lines for the upper class rumors are voiced. That may need to be worked around. Or not.

Architecturally, each castle should have an “X Castle Guest Quarters” cell with space for 4 NPCs and attached at the very least to the County Hall in some way. Sigurd’s the one with the castle architecture ideas, so I’ll leave that mostly up to him – towers, rooms, whatever. Figure if they’re rooms, they should be about the size of a pair of CastleInt1Way pieces stuck together. Secret passages, smaller courts, connections, clutter, whatever I leave to discretion. Within reason, if statics need to be created I can probably handle it.

I open the floor for ideas and suggestions.

[EDIT]

NPC List:

Anvil
1. Colin Pinder (Imperial Male) – Refugee noble from Kvatch
2. ??? (Breton male) – Provincial guy who wants to move in on Milona Umbranox
3. Admiral Graysin Umbranox (Imperial Male) – Has a treasure hunt quest.
4. Demetria (Bosmer female) – Sent from wherever to refine herself at court, spends time fighting instead.

Colin is involved in this quest revolving around the ghost of Antus Pinder and an heirloom. See same and following for Graysin’s quest.
Demetria here.

Bravil
1. ??? (Imp/Bret Female) – “Intended” for Gellius Terentius
2. ??? (??? Male) – Boon companion of Gellius Terentius
3. Rutilius Denius – father of Calidia
4. Licinia Denius – wife of Rutilius

Possible quest here to either clean up Gellius or debauch him. And/or get him married.

Bruma
1. Karn Silveraxe (Nord male) – Rustic chieftain type, helps out in the Charus quest.
2. Ragnar (Nord male) – Skyrim ambassador. Beard from here: http://tesnexus.com/downloads/file.php?id=8393 Also has Snorri Skullshatter, faithful huscarl.
3. Calidia Denius (Imperial female) – wife to be of Storig Masson, part of indeterminate love polygon.
4. Storig Masson (Nord male) – husband to be of Calida, from Rifton in Skyrim.

Option for Ragnar to join in the MQ gate closing festivities.
Elaboration on Calidia here.

Cheydinhal
1. ??? (Dunmer male) – Obligatory Morrowind tie-in. From Mournhold perhaps.
2. ??? (Dunmer female) – Girl who’s all a-flutter at Farwil’s dashing heroics.
3. ??? (Orc male) – Mafia don type with ties to Orum gang and DB. Wants to assassinate #4.
4. ??? (??? ???) – Good-guy type who wants to investigate the orc guy and put him in jail.

Orc/investigator stuff here.

Chorrol
1. Petrius Valga (Imperial male) – Husband, cousin of Charus Valga
2. Elenora Valga (Imperial female) – wife of Petrius
3. Arim Relph (Redguard male) – Horse breeder.
4. ??? (Imperial Male) – Wannabe vampire

Dynamic where cousins remind Arriana Valga of her dead husband, so she hates them but won’t send them away because they’re reminders. Discussed here.
Issues with Alessia Caro not eating dinner with dear old mom.
Possible quest here for wannabe vamp to either infect him or “cure” him.

Leyawiin
1. ??? (Khajiit ???) – Possible khajiiti ambassador
2. ??? (Argonian ???) – Possible argonian ambassador
3.
4.

Floating
1. Adrius Vanarius (Imperial male) – Steel-armored paranoid who’s convinced the emperor’s assassins are after him, too.
2. ??? (??? ???) – has mining interests in Chorrol and Bruma.
3. Florius Canstavo (Imperial male) – Elder Council rep checking on Legion resources, including city guards. Huge Ocato gossip if disposition is high enough.
4. ??? (Altmer? ???) – AU dropout, has been hiding it, playing con games in various courts and moving on.
5. Nemeus Velopiscus (Imperial Male) – Mael’s crazy alchemical merchant scheme guy.

14. Next, we need to apply a texture. But first, there’s some prep work we need to do.

Make sure your axe is selected in Object Mode, then change over to Edit Mode. Before we do anything else, we need to seperate the wooden parts of the axe from the metal parts. Using face select mode, select all the faces of the handle, including the top part. Hit the P key to seperate the selected parts.

Now, since we still have the axe blade handy, hit A to select everything. If you’re not already there, hit F9 to bring up the Editing panel in the lower pane. In the panel named Link and Materials, you’ll see a section called Vertex Groups, with a dropdown arrow next to it. Click that, and select Default:

We’ve just added one of the materials from the iron battleaxe we imported, which will save us from having to go through the hassle of defining new materials for ourselves.

Now, switch back into Object mode, select the handle, select all vertices, and repeat the process, only this time instead of the Default material, you’re going to pick the noEmit material.

One last bit before we start unwrapping. Deselect everything in the handle, and change over to Edge select mode if you’re not already there. Go to the bottom of the handle where it tapers off, and select all the edges around the top of that part. Then do the same for the top taper. Now go to Edit/Edges/Mark Seam. You’ll see the seams turn into orange lines:

This will matter in just a moment when we start UV unwrapping things.

15. Now, change the window type in your righthand window to a UV/Image Editor window:

All of the bottom menus in that window will change. There’s one that will say Image, with a dropdown button next to it. Click it and select IronBattleAxe.dds, which will appear in the window.

Now, over in the left window, you need to get to where all of the faces in the handle are selected. Now hit U and select Unwrap.

You’ll get this:

Not quite what we wanted, but it can be salvaged.

Change over to UV Island select mode:

Now select the big island that’s most of the handle. We need to fit it to the wood and leather wrap area on the image.

Rotate (R key) the island until it’s roughly pointed straight up and down, then select and drag it over into the middle of the image like so:

Now we’re ready to go to work. Switch over to UV vertex select mode (it’s two buttons to the left from UV island select). Select individual vertices and begin dragging them into position on the image. Try to get everything aligned at right angles. You don’t need to be perfectly exact, but try to be pretty close. This is going to be a pretty painstaking process, so just zoom in and keep at it.

When you get done, you should have this:

To make sure it looks ok, go back over to the left window, drop into Object mode, and take a look. When you’re done, go back into Edit mode and select everything, and we’ll continue.

You have three pieces of handle left to deal with – two fan-shaped pieces and a rectangle. The rectangle is the easy one. Use UV island select mode to drag it over onto the handle, the S key to scale it a bit, and move it into position just like the other part. After the other piece, this one should be a snap.

Now for the other two pieces. We’re going to make our lives easy, here, and not try to make a perfect half-circle out of all those vertices.

Instead, go back over to the left window, and change over to front view. Zoom in on the bottom tip of the handle, and select just the unmapped vertices at the tip, like so:

Our next step, as you can see, is to hit the U key and choose Project From View. Watch in the UV window as we suddenly get a perfectly formed half-circle that we can scale and move into position.

Now, tap either the 4 or 6 key on your numpad precisely 12 times to switch over to the other side of the axe, and do the exact same thing to the head. Now use the A key to select everything again so we can see it all. Now go back over to the UV window and use scaling and dragging to move everything into position on the map. Scale things slightly smaller than the circles on the image, just to make sure we have enough room when we add the other half.

It should all look like this:

From here, you should be able to do the axe blade on your own. Remember to go back to Object mode to select it, and use Unwrap to unwrap it. It will probably be at a weird angle and much too large, so remember to rotate it and scale it, then move it over to some unused part of the map. We’ll be redoing this part of the texture later, so it’s not particularly important where it goes, just so long as it’s not touching the wood parts.

Here’s about what you should have:

We’re going vertical with the blade so that we can make the UV island a little larger, thus getting a little more detail in. We could get even more in if we took the trouble of moving one or another of the wood ends to the other side of the map, but this will do.

One more step and we’re done. This next bit will be a semi-repetition of this tutorial. If you’re confused, refer to that tutorial, where I talk about this process in slightly more and different detail.

In the right hand window, go to the UVs menu, and choose UVs/Scripts/Save UV Face Layout. Choose your options so it looks like this:

Pick a location that you’ll remember, and save. Now, save your work and minimize or close Blender.

16. And now an interlude with texturing. Fire up GIMP or your favorite image editor (I’ll be describing things in GIMP). Next, open ironbattleaxe.dds from Data/Textures/Weapons/, and also open your UV face layout .tga you saved out of Blender. Make sure not to load mipmaps when you load ironbattleaxe.dds, as they’ll just get in our way.

In the .tga you opened, go to Layer/Transparency/Color to Alpha, and make the color white. Hit ok. Then, use the select by color tool (top far right in the toolbar) to select the transparent area, and choose Select/Invert. Now choose some high contrast color like magenta, pick a large paintbrush, and paint over the whole selection to change it to that color. The reasoning behind this will become obvious in a moment.

Now select the entire image and Ctrl+C to copy.

Now switch over to ironbattleaxe.dds, and choose Edit/Paste As/New Layer. Now we know where to put our blade when we get there.

Before we go any further, however, hide the layer you just pasted, change your brush color to black, and pick out a moderate-sized paintbrush. If it’s iron, paint over it – it will just get in our way and confuse us. When you get done, show the layer you hid. You should wind up with this:

You can go ahead and close the tga file now. You won’t be needing it. While you’re at it, save ironbattleaxe.dds as a .xcf (or .psd or whatever your native file type is) so you won’t accidentally overwrite the real thing later.

Now, hide the image layer so you just see your magenta UV layout layer. Make sure you’re working in the UV layer, select the magic wand tool, and click in the transparent area to select it. Then invert the selection (Select/Invert), and grow it by 2 pixels (Select/Grow). Now unhide and switch back to the image layer, and hit the Delete key. The blade area will become transparent. So far so good. At this point we can hide our UV layout layer, as we’re most likely done with it.

There are a bunch of different ways to do this next bit, but I’ll show you an easy one. Find a texture you like for the blade somewhere. I like to steal pieces of existing Oblivion textures when possible, but in this case none of the iron pieces are quite right, so we’ll use one from an internet texture site. One of these, say.

Once you find one you want, load it up, and scale it down so that it will in a 512×512 area (the dimensions of our workspace) if necessary. Try not to choose a texture you need to scale up, as it won’t look quite as good. Once you’ve manipulated it how you want it, copy the whole image, and paste it as a new layer into your axe image. Then move the layer down so that it’s behind the main image. If you did it right, you should have something like this:

Now let’s get fancy. Suppose we don’t actually want that brownish wood in the handle. Maybe we want something lighter.

Select all of the various wooden parts of the handle using the box select tool. Then, go to Colors/Desaturate to make the parts greyscale:

Now, change your color to something else. We’ll use #cfbe9d, which is a nice light brown. Now create a new layer with the foreground color. Make sure it’s on top. Your screen should be brown. This is good.

Next, with the new layer selected, use Select/Invert to invert the selection, and hit Delete to delete everything outside of where we’re editing. You should have slightly less brown. Now, over in the Layers toolbar, where it says Mode: Normal, change it to Overlay. And now we’ve changed the color of our wood. If we wanted, we could do the same sort of thing to the leather wrapping, or even to the blade.

When you’re done, save your .xcf image. Then, delete the UV layout layer, and right click in the Layers toolbar and select Merge All Visible Layers. Then, go to Layer/Layer to Image Size. You should now have a single layered image. Save it as a .dds file, DXT1 compression, with mipmaps. Put it somewhere in Data/Textures.

Now, you need to make a normal map. I’ve already written a tutorial on the subject, so rather than summarizing here, I’ll just direct you there instead. When you’re done, save your normal map in the same spot as your texture.

Our interlude is over, and it’s time to go back to Blender.

17. We need to change the image our weapon is using. The easiest way to do that is to go into Object mode, and click the blade. Change to the Shading Panel (F5), and to the Texture buttons (F6). There’s a panel called Image in there, with a file path pointing to ironbattleaxe.dds. Change it to point at our new .dds file instead.

Now, go back into Edit mode and make sure all the faces are selected. You’ll notice the UV window on the right has a big red bar over the image file name. Change it to your new image, and it will suddenly become happy. You’ll also see your texture appear on your axe.

Due to the way we’re sharing materials here, for the handle all you need to do is go back to Object mode, select it, go back into Edit mode, select everything, and change the image in the UV window. Now you have a fully textured axe. Make sure the UV mapping is all ok.

18. Now for the fun part. Go back to Object mode. Select the blade, then change the bottom panel to the Editing Panel (F9). Way over on the right is a panel called Modifiers, which says Add Modifier on it and has a dropdown. Change it to look like this:

You want the X and Y mirrors to be unchecked, and the Z mirror to be checked. If you check in the left window, you’ll see you now have the entire blade. Pretty cool, huh? You’re not quite done yet, however. Make sure to click Apply in the Modifiers panel so the change will take effect.

Now do the same thing to the handle. Now you have the whole axe!

19. We’re almost done. Now make sure you’re in Object mode. We don’t need our reference axe anymore, so go ahead and delete both parts of it.

Remember waaaaaaay back when I told you to move the collision to another layer because we’d need it later? We need it again now. Go over to its layer, select both parts, and move it back to the main layer where your axe is. So far so good, but you’ll need to fix the topmost collision box to cover the entire blade. Scoot it over, then select individual side faces in Edit mode and drag them around to shrink the box down.

Here’s our final axe, ready for export:

20. Time to get out Blender. In Object mode, select everything (handle, blade, both collision meshes), and go to File/Export/Netimmerse/Gamebryo (.nif) and save your new weapon .nif somewhere in Data/Meshes/. Make sure that you click the button to make it a weapon, and change the material to metal. Since this is a two-handed weapon, change the Weapon Body Location to Back. Now export.

You’re done with Blender now, so save and exit.

Now open up your axe in Nifskope. Make sure everything is the way it ought to be. It’s useful to have the iron battleaxe nif open for comparison here as well. In our case, there are a few problems:

We need to get rid of the two extraneous NiStencilProperty blocks, and it turns out that for some reason Blender exported both halves of our object with the same material property, which won’t do. The collision also looks pretty strange compared to the iron axe, but it probably doesn’t matter. If it does, we can always come back and fix it.

The NiStencilProperty blocks aren’t really hurting anything – they just make both sides of a face have a texture – but they’re in the way, so we may as well delete them. Right click one of them, then choose Block/Remove.

It removed both of them, which leads us to our next step. Go up to Spells/Optimize/Split Properties.

Now, make sure you have the iron axe .nif open. In that nif, find the NiMaterialProperty for the blade, right click it, and choose Block/Copy. Then go back to your axe, find the NiMaterialProperty for its blade, right click it, and choose Block/Paste Over. Then do the same thing for the handle.

Now we’re done here, so click the Save As button for your axe, and save over itself. You now have a complete axe that can be put into the game. You may as well go ahead and make the icon for the axe while you have Nifskope open, too.

Once you create your weapon in the CS, make sure to drop a copy of it somewhere handy so you can go check it out. Besides being awesome, you’ll want to be able to easily check for texture and normal map problems that Nifskope doesn’t show.

Alternately, it’s fun to use your new creation to kill some dudes.

A Brief Addendum Regarding Swords

For the most part, swords, daggers, and other scabbarded or one-handed weapons are the same as what I’ve described in the tutorial, with a few exceptions:

In Step 5, I talk about the 3D cursor (red and white circle) being where the center of the hand goes. For one-handed weapons, this is not always the case. There will also be a set of black arrows roughly in the center of the handle somewhere. This is where the center really is.

Swords are generally constructed as a single object – the axe was only two objects because we needed it to add materials. If your weapon, whatever it is, will only have one material this is irrelevant.

If you want a scabbard on your weapon, follow these steps:

- Scabbards should be seperate objects from the actual weapon. This will become highly important later.

- Scabbards don’t have collision, so don’t give them any.

- You should probably begin work on the scabbard after you finish modeling the rest of the weapon, roughly at the end of the first half of the tutorial just before you start texturing.

- You’ll want to build the scabbard more or less like you did the weapon – add a new plane mesh, move it into position, and delete three sides of it. It’s more or less like rebuilding the weapon in the first place, so follow those steps. The difference is that the scabbard mesh should fully enclose the weapon mesh up to the hilt (or whatever style you’re working for). If you have parts of the weapon mesh peeking through somewhere, that’s bad and you’ll need to modify the scabbard to fix it.

- It’s not particularly important to give the scabbard the same material type as the weapon, but you do need to give it the same texture for the sake of efficiency. You’ll probably need to create a UV map image for it like you did for the axe blade in the tutorial.

- When you export from Blender, if the weapon is supposed to be worn at the side, then Side is the button you need to hit, not Back.

- Once you export the completed weapon + scabbard, there’s an extra bit of wrangling you need to complete in Nifskope:

You’ll need to create a new NiNode under the root NiNode. Make sure it’s a child of the top NiNode (you can change this by editing the NumChildren and Children list properties of the topmost NiNode). Make sure that this new NiNode is named Scb. Then, you need to make the NiTriStrips for your scabbard mesh a child of the Scb NiNode. If you load up any given longsword, the structure for this will be pretty obvious. Just make sure you duplicate that structure, and you’ll be fine.

I’ll be condensing and working from information contained in this Blender 3D: Noob to Pro tutorial, as well as this CS Wiki tutorial.

This one’s going to be very long, two parts worth of long, in fact, so grab something to drink.

1. First step, find yourself a reference picture. It doesn’t need to be huge, but make sure it’s of some reasonably good resolution – you need to be able to at least guesstimate as to depth later, so something well-lit is good.

Your reference picture needs to be oriented straight up and down with the hilt or handle at the bottom and the edge of the blade facing right. Fire up an image editor and rotate it if you need to. When you’re done, save it somewhere you’ll remember. We’ll come back to it in a bit.

I’ll be using this picture of a Viking axehead, manipulated a bit from the one found at Wikipedia:

Also, it’s worth noting that the Noob to Pro tutorial recommends creating three reference images: side, top, and front. For the purposes of weapon creation, top down is sufficient. For other projects (such as creatures), you may want more images.

2. Now, start up Blender, and import a weapon of a similar size to the one you want to make. We’ll be stealing bits and pieces of this during later steps. We’ll be using the iron battleaxe from Data/Meshes/Weapons/iron/.

Once you’ve imported the weapon, you’ll see some collision around the handle and the head. Right click to select both pieces, then use M to move both of them to a different layer. Don’t delete them, as we’ll be coming back to them later.

3. Next, split the view window in two. This will be important later.

4. Now, let’s put in our reference picture. In the left-hand pane, go to View/Background Image.

Click the pointless Use Background Image button in the dialogue that comes up, then click Load and go find your image. Your image won’t show up right away, but it’s there. Click or change the view or some such to make it show up. You should wind up with something like this:

That’s much to big, so we’re going to fix that.

5. In the Background Image dialogue are three sections labeled Size, X Offset, and Y Offset. Using these, move the background image so that it lines up with the weapon you imported.

You need to pay special attention to where the handgrips are. For two-handed weapons like this, the middle of the top hand will be where the origin point/3D cursor (the red and white striped circle thing) are. For one handed weapons, this is sometimes the case, but there will also be a pair of black arrows that signify where the hand goes. You can always move your weapon around later if you run into problems, but it pays to get this right at the beginning of things.

When you get done, you should have something like this:

6. Unfortunately, you can’t really see most of your image, because the imported weapon is in the way. To fix that, select both parts of the weapon, hold down Ctrl, and use the red arrow to drag the weapon off a few units. We could delete it, but it will be a useful reference for later.

At this point, your workspace should look about this:

The left window, with our reference picture, should be in top view at almost all times. The right window, which I have in front view, is where we’ll be doing anything that needs viewing the object in anything not top-down. Over the course of things, you’ll be changing views in the right window a lot – to check the reference object, create depth in your mesh, or what have you. This will make more sense once you start doing things.

7. Now it’s time to create things. Go to Add/Mesh/Plane. Drag it up until one edge of it is more or less flush with one edge of your reference picture, like so:

It is incredibly important that any time you move things around in top view that you use the red and green arrows and NOT the blue arrow or the orange circle. This can take some getting used to, but it’s important that this part of the mesh stay at exactly 0 on the Z axis for reasons I’ll explain in a bit.

8. Now, change over to Edit Mode. Using Edge Select Mode, select the three sides of the plane that aren’t flush with the reference image, and delete them:

9. Now, change over to Vertex Select Mode, and drag the vertices on what’s left of the plane until they outline part of your image. Then, using Extrude (E key), continue creating edges until you’ve outlined your entire image.

Something like this:

Notice that I’m deviating a bit from the plan: I’d rather have the axehead attach to the shaft with a circular iron piece rather than the method it currently does, so I’ve outlined where I want it to go. I’ve also moved the shaft from the reference axe up so that I have something to measure by. I probably won’t end up using some of those vertices, but they’re there so that I know what I will do.

Speaking of which, let’s do something about that.

10. If you’re working along, go back to Object mode, and select the shaft. Over in the right window, set the view to front, zoom in on the shaft, and count how many sides it has – 12. We’re going to need something with that many sides.

Add/Mesh/Circle will create a new circle mesh. Make sure vertices is set to 12, and make sure you uncheck Fill if it’s checked and then hit Ok. Your mesh will appear. It probably doesn’t look like a circle in the right window, so let’s fix that.

Make sure the circle is selected (it is by default), then change the right window to side view. Hit R to rotate, and hold down Ctrl and move the mouse to make it vertical. Now change back to front view. Hit S to scale, and scale the circle until it just fits inside the shaft’s circle. It’s not particularly important to make the rotation in this angle match – most of these vertices are going away soon.

You should end up with this:

Now switch over to Edit mode, and change to Edge select mode. Select ONLY the bottom 5 edges of the circle, and delete them. If you did it right, both edges that are straight up and down should still be there.

Change over to Vertex select mode and select the bottom two remaining vertices, then use the blue arrow to drag them up while holding Ctrl to snap to grid. Usually you can get them exactly flush with 0 Z axis, but sometimes you’ll need to go to Mesh/Snap/Selection to Grid to them there.

When you get done, you should have this:

The reason we’re doing things this way is that we’re only going to make half the weapon now, then we’ll mirror that half to make the other half, saving us a ton of work.

Almost done with this step.

Well, sort of. Make sure the entire half-circle is selected. Still in the right window, hit Shift+D to duplicate the half circle, then immediately right click to avoid moving anything. Then, use scale (S key) and Ctrl to scale out a few gridlines. Then use the blue arrow to move it all up a few gridlines. Make sure you edit the bottommost vertices flush with 0 Z axis like I described earlier. Ultimately, this will all make up the iron band around the handle.

You should have this:

Really, we’re almost done.

Now, go back over to the left window. Remember those edges we drew in the outline that I said we were going to delete later? Go back to Object Mode, select the outline mesh, go to Edit mode, select those edges and delete them.

Now go back to Object mode, click the reference shaft, and drag it back on out of the way. Then select your half-circles, and use the green arrow and Ctrl to move it up to where the axe is.

Still in Object Mode, select both the half-circles and the outline mesh, then use Ctrl+J to join them together.

Now in Edit Mode, switch to Vertex select mode, and in the right window select the bottom vertex in the outer circle on the side closest to the outline. On the left side, select the vertex in the outline closest to the circle. Now hit the F key to join them together.

The result of our efforts is this:

Now go back and select both half-circles but not where they join up with the blade outline. In the left window, hit Shift+D to duplicate, then hold Ctrl, and move the duplicates up to where the iron band will end at the top of the axe.

11. Now let’s add some depth. Find somewhere with a slope on your reference image, and make sure you know where it goes.

In the left window, go find a vertex somewhere in the outline part of your mesh near the slope, and extrude that vertex up to the edge of the slope. Now do it again to begin outlining the top of the slope. Switch back to the first edge that you made and delete it. You should end up with one edge that’s disconnected from the main outline. Now reselect it and finish outlining the whole image.

Sort of like this:

So far, so good, but it’s not very deep yet, is it? We can fix that, but we need to see how much depth to give it. Go back over to side view and pan over to your reference weapon. See how thick the blade is. We’re going to replicate that.

Over in the left window, select your entire inner outline (click one edge, then Ctrl-+ until you get it all). Over in the right window, use the blue arrow + Ctrl to move everything up a few notches until you’re reasonably close.

You should have something like this:

12. Now we fill everything in. You’ll mostly be working in the right window for this, and you’ll be rotating the view freehand for a while. We’ll do the iron band first.

Switch over to Vertex select mode if you’re not there already. For each vertex in the outer circle, find its opposite on the other outer circle, select both of them, and hit F to make an edge between them. Don’t bother with the bottommost one on the blade side, since nobody will ever see it.

Next, switch over to edge select mode. For each edge on the outer circle, select it and its opposite on the inner circle, and click F to make a face. When you get done, if any of the faces are invisible, hit the W key and Flip Normals to fix it.

Our axe thus far:

You need to keep going in this vein, making faces until you’ve covered the entire axehead and shaft. It may be easier for you to create edges first, then make faces from those. Don’t worry really hard about triangles versus quads – we’ll deal with that later.

Remember that you can subdivide edges when need be.

First step:

Second step:

Third step:

At this point, we’re done with the axe blade.

13. Time for the handle. This part is a lot easier, since we already have most of what we need. First, drag the reference handle over in Object mode so we can see it. Then select the axe blade.

Now go into Edit mode, and select all the edges of the inner circle. Extrude them (E key) down to about where the top of the black bit is before the leather wrap. See how it flares there? We need to create that.

Extrude another little bit to where the end of the flare is. Now use scale in the left window to flare the edges out a bit. However, watch out for something:

Fix the bits protruding down below there.

Now select all the edges again, and extrude down to the bottom of the leather wrapped bit. The Bethesda version of this handle is subdivided into two pieces to cut down on how big the leather wrapped part is. We may or may not get to that, depending on how we end up UV mapping the axe.

Now we need to do the bottom flare. Instead of extruding again, go back up the handle to the other thin part, and select all the edges there. Duplicate them (Shift+D), drag the duplicates down to the bottom where the flare will end, and use F to make enough faces to make the flare.

Now reselect all the faces at the bottom of the flare and extrude the very bottom of the handle out to just before it tapers off to the end. Now, extrude out again to the end of the taper. Next, scale them slightly inwards.

You should now be looking at this:

We’re going to make two triangle faces to fill in that gap. Select a vertex at one corner of the triangle, and extrude it out to very close to where the other side should meet it at a 90 degree angle. Doesn’t need to be perfect, just close. Then use F to make the faces:

Now to fill in. Select the bottom two innermost corner vertices and form an edge that should run along the bottom of the open gap. Hit the W key and choose Subdivide. Now start making faces until you fill everything in:

One last step. See how the iron axe shaft has that tilt to the bottom cutoff? We’re going to make that.

Select the innermost circle edges. Then, in the left window, rotate until you get something that looks good:

Last but not least, we need an end cap on the top of the axe. I leave this as an exercise for the reader – if you’ve made it this far, you should be able to do this, too.

At the end of the day, we have an axe:

Take a break, get up and stretch, and continue on to Part 2.

There are two good icon tutorials that I’m aware of. One for Photoshop is here, and one for GIMP is here. Both cover inventory items, but not quests.

Inventory Icons:

Before you do anything else, fire up GIMP or Windows Texture Viewer, and have a look at a few of the icons for whatever item you’re making. You can find them in Textures/Menus/Icons/. Most items are positioned a certain way, and it’s best to mimic that when creating your icon.

Having done that, let’s make an icon.

1. Open up GIMP if it’s not already open, then open up the .nif file for whatever you’re making the icon for in Nifskope.

2. In Nifskope, zoom out until you can comfortably fit the entire item in the view window. Rotate it around until you have it in the proper pose. Make sure that your background color is something that contrasts with the item, but not too strongly. Grey is usually good. High contrast colors are bad, because it’s hard to get all traces of, say, magenta out of the picture you’re going to take, leaving a halo or noise around the item. If you have the axes, havoc, or anything else showing, turn it off under the Render menu.

3. Hit Printscreen to take a picture of your item. Switch over to your empty GIMP window and hit Ctrl+V to paste. You now have a snapshot of your desktop.

4. Using the box select tool (top left corner of your toolbox), select a perfectly square area around the item. This should be exactly how you want your icon to look – don’t leave extraneous space, make rectangles, or cut the item off. It doesn’t particularly matter how big the selection is, since we’ll be resizing it later anyway. You should end up with something like this:

The green box is my selection, the red arrow is the precise size in pixels of said selection. Don’t worry about framing the selection perfectly, as you can click and drag to move it around once you get the size right.

Once you have a selection that you want, Ctrl+C to copy, Ctrl+Shift+V to paste as a new image. You’re done with the other image, and you can close it now.

Note that if you’re making icons for multiple items, there’s a handy trick you can use. Open several Nifskope windows, scale them and pose them the same (middle mouse button makes this easy), then line them all up in a row. After you screenshot the entire screen, go through this step, but don’t close the window after pasting the new image – go back to the screenshot, and simply move your selection box over to the next item and repeat the process, then stash the windows you want to work on later out of the way until you want them.

5. Now, in the new image, click the magic wand tool (second from the right in the top row of the toolbox), then click the grey area of the image. Hit Delete, and everything that was grey should be replaced by the checkered pattern GIMP uses for transparency. You should have this:

So far so good. Now, go to Image/Scale Image. Change the height and width to 64×64 pixels (just change one, the other should change automatically). Hit the Scale button. Your image just got really small.

6. Next, over in the Layers, Channels, Paths toolbar, right click under your current layer (probably called Pasted Layer), and click New Layer. Make sure it’s 64×64 pixels and transparent. Hit ok. If it appears above the layer with your image on it in the list, drag it below. Then switch back to the image layer by clicking it.

7. Now, go click your foreground color in the toolbox to bring up the color picker dialogue. Where it says HTML notation, change whatever’s there to EFE7AD, and hit ok. Now select the magic wand tool, and click the transparent area to select it. MAKE SURE YOUR IMAGE LAYER IS SELECTED, or you’ll get unhappy results. If a dotted line appears around the image, things are going well.

8. Now go to Select/Invert to invert the selection – now you have the item and just the item selected. Then, go to Select/Grow, make sure it’s set to 1 pixel, and hit ok. Now, and this is important, go over to the Layers toolbar and select your new layer, which is probably called New Layer.

9. Select the Pencil tool (looks like a pencil), and change the brush to something really big. Now color in the entire area under the item. You should see an outline appear around the item.

10. Now go back to the Layers toolbar, right click it, and select Merge Visible Layers with the default settings. Hit Merge. What you have should look like this (mine is zoomed in):

11. Now save the image. Icons MUST go somewhere in Textures/Menus/Icons/. Accepted practice is to make a folder with the name of your mod (or one with your name, then another folder with the name of your mod), and then organize your icons however you like in there. Call it some descriptive name and make sure you’re saving as a DDS file.

When you hit Save, a menu will pop up with options. Icons should be saved as DXT3, WITHOUT mipmaps – if your icon has mipmaps, the CS will complain loudly. Hit ok, and you’re almost done. We just need to create the Menus80 and Menus50 versions.

12. To create a Menus50 icon (32×32 pixels), go to Image/Scale Image, and change width and height to 32. Hit scale. Now, save it in the precise same spot you saved your icon before, only instead of under Textures/Menus/Icons/, you’re going to put it in Textures/Menus50/Icons/. Make sure you get the right directory, or this won’t work. Also make sure to not save over your 64×64 icon. If you DO save over your 64×64 icon, just keep hitting Ctrl+Z until you get back to it, then save again.

13. Now hit Ctrl+Z to undo the scale change. It’s time to make the Menus80 icon. Same deal as the Menus50 icon, except a different location and a bit more annoying.

14. Use Image/Scale Image to change the size of the icon to 51×51 pixels. Then, go to Image/Canvas Size, and change things back to 64×64 pixels. Your image should now occupy the top left corner. This is good. Hit Resize. Now go to Layer/Layer to image size. If you don’t do this, you’ll get an error when saving. Now save the image. Same deal as before, but it needs to go under Textures/Menus80/Icons/.

Now you’re done. It should be noted that it’s not entirely certain if you really need to go through the motions of creating the Menus80 and Menus50 icons, but it doesn’t take long, and won’t hurt anything.

Quest Icons:

Rather unlike the inventory icon process, creating quest icons is a lot more of an art form. What this means is that while I’m going to tell you some things here, you should treat them as guidelines or helpful suggestions, rather than laws of the land.

That having been said, to start with, open up GIMP, and load up a good 3-4 quest icons, which live in Textures/Menus/Icons/Quest/. The Settlement Quest icon is particularly good for this because it has a nice border we can steal.

1. With Icon_Settlements_Quest_Icon selected, go to View/Zoom and pick something nice and big like 4:1. Now, using the box select tool, select the area inside the border, roughly like this:

Note that the dark area around the border is NOT included in my selected area.

2. Now, use Select/Invert to invert the selection, Ctrl+C to copy, and Ctrl+Shift+V to paste to a new image. This will one day be the border to our icon.

3. Still in Icon_Settlements_Quest_Icon, select the eyedropper tool (second from the right on the second row), and click somewhere in the sort of medium blue area near the top to change our foreground color, which we’ll need in a second.

4. In your new untitled image with just the border, create a new transparent layer, and make sure it’s under the Pasted Layer. Select the bucket fill tool (looks like a bucket), and with the new layer selected, click to fill it. Everything will go blue, and you should have something like this:

This is going to be our base image. We’ll come back here later. For now, we need something to put in the center. There are a couple of rules of thumb here:

- You want a balance between blue and tan in the image. Too much empty blue space looks bad, but so does too much tan. Pick something that will be able to balance these out.

- At the same time, don’t get too intricate. This is a 64×64 icon, and you simply can’t get a lot of detail at this resolution.

5. That having been said, go find some stuff you want to make up your image. There’s no particular need to draw anything here – symbols from in-game, clip art, and the like can be combined to make a good icon image. For our purposes, we’re going to use .nif models – ARStatue01 and ARArch02.

Just like you did for the inventory icons, open up your model in Nifskope, get it to a zoom distance and angle that you want, and Printscreen and Ctrl+Shift+V to get an image. Then copy out the bits you want and magic wand + delete to get rid of the background. Unlike inventory icons, it’s not overly important to make things precisely 64×64 – if you need something smaller, bigger, rectangular, whatever, do it so long as the image you wind up with will fit in a roughly 55×55 interior area.

As you create and resize new images, go back to your base image and use Edit/Paste as new layer to paste things in. Reorder the layers in the order you want them to appear in. You may have to do this a few times in order to get something that looks ok.

What I wound up with was this:

Make sure to keep all your old image windows open all during this process, because there’s a good chance you’ll need them again.

6. Now you’re going to create a whole bunch of new transparent layers. For each actual image you paste in as a layer, you’re going to need 3 transparent layers, organized like so:

- Transparent Layer #1: this will eventually be for blue detailing.
- Transparent Layer #2: this will be for making a tan version of your image.
- Your pasted image.
- Transparent Layer #3: this will be for dark blue shadowing later.

If it helps, you can rename layers by double clicking their names in the Layers bar. You can also hide layers by clicking the eye icon next to a layer. This will become important later.

When you get done, you should have something like this:

This looks like a lot, but taken in steps it’s not bad.

7. Now, hide everything except one of your image layers and its tan layer. In the statue layer, use the magic wand to select the transparent area, then Select/Invert to invert the selection. Now switch to the tan layer. Remember this bit from the inventory icons? Yeah. Now, change your foreground color to some sort of tan – e9d6b0 works well enough, or click somewhere in the brighter tan area of one of the icons you opened for reference with the eyedropper tool, and give yourself a nice large paintbrush with the paintbrush tool. Color in the entire selected area over your image. If you get any color outside the area you want, use the eraser tool to fix it.

8. Now, use Select/Grow to grow the selection by 1 or 2 pixels. Use the eyedropper tool to select a pretty dark blue from one of the shadowed areas of another icon (but not the darkest blue), then change to the paintbrush tool. Change the opacity to 50 or 60. You want a small brush for this, Circle 05 or so.

Now switch to the shading layer under the image (make sure to unhide it), and outline the image with blue. It’s ok to have some fading and blotches in spots, but try to keep things reasonably uniform.

Now repeat the process for the other images. When you get done, you should have something like this:

So far so good, but that’s way too much tan. We can fix that by adding in detailing.

9. Pick an image, either image. Hide the tan and shadow layers, make sure the image is visible, and unhide and switch to the details layer. Switch to the paintbrush tool at a very small brush size (usually 1) and 80 opacity or so. Find dark, regular areas of the image. Joints between bricks, wooden posts, anywhere you can break up the tan. In my case, I’m going to be using the arch joints and the wing feathers of the statue.

Run the brush a pass or two over the areas. Try not to make entirely straight lines – click a couple of times along the sides to make things uneven if you need to. Every so often unhide the tan layer to see how you’re doing.

And here’s our mostly complete image. Still a bit too much tan in the arch though. We can be a bit more subtle in fixing this.

10. Create a new layer, and position it between the tan and details layer. In this case, I’ll put it between the arch tan and details layers. Hide the tan layer and switch to your new layer. Go to the color picker tool and get a slightly darker tan shade. Then, get a mid-sized paintbrush (Circle 05 or 07 work well) and change to 50 or so opacity. Then find areas in the original image you might create a few blotches, and do so. In my case, the mossy areas of the arch are perfect. You may want to use the magic wand to select the transparent parts of your tan layer, then invert the selection to use as a guide.

Sometimes you just have too big of a tan area and need to rethink what you’re doing, but for the moment we’ll believe otherwise.

Ultimately, your final icon might look like this:

From here, make sure all layers are visible, merge visible layers, and save as .dds in Menus, Menu80, and Menus50. Remember to do the Layer/Layer to image size command, or you’ll encounter problems. You might also wish to save a copy with layers unmerged as an .xcf file somewhere OTHER than Menus so that you have a backup you can work on if need be.

Two plugins of note:

GIMP DDS Plugin

GIMP Normalmap Plugin

First, the brief description of normal maps, then I’ll talk about a good process to create them in GIMP.

Normal maps are important in an Oblivion context for two reasons:

1. They act as a heightmap for a texture, allowing it to appear as if it had some semblance of depth. There are a couple of different ways to get this information, and I’ll be demonstrating one of them.

2. They act as a reflection map for a texture, controlling how it reflects light. This is based on the alpha channel of the texture – white alpha channels are highly reflective, black alpha channels are nonreflective. This turns out to be a problem because most Oblivion textures lack an alpha channel by default, thus GIMP makes them white. We’ll come back to this.

Before we go any further, I want to note this tutorial, which is fairly old and out of date in many ways, but contains a useful Appendix discussing various filter types for the normal map plugin.

This image is a basic texture, referred to in some places as a colormap.

This, meanwhile, is the normal map that comes with the game:

There’s two basic problems here. The first is that this map doesn’t even really go with the texture – try to line up the various cuts and things, and you won’t be able to do it. The second is that it’s completely solid, which means extraordinarily high amounts of gloss. If you look at this texture in-game, it will be blindingly bright.

So, obviously, we’re going to need to make our own. But first, we’re going to need to talk about how the normal map plugin does things. With a base texture open, go to Filters/Map/Normalmap. If that’s not there for you, you need to install the plugin correctly.

A window will appear. It will have a blue square in the preview window, and a bunch of options:

- Filter can be several things, and starts as 4 sample, which is a low detail map. 3×3, 5×5, 7×7, and 9×9 get progressively more smooth. Somewhat contra to the tutorial I linked, I prefer 5×5 or so for a fairly detailed map.

- Scale is another determinant of how detailed the map is, with higher scales being progressively more detailed. I generally prefer values of 5 to 15 for fairly detailed maps.

- Height Source can either be Average RGB or Alpha. Almost all of the time you want this to be Average RGB.

- Alpha Channel is where we’re going to do some things with height maps later. For the moment, we can ignore it.

So, supposing you set up a 7×7, scale 10 map. Hit ok. You’ll get a map that’s more in tune with what this should look like:

The only problem is that it’s going to be astonishingly shiny in game. For some things, that’s perfectly fine, and this is as far as we need to go. For other things, like terrain, this would be impossibly bad, and we’d end up with glowing ground.

Fortunately, there’s a way to get around this, but it’s fairly long and involved.

1. Open up your base texture. When prompted to load mipmaps, don’t. Having them doesn’t hurt anything, but they’re really just in the way.

2. Go to Colors/Components/Decompose. Change the color model to RGBA, and make sure Decompose to Layers is checked. Hit ok.

3. You now have a greyscale image with several layers. Ignore it for the moment and go back to the color image. Go to Colors/Desaturate. You can play with the radio buttons to get an image you like, but Lightness is good. Hit ok.

4. Ctrl+C to copy the desaturated image.

5. In the RGBA image, deselect the R, G, and B layers, and make sure the alpha channel is selected (it should be pure white). Ctrl+V to paste in the desaturated image, and right click and Anchor Layer in the layers dialogue to make it the new alpha channel.

6. At this point, you can use Colors/Brightness-Contrast to play with how dark the alpha channel is. What we have now is probably good.

7. Still in the RGBA image, click Colors/Components/Compose. Make sure the color model is RGBA. Click ok. You’ll get a very crazy looking image, which is ok.

8. You should now have three images open – your base texture (terrainsand03.dds in my case), terrainsand03-RGBA.dds, and rgba-compose.dds. Make sure you’re in rgba-compose.dds (it’s the messed up looking image you just made). Go to Filters/Map/Normalmap. You want your dialogue to look something like this:

Of note, make sure that where it shows the Alpha map, you have it set to the alpha channel of the RGBA image, and not one of the other channels.

Hit ok. Your rgba-compose image will now be a functional normal map:

9. Select File/Save As, and save it in the same spot as your source image, attaching _n to the end of the name. For example, for terrainsand03.dds, I’m going to call the normalmap terrainsand03_n.dds. You’ll get a popup for DDS options. Make sure Compression is set to DXT3 or DXT5 (these formats save alpha channel information – DXT1 does not, and is what you should save your base textures as. Make sure Generate Mipmaps is checked, and hit ok.

The last step is to go test things out in game. You may need to come back and redo the process with a darker alpha channel or some other change, so if you have the computing power for it, it pays to keep all your GIMP windows open while you test.

I have furthermore just had the following exchange:

[21:01] serenadingwords: Were we talking about anything?
[21:03] Dwip: Yes.
[21:03] Dwip: The sociopolitical impacts of squirrel aliens on the countries of 17th century Asia.
[21:03] serenadingwords: The sociopolitical impact of squirrel aliens…
[21:03] serenadingwords: DAMMIT YOU TYPED FASTER

Which is a reprise of an exchange from 2005.

Which then brings to mind this exchange, which I think says everything that needs to be said, since anything more would be seriously unfunny:

[23:24] Dwip: Back when we were funny, &c.
[23:25] Dwip: I’ve saying that about roughly 3 years in the past for about 6 years now.
[23:25] serenadingwords: Yeah.
[23:27] serenadingwords: It’s sort of at the point where you were saying it in 2007, and now you’re saying it ABOUT 2007, so.
[23:28] Dwip: Yeah.

This one’s simple. I got really tired of playing dress up with my newly created NPCs, and finding proper hair colors was annoying, since you had to find that one guy with the right color and copy his values.

So, I made a series of reference images to help me, and maybe help you.

Hairstyles

Note that these are CS screenshots, and actual hair looks moderately better with shaders applied in-game. Names in ALL CAPS are the names of hairstyles from the in-game chargen slider, while lowercase names for hairstyles found in the CS but not in-game. The numbers underneath each head are the RGB values for the hairstyle.

Note that Platinum, Orange, and Adoring Fan are appropriately very rare.

Lower Class Clothes

Found in the CS under Clothing/LowerClass. Where clothing for only one sex is shown, it means the other is similar enough not to warrent a picture.

Middle Class Clothes

Found in the CS under Clothing/MiddleClass.

Upper Class Clothes

Found in the CS under Clothing/UpperClass.

Robes

Found directly under Clothing in the CS, mostly prefixed with “Robe” and a name.

Shivering Isles Clothes

Clothes from Shivering Isles. CS locations vary:

LC Clothes are in Clothing/LowerClass.
UC Clothes, plus Duchess are in Clothing/UpperClass.
Sheogorath’s Regalia, plus Apostle, Order and Zealot robes have their own categories under Clothing.
Haskill’s suit can only be found by searching for SEHaskillSuit in the main Clothing list.

I’ve spent a hell of a lot of time describing the good parts of the 2.x series elsewhere, so I won’t keep harping on the point, except to say that I’m really proud of what’s been accomplished here. AFK_Weye 1.21 was a pretty good mod by all accounts, but I’ve gotten a lot better at this modding thing in the last year or so, and I think it shows. Better modeling, texturing, some sound work, and most of all the quests.

Which is all to say that you should go check it out, if you have thus far not heard of it from me.

Me: *sits up in chair* I know knife subdivide. Whoa.

Hanaisse: Are you going to teach us this so called knife subdivide, oh master?

Samson: WTF is “knife subdivide”?

Well, since you all asked, let me tell you. In the interests of completeness, let me expand the lesson to include a few other ways to subdivide a mesh in Blender, because they’re useful and it might help to know them.

First off, most of the commands I’m talking about are accessed via the subdivide menu – W key, followed by choosing something off the menu.

Second, if you want to play along at home, start up Blender, delete the starting cube, and add a new plane (Add/Mesh/Plane), then zoom in on it so you can see it well. More or less like this:

Third, it’s worth noting that subdividing works differently depending on what you have selected. It’s pretty intuitive once you figure it out, but for now accept that subdividing a single edge is going to look different from subdividing a single face is going to look different from subdividing a group of faces.

Now, to start out, here’s some simple subdivision. Select the entire plane with the face select tool, and Subdivide with W key, Subdivide.

You should wind up with four square faces instead of the one square face that you had. Notice that Subdivide splits each edge at the center.

I said a moment ago that it looks different if you subdivide an edge instead of a face. Hit Ctrl+Z to undo what you just did, then select the top edge of the plane with the edge select tool. W key, Subdivide to subdivide the edge, and you get:

Instead of four squares, we get three triangles, divided by edges that terminate in the center of the edge we just subdivided. You can hit Ctrl+Z to undo the subdivision, the play around with different combinations of edges if you want.

Using that same edge, let’s talk about Subdivide Multi, which subdivides an edge or face multiple times. Select the top edge again, then W key, Subdivide Multi. It will ask you how many cuts you want to make. 2 is the default, and that’s fine. You’ll end up with this:

Instead of cutting in the center, Subdivide Multi creates cuts at precise intervals along the edge. If we had performed the command with the whole face selected, we would have ended up with nine small squares.

The last subdivide command I want to talk about is Subdivide Multi Fractal. This is like Subdivide Multi, except that it warps the newly created verticles randomly.

Ctrl+Z until you have the blank plane. Use the face select tool to select the whole face, then W key + Subdivide Multi Fractal. Like Subdivide Multi, you’ll be asked how many cuts you want to make, but you’ll also be confronted with a thing called “Rand fac” that’s set to 10 by default. That’s the random factor, and it’s how much warping you want. Lower numbers mean less, higher numbers mean more. For our purposes, the defaults are fine. You wind up with something like this:

Nine small squares, with the points warped in space. This has a lot of utility when you want to create rough walls and the like, but usually needs some sort of subsurf or multiple divisions to make it believable.

For example. Ctrl+Z to get back to the blank plane. Select the whole face, then Subdivide Multi, then Subdivide Multi Fractal:

You get a much smoother, more real-looking roughness. You have to be careful, however, as this is a very quick way to create excessively high-poly objects.

There are a bunch more things in that subdivide menu, but those are the particularly important ones for our purposes.

Lastly, let me talk about that whole knife subdivide thing I mentioned. To get my point across, I’ll be using this beam here:

You can replicate what I’m about to do by adding a cube mesh and selecting the whole thing.

The K key brings up the Loop/Cut menu. You want Knife (Exact). The cursor changes to a crazy little knife thing, and you get some instructions along the bottom of the window. They’re more or less gibberish in the way of Blender. The important thing is this:

- Click the left mouse button and let up, then move the mouse. A little purple line will follow the cursor. That’s your cut line. Click the left mouse button again to actually create the line, and start another one.

- If you click and hold the left mouse button and drag, you’ll draw the cut line freehand.

- Hit the Enter key to actually make a cut.

The way this plays out is like this:

That angular line to one end is where my cut line was. As you can see, the knife tool is a lot less constrained than the various subdivision tools.

In practice, the way this plays out is this:

- The various subdivision commands are primarily useful in mesh creation, where exact precision is less useful than creating a bunch of smaller faces.

- The knife tool really shines when modifying existing meshes, especially UV mapped meshes. For example, that beam at that length was useful for one part of my mesh – vertical supports. But I also needed crossbeams in between two vertical supports. Creating and UV mapping an entirely new shorter beam would have been pretty onerous, and simply moving edges would have distorted the UV map. With the knife tool, however, I can simply copy the long beam, move it into position where I want it, and the use Knife (Exact) to divide the part I don’t want so that I can delete it. The UV map is preserved, and I just saved a ton of time.

The basic problem is this – the yellow arrow for player location on the world map (and probably the local map) is only ever accurate when the player is facing precisely Northwest.

If the player isn’t facing northwest, the arrow becomes some level of inaccurate, like so:

The red cross is facing NW and is accurate. Going clockwise, the blue dots are N, NE, E, SE, S, SW, W. This holds true for both menus and menus80.

That was after my first playthrough. We’re now a second playthrough, the expansion, and all of the DLC later. I’m pretty late to the party here, but as you might imagine, I have a few more things to say about it now. Let us begin.

Writing and Plot:

I’m going to be pretty enthusiastic about a lot of this game, but let me start by gushing over how great the writing is for this game.

On the face of it, this is pretty generic worldbuilding – I mean, name me another fantasy RPG where you fight demonic hordes to save civilization, said civilization including the remnants of a collapsed dwarven empire and some forest-dwelling elves who got bushwhacked by humans. It won’t take you all that long, so I’ll wait here for you.

Oh, you came up with one already? Well ok then.

Any seeming deficiencies in the worldbuilding, and I actually don’t mind the worldbuilding because it works, are more than made up for by the plot. There’s so much going on here I don’t really know how to begin.

Obviously there are the titular origin stories, five of them. I’ve been through the Magi one (my first) and the Dwarf noble one (my second), and both were of excellent quality, and really gave some great backstory to my character and the world. People generally single out the Dwarf noble story as being the best of the bunch, involving as it does political intrigue, fighting, and danger, and it really rounds out the Orzammar storyline. For myself, I thought the Mage story was just as good, and I really got a lot out of the interplay between the Circle, the Templars, the Chantry, and the Fade. I can only hope the others are just as good.

Without going into a lot of detail on the rest of things, let me just say that the main plot is suitably epic, and branches out well in ways I don’t think anybody’s ever really accomplished before. Grand betrayals, redemption, political intrigue, and regret are all here. The choice you make at the end is actually hard, and I really had to think twice before making it – in both of my playthroughs. That’s hard to do in a game, and I salute Bioware for pulling it off as well as they have here.

I also want to put in a plug for the characters here. I like some of them more than others, but there aren’t any of them that actively annoy me. I might not like them as people, but as characters they’re uniformly well done. So far I’ve had Alistair (twice), Oghren (twice), Leliana, Shale, and Wynne in my party, and enjoyed my time with all of them. I’m not as unabashedly a fan of the romance plotlines as I was in BG2, but they’re well done enough here, and any lack is made up for by the interparty banter. Oghren and Shale are worth it just for their banters, as are the Allistair/Morrigan duo.

What I’m really trying to say here is that this game is worth it just for the story experience alone, and despite what I’m about to talk about, this is some of the best writing video gaming has ever seen, and you should experience this game just to see it.

Gameplay and Interface:

Where do I start here. I hate to be negative about anything in this game, but the gameplay is a mess.

Let’s start with the interface, because it’s less of a trainwreck than the rest. Since Neverwinter Nights, Bioware has ridden the isometric 3D train pretty hard. They’re finally getting pretty good at it. Unlike Samson, who wants the camera to die in a fire, I’ve spent enough time with Bioware games to recognize that this is a camera that you can mostly use. Unlike the bad old days, there’s no ridiculous camera spinning bullshit, and you only get caught behind scenery sometimes. Zoomed out, the only thing I wish is that you could zoom out a bit MORE to be able to see the entire battlefield. While I think it’s fairly clear that for party-centric games top down is the way to go, what we’ve got works far better than any of its predecessors.

Beyond the camera, the rest of the interface is more or less intuitive if you’ve ever played an RPG. Journal, quest markers, pretty good map, quickslots, it’s all here. Of note is the codex, which stores non-plot critical information about creatures, characters, the world, and other things. I don’t know that I think minor sidequests should be stored here, but it’s fine enough, and hunting down codex entries makes for a fun little sidequest during the game.

Now, as to the gameplay. I just wrote a big post discussing Valve’s level design, and playing Dragon Age: Origins in light of what was discussed there is pretty disheartening. At the root of things is the fact that the combat model in this game stinks, which in turn boils down to the fact that balance pretty much doesn’t exist. Not all classes and powers are created equal. With a mage, you can practically solo the game. If you’ve got the right powers, you definitely can. Two-handed warriors are basically inferior to twin weapon anything and mages, and essentially the entire slate of sword and shield warrior powers is useless, which was intensely frustrating to my dwarven noble. Meanwhile, enemy powers range from the I-didn’t-even-notice-what-that-did to the colossally overpowered Overwhelm and Grab powers, which boil down to you watching helplessly over the course of 30 seconds while your character is brutally killed while you can do absolutely nothing about it.

It’s fairly difficult for me to see how anyone believed this was a good idea, but the combined effect is that it makes combat boring as all hell unless you take some of the fun powers. My mage was a whole lot more fun to play than my mace and shield dwarf mostly because, as a mage, he actually had things to do in every fight rather than just stand there and swing.

If that wasn’t bad enough, you might remember me talking in that level design post about varying things up between combat and puzzles, or combat and anything else at all. In Dragon Age, combat is broken up mostly by more combat and occasionally by story elements. When the story elements are in play, things are a lot better, because at least you feel like you’re fighting for the destiny of your or somebody’s people. Unfortunately, long stretches of the game, including the vast majority of Orzammar and the Dalish plotlines, are loooooooooong bland dungeon crawls filled with bad guys and nothing else. It’s a horrid way to run a railroad. Fortunately, outside of the bland dungeons, combat is at the very least tolerable, and usually more like what it should be – challenging and fun.

Lastly about the dungeons, and going back to my point about balance, the game is sort of partially level scaled. Badly. To quote Shamus Young, “People usually prefer a difficulty CURVE to a difficulty SAWTOOTH.” This is pretty much it – fights vary between cakewalks and soul crushingly hard TPK fests, sometimes over two rooms in a dungeon, sometimes the same fight with the same character at the same level with a different player. There are ways to do partial level scaling correctly, but this isn’t it.

I could keep going here. After playing so much Oblivion and Fallout 3, it’s really hard to go back to Bioware’s mostly non-interactive and linear dungeons. The various branching narratives and side quests of Dragon Age’s middle game are highly compelling, but again, the dungeons are some of the worst I’ve seen, made worse by the fact that I know they can do better, as you can see by playing through basically any given dungeon in Baldur’s Gate II.

It’s worth saying that a lot of this can be overcome by playing on Easy. I don’t really like that, and I wish there was a way to have a better combat experience without making it cartoonishly simple, but short of wholescale redesigns of the level scaling system and entire dungeons, what you see is what you get.

Graphics and Sound:

To briefly cover sound, the best compliment I can give to it is that everything sounds like it ought to. Swords clash, people scream, spells sound like spells. The soundtrack, when used, is mostly used compellingly and well. I’m not really doing it justice here, because you really have to hear it. Too, the voice acting is generally quite good, although as Penny Arcade once said, it’s pretty difficult to get around the accentless dwarves and elves (well, the elves. Dwarves varied pretty oddly between Texas and…somewhere). It makes them a little bland, and is hard to understand considering that the Orlesians get a distinct voice, as do the Antivans.

That having been said, I feel I should mention that the game does not, or didn’t always play nice with Realtek sound drivers. When I originally played the game on Vista, I had to use some fairly annoying workarounds every time I wanted to play to avoid stuttering and generally atrocious audio quality. The same hardware and the same game version played on Windows 7 were flawless, and I’m left to wonder if this was solely a Vista issue or what, since many many other people reported the same problems.

Now, as to the graphics:

(from right to left: Badass, unbelievably badass, awkward, and hiding his face in utter shame and humility)

That picture should tell you most of what you need to know right there. In general, the graphical quality of the game is pretty high, although with a puzzling lack of any sort of normal mapping. That’s not really a problem with the armor, but every single suit of clothes and almost all of the robes look flat and painted on, and generally like they’ve just stepped out of 2003. It’s very difficult for me to understand what could have possibly compelled them to not deal with this issue, and the only reason I can think of is that work on the engine dates from at least 2004. The ultimate result is that this looks like a game with great graphics from 2006, as opposed to 2009. However, since even 2006 graphics are still fantastic, this is a relatively minor complaint.

The other unfortunate thing is that while heavy armor looks badass and awesome, leather armor at best looks lame, and anything for mages looks unfortunately ridiculous. While I understand a lot of this has some realistic basis (have you ever seen real medieval headgear? It’s pretty bad), a little bit of self respect would have gone a long ways.

Spell graphics are almost entirely excellent, however, with the unfortunate exception of a few things like the Tempest spell. It’s very hard to get magic that looks good in a game, and Bioware has done well for themselves here.

Likewise excellent is the animation. There are a couple of random glitches, but nothing that should give you worry. I should say that while actually playing the combat is an often frustrating experience, it’s a real joy watching the combat, and I can take comfort in the fact that even as my characters are being slaughtered by the hax Overwhelm power, it looks suitably terrifying and horrific (seriously. Watching spiders crawling over my guys creeps me the hell out). I wish I had animation this good for Oblivion.

I don’t think anybody really does cutscenes as well as Bioware, and that’s good, because there’s a lot of them. Cutscenes in games are often a real drain to sit through, but I think these are about as good as they can possibly get – short, to the point, well-animated, well-acted, and they drive the plot perfectly. It’s really hard to get this right, and kudos to Bioware for doing so.

In direct contradiction of Samson’s take, let me just say that I enjoyed the Fade, both as a visual experience AND as a gameplay experience. One of the reasons I’m very glad I played a mage first is because it made the Fade sequences hit a lot harder than they might have otherwise, and the Fade section of the Circle Tower is probably the best realized dungeon in the entire game.

Having griped about the normal mapping thing, I should say that despite that, the environments are well done. The cities are nicely urban, the nature feels natural, Orzammar was a bit blocky for my taste but still good, and in general everything just looks like it should. Unlike Samson, the lack of any sort of sky doesn’t overly bother me. It’s a Bioware game, and they haven’t bothered with day/night cycles since Baldur’s Gate II. I don’t really like that particularly, but it is what it is.

To sum it up, this game is an audio/visual treat, but not nearly to the same degree that it could have been.

Online Components:

Let me say this. If you’re going to have a game that has an online component, this is just about the best way to do things. While I’m not particularly in favor of the needing to create an online profile thing, it is at the very least a one-time annoyance, and it was easy enough to turn things off that I didn’t want. In practice, the whole thing is pretty unobtrusive.

As to the things it does, screenshots and achievements, both are kind of nice, although you’re either going to care or you’re not. In the case of screenshots, I don’t really mind having them automatically taken (a nice touch, if somewhat flawed), but I don’t bother to upload them (who cares?). As for achievements, this is either the sort of thing you care about, or the sort of thing you don’t. I find, somewhat to my surprise, that I actually do care about them. They’re fun little goals to set for yourself, and add a little value to replays. Again, no sharing.

The entire social system here isn’t really for me, but I think if you’re going to do it, this is about the best way that I’ve seen.

Dragon Age: Origins – Awakening:

And now a word about the expansion, Awakening. In short, this was a worthy addition and extension of the game. The high quality plot was continued, and I found the revelations it contained to be, if not exactly shocking, highly compelling and interesting additions to the lore of this world Bioware has constructed. Too, I enjoyed the characters, both new and old. Like the set in Origins, they’re well-written, interesting, and likable. There’s a new system of interaction with them that replaces the party camp mechanic as well, and if it still has a few quirks to work out for Dragon Age 2, it’s still a pretty good idea.

The other thing Awakening does is to mostly quiet my angst about Dragon Age’s gameplay. I’m unsure as to if this is merely a symptom of the fact that there are many fewer levels to deal with in Awakening, or if Bioware has figured out their system better. I suspect the latter, as there were many, many fewer areas of egregious combat annoyance here, with much better balance overall. Dungeons were a lot shorter and more interesting than they were in Origins, and the whole thing is much more fun and less of a slog.

I do want to raise one sour note, and that has to do with the crafting system present in Origins and expanded upon in Awakening. In general, I found the one in Origins to be a good start, especially when it came to runes, but wanted the ability to make my own weapon runes, and a system of armor runes, both of which were created for Awakening. The flaw is this. Outside of actually crafting, you can’t see your recipes. Unless you took crafting skills with your main character, however, you can’t access crafting in camp/in your keep, which is where all the vendors for the various crafting components are. This leads to a horrendously annoying experience of going outside, reading recipes to discover what you need, going back inside, buying it, going back outside to craft some runes, going back inside to enchant, then going back outside to go play. At best, this is five or six loading screens worth of pain, and if you forget something, God help you. There’s absolutely no reason for this sort of bullshit, and especially in Awakening where all the other merchants are outside, there’s literally no point in this except to grief the player.

On the whole, 8/10 aliens.

Mods:

Like any good game in this day and age, Dragon Age can be modded. I haven’t really spent any time with the toolset, so I can’t very well speak to it, but as compared to anything Bethesda has ever made, mod installation is a disaster. There’s a tool called DAUpdater that comes with the game, and if you’re lucky and the planets and stars align, it will work, and if you’re not, the results will be painful. There’s a third party tool called DA Modder that seems to work better for the most part, but again, it’s entirely possible to mangle installation seemingly at random, and once you do it’s a pain in the ass to figure out what went wrong. Compared to the esp/esm format and the tools available for Bethesda games, mods for Dragon Age are like something out of the Dark Ages.

DLC Overview:

Also like any good game nowdays, Dragon Age comes with a full slate of downloadable content. At time of writing seven DLCs have been released, and we’re due for at least a year and a bit more. This is almost unimaginable largesse from the point of view of somebody who likes the game, and I wish more developers would do this.

Too, as a system, I find the way that the DLCs were implemented to be almost seamless and foolproof, with one glaring exception – the fact that characters and locations were inserted into my game to compel me to buy the DLCs is immersion breaking and generally a dick move of the most gigantic proportions, and I want to make it very clear that I own all the DLCs very very much despite this move rather than because of it. I also want to make it very clear that the person who came up with this idea deserves to be shunned in polite society for all of time.

That having been said, as a way to turn cash into gameplay, this is about as good a system as they could have come up with outside of Steam. While you need to have an online profile to get the DLCs, and it will want you to be logged in to play, which is annoying but mostly understandable, this as far as it goes. Too, while Bioware went down the road of selling “Bioware Points” for cash, and then selling me DLCs for Bioware points, the point costs of the DLCs are such that you don’t waste points, which is a refreshing touch for anyone who watched the Fallout 3 debacle. With the exception of actually buying the Bioware Points, you can purchase, download, and install the DLCs from within the game itself, almost at a single click, which is a very nice touch. If only mods were this easy.

All griping aside, I want to say that overall I enjoyed the DLCs thus far a great deal, more than any other slate of DLCs I’ve been exposed to that I recall. To discuss them individually:

Blood Dragon Armor:

This was one of two freebie DLCs that were included in my copy of Origins, and it’s sort of the big brother of a whole host of other item freebies that were released prior to the game for various editions. This one is a full suit of platemail, most of which is available in the party camp for sale. It’s solid mid-game armor, and my warrior spent most of her time in it from when she met the requirements until the end of the Orzammar plotline. Nice statwise, I thought it was a little oddly styled, which I assume is to help it fit in to Mass Effect 2 (WTF?).

Handy, but not really critical. 6/10 aliens.

The Stone Prisoner:

This was the other DLC that came as a freebie in my copy of Origins, and unlike the annoying plugs for the other two launch DLCs, this is the one that actually got me to buy the others. Basic idea is that you can recruit a very small stone golem, Shale, for your party. Shale has some fairly unique items available for her, and is unique in that she can fulfill almost any role in the party – tank, damage, ranged, support buffs, she can do it all. In my second playthrough, she was a tremendously effective brawler, singlehandedly responsible for about half my party’s damage over time.

The other thing about Shale is that she’s hilarious. Even if she wasn’t awesome statwise, she’s worth having in the party for the hilarious banter. Her personal side quest also has the benefit of being the one part of the Deep Roads excursion that I actually cared about, and her one area was better than the other four combined. If I have one complaint, it’s that finding higher tier crystals was pretty difficult. Also that it ended. Endings make me sad.

Totally worth your time. 10/10 aliens.

Warden’s Keep:

This was the other launch DLC, and I went to the trouble of actually buying it. Samson was pretty critical of it, and it’s not hard to see why. The party storage chest was nice, but there’s a mod for that for your regular camp. The merchant is the best one in the game, so there’s that. But for a supposed Warden base, it’s a little ridiculous that you can’t treat the damn place like a second camp. If I ever make a mod, it will be to do just that very thing.

On the other hand, the equipment is super cool, for style if nothing else, and I found the story compelling. It really rounded out the Grey Wardens for me, and gave me an insight into their character that you don’t really get from the kindly father figure who is Duncan. If for no other reason than the story, I found this to be an enjoyable DLC, and I’d miss it if it weren’t part of my game.

Fun, but deeply flawed. 6/10 aliens.

Return to Ostagar:

In which you return to the site of your greatest defeat (well, second greatest for my dwarf, but) and avenge some wrongs. I really enjoyed this one. It was about the best reboot of a map you could hope for, had a great story to it with lots of emotion for my Warden and Alistair, and the equipment was cool. In two playthroughs now, equipping Alistair with Cailin’s gear has simply felt correct to me, as if by the mere act of doing so I am fulfilling some sort of prophecy. It is, at heart, a dungeon crawl, but an above average one in Dragon Age terms, propelling you forward with a sense of history and story.

Flawed as always, but who cares? 9/10 aliens.

Darkspawn Chronicles:

I expected to heartily dislike this one but bought it anyway on the strength of the others. I was fortunate to find myself happily surprised. In this DLC, you play as a hurlock vanguard during the last climactic battle, and your goal is to kill all the main characters and save the main evil badguy.

For a DLC that was entirely combat, this was actually a pretty fun diversion. The combat wasn’t ridiculous at all, the mechanics of controlling darkspawn were fun (yes, I gleefully overused the Overwhelm power on the Shriek), and got a nice thrill out of killing off various party members and NPCs from the main game, especially Alistair. Killing Bann Teaghan always makes me sad, though, for some reason. I like Bann Teaghan.

Overall, it was fun as a diversion, but it is what it is, probably the weakest DLC. 5/10 aliens.

Leliana’s Song:

I may be overenthusiastic having just played it, but I find it hard to find anything bad to say about this one. Leliana’s backstory was one of the better companion stories in Origins, and getting to live part of it is neat. For a DLC that’s mostly thieving hijinks, the style is about the closest thing to a heist movie you can get in Dragon Age, sort of Oceans Elven meets the Thief series. The focus on semi-stealth and trickery over combat is hugely refreshing, and my time spent here was hugely enjoyable, and my only complaint is that Leliana ditched the hairstyle she has here for the crappier one in Origins.

Worth every second. 10/10 aliens.

Conclusion:

I’ve spent a lot of time being negative about various aspects of Dragon Age, but it’s very much an I criticize because I love sort of thing. Of its various flaws, only the griping about combat approaches serious, and this can be fairly well mitigated through the difficulty slider when necessary. The world Bioware has created, and the stories they have chosen to tell in it, are hugely compelling things to me, easily the best in that regard I’ve experienced in any RPG since Baldur’s Gate II. This is RPG gaming at its finest, and you would be utterly remiss not to experience it.

9/10 aliens.

Neverwinter Nights

Icewind Dale 2

Because here at the Ravings of Demented Rabbits, we pride ourselves on bringing you up to date and timely game commentary.

But there is one thing I want to talk about, because at least a few of the people reading this are modders, and that’s Valve’s approach to level design, which I think has some good application for other games, including Oblivion.

Put roughly, the approach taken by Valve for the Half-Life series boils down to this:

- Have an area bounded in some way. Usually this is a building, or a complex of buildings. Usually interiors, but often exteriors or partial exteriors are included.

- Often there will be what Valve refers to as vistas that bookend the level. Think of them as scenic overlooks, showing the player where they’re going, showing them what they’ve accomplished, or sometimes both. A lot of the time they use these as a sort of reward for accomplishing something, because they look cool.

- Somewhere in the level will usually be what Valve calls an arena. Big, usually open space, bounded by exits, and usually the site of some kind of boss battle.

- Each area usually has some sort of overarching goal. Usually this is a puzzle of some sort, but sometime’s it’s just to make it through one end and out the other side. This is usually broken down into several sections which take part in various sub-levels of the big level, and is broken down into several sub-tasks to accomplish the big task. Once the puzzle is solved, there’s usually some kind of boss battle, then you move on.

Inside each sub-level, there will usually be stretches of combat, broken up by little mini-puzzles – how do I get up to that pipe, or how do I get past the door, or what have you. Valve talks a lot about how they do this to combat player fatigue – too much combat gets to be unfun, while too much puzzle solving is likewise unfun. If you mix the two, the whole is greater than the sum of the parts.

- There is, of course, also the constant thrill of exploration and finding new stuff.

The Valve Example:

To demonstrate, let me use some handy Youtube footage of the Episode Two chapter Freeman Pontifex, about half-way through the game. This whole video series is about 30 minutes, so feel free to skip to the time stamps I note, and skim the rest.

The video starts with a vista, showing a bunch of enemy troops marching on your side’s base. It’s meant to serve as a reminder of what your ultimate goal is in the story, having just spent the previous chapter rescuing Alyx Vance, your companion.

Immediately thereafter it moves into a fairly large arena area where the player fights some random antlions, and beginning at about 2:35 turns into a boss fight with two large antlions. You’ll notice the large structure approached at about 2:30, which is a sort of safe zone the antlions can’t kill you in, but from which you can’t kill the big ones – you need to go grab the explosive barrels for that.

As a boss fight in an arena, I find a couple of things intriguing here. The first is that multiple enemy fights are almost always more interesting than single enemy boss fights, although you need some of those, too, from time to time. The second is the interesting terrain. Wide open spaces are pretty boring, but being able to dodge, snipe, and hide when necessary makes things more fun.

This whole fight is sort of the concluding act of the last chapter, where you got chased around a lot by the big green antlion guard, but couldn’t fight back.

Starting around 5:25 is a short little puzzle. It’s not one of Valve’s best ever, but it serves as a nice little transition between areas, and gives you something more exciting to do than just walk somewhere.

After a short bit of random hallway and some neat atmospheric storytelling at about 6:53, you hit another vista at 7:20. This shows you the area you’re about to traverse, which looks cool and gets you excited about the next part. Note Alyx pointing out your immediate goal. One of the neat things about Half-Life 2 is that your companions don’t suck, but that’s for another post.

Starting at about 9:00, note the combination of very brief little environmental puzzles with brief amounts of combat – usually an enemy or so at a time. Also note the use of vertical spaces. It’s not as in evidence here as it is in some other multilevel areas, but it is there.

In the second video, the first three minutes or so is more of this environmental puzzle interweaved with combat thing in very short bursts, followed by a section at 3:06 or so. It’s combat, but rather uniquely your companion gets to help out. I’m not sure I know a way you could accomplish the same thing in, say, Oblivion. Similarly, it would be reasonably difficult to add in the use of the environment by yourself and enemies, exemplified by 4:14.

At 5:00, we transition from indoors to outdoors. Hard to do in Oblivion (although not impossible), but the point isn’t indoors/outdoors so much as terrain variance. You CAN change tilesets.

The bit from 5:18 on is another arena, now with companion assistance and a bit of fun with the environment. Having been through it, I’ll say that this is a pretty unique fight scene. The guy in the video is a lot more proactive about it than I am – I lured them all out so Alyx could shoot them. I think the point here is that the choice of using differing strategies can offer a lot.

From about 7:50 on, we change back to puzzle/exploration mode. There are a couple of enemies, but nothing serious – the goal is to get out of the room. Note that the solution of pulling the lever never even occured to me – I always treated it as an environment puzzle, and stacked boxes on top of the forklift to jump from. The claustrophobic vent tubing is, of course, a nice change of scenery, leading to another nice change of scenery to outdoors at 9:10.

The third video begins almost immediately with what would be a retread of the last arena with Alyx and zombies, but also features some toxic sludge as an environmental hazard. It doesn’t work as well as it might, but it does work well enough.

At 1:38 they start combining things – large environmental puzzle, plus combat. Alyx can only sort of help you out of this.

At 3:35 or so, you’re up the other side and to another vista if you care to look, showing you all that you just crossed to get to this point. 4:10 continues this, with a stunning reminder of what’s going on in the main plot that also sets up the grand finale puzzle. The fact that everything in this section looks awesome is a further reward for everything you just went through.

Everything from 5:35 on is exposition, setting up for the next chapter and acting as a reward for having braved all that danger.

Applying It To Oblivion:

Now, let’s talk for a bit about how one might apply this sort of level design to a game like Oblivion. I find, somewhat to my surprise, that it actually maps quite well, although there aren’t any particularly good examples of it in the vanilla dungeons that I can think of. So for an example, I’m going to reach into AFK_Weye a bit, and talk about the entire dungeon of Kerrach.

It will be helpful to keep the CS open for this bit, as I’m not going to screenshot any of it.

Start in the Breakneck Lair (AFKKerrach01). There are a couple of major features of this part of the dungeon that I want to highlight, which I think are in keeping with the superior way Valve does things:

- There are, more or less, two types of architecture present here. The normal cave parts, plus the addition of some fort ruin architecture. This provides a contrast to almost all other caves in the game, which use nothing but the cave architecture, which can become monotonous when used over and over again.

- Use of vertical space. This is hard to do in Oblivion, because unlike Half-Life 2, you can’t, say, have soldiers rappeling from rooftops, and the AI doesn’t always handle this well, but by adding the central tower area, I did attempt to provide a little vertical space. This could have possibly been handled better, but was hampered by the other drawback that the modeling in Oblivion really isn’t set up for it very well. I try to take advantage of this where I can, however.

- In Oblivion, it’s much harder to make use of environmental objects as weapons or impediments as in Half-Life 2. What we do have, however, is traps, which can either be used by or against the player. In the Breakneck Lair, there are two log traps that provide some amount of flavor. These probably could have been handled differently, and there are more varied ways of doing traps, but they do serve to spice things up a bit.

- In the cave tileset especially, it’s important to spice things up so that you don’t end up with bland walls everywhere. In the Breakneck Lair, the way I did this was obviously through the tower, but also through lighting, such as the ceiling shafts, and the fires which indicated goblin activity.

- I mentioned two types of architecture here, but there’s actually a third. Behind the locked door at the very bottom, in the basement of the tower is a bit of Ayleid architecture, with the dremora who provides a boss for the dungeon. This is a tease, meant to entice players and draw them deeper into the dungeon.

- This part of the dungeon doesn’t feature vistas or arenas per se (although the dremora room might count if you squint hard), but it does have an overarching goal – to rescue Elahai, and find Thalonias’ shipping manifest. Both of these break up the combat a little bit, and give us a reason for being there.

- The other thing lacking here is puzzles of the sort found in Half-Life 2. This is something of a failing throughout Kerrach, although I sneak a couple in.

Now to Kerrach Molag Garlas (AFKKerrach02):

- The area where you first enter serves a number of purposes. First, it’s a vista of sorts, combining Ayleid architecture with caves and lava to give the player a glimpse of what lies ahead, and to entice them further in. Second, the fight here with the dremora archers over the mined bridge is meant to provide a different sort of combat experience. It’s fairly rare to fight more than two enemies in Oblivion in the first place, and usually not with many if any environmental concerns. Perhaps misfortunately, this section is somewhat hampered by the lighting and enemy AI (they rarely shoot at you across the chasm).

- Most of the rest of this fairly short level is straight up exploration and combat, ending in the large room at the top of the stairs, which serves as a not quite Valvesque arena with multiple enemies, which is again fairly rare in Oblivion. The thing missing here for the full experience was some sort of environment, but the spectacle of dremora in an Ayleid ruin is hopefully enough.

- As a whole, Kerrach Molag Garlas is basically a connecting area, with no real goal except to get through it. There’s a minor goal when confronted with the locked door to move on through to Kerrach proper, which can be solved by getting the key from the boss, but this can be bypassed if desired.

Next up is Kerrach (AFKKerrach03):

- This is the main part of the dungeon, it starts with a sort of vista, of the Ayleid king fighting dremora (which the player can simply watch if they choose), which turns out to be a scene where the player recieves two quests and a number of goals – rescue several ghosts, plus retrieve the two artifacts needed to go on and defeat the boss. Oblivion doesn’t make much use of jumping obstacles, so I made sure to use one here to give a little flavor.

- The balcony overlooking Kerrach is an almost perfect vista in the Valve sense. It gives the player the opportunity to survey the areas they are about to enter, itself a large arena, and offers the prospect of an entire Ayleid city drowned in a giant lake of lava inside a cave, a spectacle rivaling many outdoor scenes. This is all means of exciting the player (and a bit of a reward for the last two dungeons).

- The islands at the bottom of the lake are a large arena, filled with enemies who aren’t particularly threatening by themselves, but together offer a different sort of combat experience – many weak enemies. This includes some dremora up in the towers with bows – a little bit of vertical space usage for flavor. As an arena, this one features walls to dodge or hide behind, and it’s sometimes possible to knock the dremora into the lava, which is also a danger to you – use of the environment.

Going into the leftmost tower takes you to Kerrach Ageasel (AFKKerrach04a):

- Most of this level may be thought of as an exploration puzzle punctuated by a few traps and some combat.

- The very first room offers up the puzzle: clearly to proceed you’ll need to choose a portal, and two of the portals warp you back to the same room, suggesting that this is a maze. In addition to the goal you already have (retrieve the item that lets you proceed), the Ayleid ghost in this room offers you another goal for the level, another thing to solve inside the main goal.

- The second room, with the lightning stone, is a sort of environmental hazard arena, where players are forced to run through an open area to get past the lightning-shooting stone. Besides acting as a possible change of pace, the stone serves as an inducement to players to pay attention, since warping back several times to this chamber will prove costly to them. The other chambers in this section of the area are mostly straight up combat, which serve to punctuate the other hazards, and include a mini-arena that recalls the bigger one at the end of Kerrach Molag Garlas. Over everything, of course, is the meta puzzle of finding a way through the maze.

- The third room offers a different vista (lava) from the first two, keeping with the idea of varying the interiors as much as possible to provide a unique experience. The summoning circle with the daedra in the middle is a trap, meant to keep players a little alert and spice things up a bit after (or during) the dremora fight. Again, it’s possible to use the environment here, bottlenecking enemies or knocking them in.

- The fourth room is yet another differing vista (trees), and aside from the hostile deer, the sights and sounds are meant to provide a respite from the traps and combat of previous rooms, although the puzzle is still intact, and indeed one of the goals for the level can be found here if searched for – a bit of a mini-puzzle.

- Apart from the dremora, the fifth room is meant as a puzzle within a puzzle, a capstone to the area. This is one of the very few logic puzzles in AFK_Weye, let alone Oblivion, and provides another unique change of pace. In addition, the almost pure Oblivion cave provides a much different feel than the Ayleid ruins previously.

The Storm Spire (AFKKerrach04b) serves a number of roles in one small area:

- As a vista. The spire is fairly unique, being as it is in a void, and its un-ruined appearance provides a contrast to everything the player has just gone through, as well as a glimpse of powerful magic beyond their knowledge.

- Use of vertical space. The entire area involves going up while surviving the environment.

- As a use of environment for and against the player. It is possible to be blasted by lightning while climbing the spire stairs, and falling from them results in death. This hopefully heightens tension, while providing a possible asset or hinderance in the battle to follow.

- As an arena. The spire is the boss battle for this entire subsection of the dungeon, and as you enter the topmost portion, you are immediately confronted with a hostile dremora who shoots you with lightning from a distance and summons an atronach. This provides a very tight and hectic battle, where the environment plays a large factor in what occurs.

The rescue of the ghost and restoration of the well with the Waters of Anu serve as the emotional climax and reward for this section.

Going in the rightmost tower takes you Kerrach Buroseli (AFKKerrach05a):

- Thematically, this is much less varied than the Ageasel, and closer to what you saw topside – Ayleid ruins and lava, but now punctuated with some Oblivion cave architecture as well.

- This area is mostly combat, with some puzzles and traps for variety.

- The first couple of rooms serve as the thematic introduction to the area, before breaking things up with the room with the hole in the middle, which gives the first serious fight, as well as a reward in raising the cave and rescuing the ghost.

- A brief fight or so later sees the player at a potentially hazardous environmental puzzle – platforms to jump on, which may or may not be safe. Continuing left from the other side, the player is exposed to another trap – falling rocks. After that, a hallway full of skeletons to break up the traps with a more conventional combat experience. After that, the player hits one of two traps, which breaks up the skeleton combat with the dremora combat at the end of the area.

Like the Storm Spire, the Rending Spire (AFKKerrach05b) is designed to fulfill a number of roles:

- Although not as vista-worthy as the Storm Spire, the Rending Spire breaks things up a bit by offering slighly more pristine Ayleid ruins. As use of vertical space, the elevator is fairly unique and offers something new and different to the player. It leads up to an arena, where the player is shortly trapped in a very small area with a powerful and homicidal dremora. As an added environmental danger, it’s possible to fall through the hole in the center, which provides an interesting challenge.

- The reward for all of this is later, when the player returns the pommelstone to the statue, which gives a nice display of fire, and adds to the Kerrach vista.

The Spire of Shattered Hopes (AFKKerrach06a) and its sub-areas all provide yet more varied themes for the player – Ayleid ruins and Oblivion towers, a sort of nightmarish daedric garden with clannfears as pets (straight combat), a bit of quest dialogue (the ghost) to break things up, some more straight combat, and then some more quest dialogue and a section of pitch black traps to round things off before the player can exit back out to Kerrach.

The other side of Kerrach offers the player a reward in that they can now look back out over Kerrach and see what they’ve just accomplished, as well as witness the fire of from the statue, which is also a little bit of a mini-puzzle while players learn they need to fix the statue before they can proceed.

Like the tops of all Oblivion towers, the Throne of Destruction is a large arena, populated with several dremora in a multi-level battle. It differs slightly from normal Oblivion towers in that there are a few more dremora and the main one is harder, but otherwise it’s more less a fairly straight up battle with a couple of neat visual differences to the arena.

The cutscene at the end of Kerrach is a showy reward for a job well done. It reuses existing areas, providing some retroactive importance to them, and by having the player go back up and down the stairs on the other side of Kerrach, it lets them see a new version, allowing them to see what they’ve accomplished.

Some Concluding Thoughts:

As I said at the beginning, there are a number of admirable elements to Valve level design that I think work well in Oblivion.

- Vistas are somewhat difficult to pull off in the Valve style, since so much of Oblivion is confined to interiors, and much of the exterior space already offers excellent and scenic vistas. Through very large interior cells and small exterior worldspaces this is still possible, however, and can also be achieved by creating a new look for an existing tile set or area.

- Arenas are pretty common to all games of the type, of course. Where Oblivion differs from Half-Life 2 is that generally Half-Life 2 arenas are much more varied, with more interesting combats, and often offer scripted sequences where enemies will break down doors or use the environment in some fashion.

- Dungeons in Oblivion should have goals, generally quests. At this late date, there’s very little less exciting than a pointless dungeon crawl.

- Rewards should involve more than mere loot, and should include things like character interaction and approval as well as interesting cutscenes and the use of vistas.

- Environments should vary where appropriate. Combining tilesets is an easy way to accomplish this, and provides a lot of bang for the buck. Breaking up the monotony of a large dungeon is important. Use of vertical space where appropriate and applicable is nice.

- Perhaps most importantly is varying the tempo of the dungeon. Mix up the types of enemies, mix up the types of tasks – add some puzzles, add some environment challenges, add some traps. Combine when necessary. Although Oblivion lacks the robust Source physics engine, it’s still possible to create interesting puzzles through hidden areas, switches, and logic. It’s also possible to create scripted traps that behave differently than vanilla Oblivion traps, or to use the old traps in new ways.

But first, a digression on how to delete records from esps.

The easy way is this. Load up the CS, and select the file you want to edit, set it to the active file. Instead of hitting OK and having it load the esp, instead, make sure the esp you want to edit is selected, and hit the button that says Details. You get a list that looks like this:

Find the record you want to delete (it helps to know this beforehand), select it, and hit the Delete key. You’ll be confronted with a popup. Say yes. A “D” should appear next to it on the list. Keep deleting things until you’re happy, then hit the Close button, and hit the OK button. Your esp will load. You should then be able to immediately save it and close the CS.

Leveled Lists, Example 1:

Now let’s talk about how leveled lists relate to each other.

You may recall our friend from last time, LL0NPCWeaponMace100:

You’ll recall that this particular list picks any given item off the list that’s in a range between (player level – 8) and (player level), and has multiple items at each level. The reason it does this is so that it can gradually phase in and phase out weapons, as I demonstrated in our last episode.

Now, go find it in the LeveledItem category under Items in the CS. Right click it, and select Use Info. As you scroll through the list, you’ll notice that it’s used in a number of other leveled lists, almost all of which are LL1 lists or quest related lists. You’ll also notice that it’s called by a number of NPCs directly. That means that those NPCs will always carry a non-magical mace from this list.

Now let’s go look at one of the other lists, LL1NPCWeaponBlunt100:

This one is set up differently. It can pick from any one of four other leveled lists, one of which is our old friend LL0NPCWeaponMace100. The other three should be fairly self-explanatory by this point.

What happens at the LL1 level? Well, let’s go look at the Use Info for this list. You’ll notice that this time, the list is used by a mere three other LL2 level lists, and by far more NPCs. This is fairly common, to give NPCs a degree of variety in their weaponry – BanditMeleeMale4A, for example, could be carrying any sort of random nonmagical blunt weapon – a battleaxe, a waraxe, a mace, or a warhammer.

Now, what happens if we kick it up a level? Here is LL2NPCWeaponBossCombBlunt100:

This list is set up like the first list we looked at. Based on level, it picks an LL1 level blunt weapon list. Starting at level 3, it begins to phase in magic weapon lists, until at level 9 you have a 50/50 chance of getting either LL1NPCWeaponBlunt100 (our example above), or LL1NPCWeapon0MagicBlunt100 (it’s counterpart for magic weapons).

Now, if you look at the Use Info list for LL2NPCWeaponBossCombBlunt100, you’ll note that it’s used by a further LL2 list, LL2NPCWeaponBossCombAllMelee100, which if you’ve been paying attention to the naming conventions thus far, means that said list picks ANY melee weapon (have a look at it). It’s also used by a large number of NPCs. Most lists used by NPCs are LL1 or LL2 lists, because this gives the maximum variety.

Leveled Lists, Example 2:

Now, if you’ve been paying attention, you’ll notice that alongside our previous examples, there has been a whole other set of lists with the exact same names, except they also include “Lvl” in the name.

Here’s the difference. This is LL0NPCWeaponMaceLvl100:

If you remember back to the first tutorial, you’ll see that, unlike LL0NPCWeaponMace100, where you might get a range of different materials, for this list, at each level you have one and only option. At level 12, you will ALWAYS get a glass mace. This is, as you might have guessed, much less random.

If you follow the chain up, it looks much like the chain of lists from the first example, only with Lvl version lists instead of the other type.

Leveled Lists, Example 3:

Now let’s have a look at the way AFK_Tweaks does things:

These are the lists AFKBanditBootsLOW, AFKBanditBootsMID, and AFKBanditBootsHIGH. I didn’t use the naming convention, but these are the equivilents of LL0 lists, and they aren’t called by anything other than other leveled lists – namely themselves and a list called AFKLL0BanditArmorLightBoots100:

This list is actually misnamed – it should be an LL1 list. Note that I’ve prefixed it with “AFK” to differentiate it from Bethesda’s lists. If you check the Use Info, you’ll see that this is the list that actual NPCs carry around.

Here’s how the whole thing works.

In AFKLL0BanditArmorLightBoots100, one of the AFKBanditBoots* lists is picked, based on level. Level 8 characters would pick from the MID list, for example.

Then, in AFKBanditBootsMID, you have a chance of picking any one of several types of armor, in roughly equal chance. There are also calls to the LOW and HIGH lists, which means it’s possible for a level 7 bandit to wind up wearing glass (from the high list) or fur (from the low list).

You’ll notice that each list has exactly 10 entries, and several entries are repeated, which gives certain armors better chances of appearing. It works like so:

10% chance – call AFKBanditBootsLOW
10% chance – LeatherBoots
30% chance – AFKLamellarBoots
40% chance – ChainmailBoots
10% chance – AFKBanditBootsHIGH

If you’ve been paying attention in previous examples, you’ll recognize that this particular list structure ensures extreme randomness – spread over several pieces of armor, any given NPC is likely to have widely varying pieces of armor. The reason it was constructed this way is to maintain randomeness at high levels – if you look at, for example, you’ll see that using the equivilent Bethesda list, LL0NPCArmorLightBoots100, a character of level 20 or higher will ALWAYS get one of three armor types – Glass, Elven, or Mithril.

This is what leads to bandits wearing full suits of glass day in and day out. As one of the things AFK_Tweaks set out to fix was this percieved randomness problem, lists were constructed that featured this extreme randomness – now you’ll see scale, chain, lamellar and leather again as well.

A Note On List Design:

Of course, the way I chose isn’t the only possible way. You could, for example, redesign LL0NPCArmorLightBoots100:

1 1 FurBoots
1 1 FurBoots
1 1 FurBoots
3 1 LeatherBoots
4 1 LeatherBoots
5 1 LeatherBoots
6 1 ChainmailBoots
7 1 ChainmailBoots
8 1 ChainmailBoots
8 1 AFKLamellarBoots
9 1 AFKLamellarBoots
10 1 AFKLamellarBoots
10 1 MithrilBoots
12 1 MithrilBoots
12 1 AFKLamellarBoots
13 1 AFKLamellarBoots
13 1 MithrilBoots
15 1 ElvenBoots
18 1 ElvenBoots
20 1 GlassBoots

Both check boxes remain checked.

This version of the list weaves in AFKLamellarBoots, and also makes chainmail become a little more prominent. At high levels, you’ll notice a chance for both Mithril and Lamellar roughly evenly, with a slightly increased chance for Elven as you level, topping out at Glass with level 20.

At level 20, then, you have: 2 in 7 chance of Mithril, 2 in 7 chance of Lamellar, 2 in 7 chance of Elven, and 1 in 7 chance of Glass. Overall, Glass becomes about half as common, Mithril and Elven stay the same, and Lamellar is added in. Chainmail, Leather, and Fur disappear entirely.

On the other hand, if you wanted to maintain the previous AFK_Tweaks structure, you could change around the armors present in each list, which design exercise is left as a project for the reader.

There are a couple of things to keep in mind when designing your own leveled lists. In the stock Bethesda lists, certain materials become available at certain predetermined levels.

For armor:

Level /  Light  / Heavy
  1       Fur     Iron
  3     Leather   Steel
  6      Chain    Dwarven
 10     Mithril   Orcish
 15      Elven    Ebony
 20      Glass    Daedric

And for weapons:

Level / Material
  1     Iron
  2     Steel
  4     Silver
  6     Dwarven
  9     Elven
 12     Glass
 16     Ebony
 20     Daedric

You should keep these in mind when designing your own lists. In general, it pays to spend some time browsing the vanilla leveled lists and UESP when designing your own lists to make sure you’re staying within the guidelines for material quality and magic item quality versus level.

I now open the floor for questions and comments.

I’m going to be covering information talked about in the CS Wiki Leveled Item article, with some digressions into AFK_Tweaks and a few other things.

Let us start with the definition. Leveled lists are how Oblivion containers and NPCs/monsters generate random items. Any type of items – armor, weapons, scrolls, gold, you name it. It can do this in a few different ways, and we’ll be covering them shortly.

First, let’s go over the structure of a leveled list:

From the top:

ID – This is the name of the leveled list, and can be anything you want it to be. Most but not all Bethesda leveled lists follow a particular naming convention.

At the bottom are lists named LL0xxx# (say, LL0ClutterBones75), where xxx is some kind of descriptive name, and # is the chance of the list generating an item (more on this in a sec). In the case of LL0ClutterBones75, the list holds bone items, and has a 75% chance of generating one. LL0 lists NEVER contain other leveled lists, and ALWAYS contain items and only items.

LL1 lists are named the same way as LL0 lists, except that they contain a mixture of actual items and LL0 leveled lists.

LL2 lists contain LL1 or LL0 leveled lists. They never contain actual items.

There are some lists with “Lvl” inserted in their names, such as LL1NPCStaff4CombatOnlyLvl100. I’ll go a bit more into detail on those later. For now suffice it to say that these lists work slightly differently than other lists.

Calculate from all levels <= PC’s level – Changes list behavior. This is easier shown than explained, and I’ll do so in a bit.

Calculate for each item in count – If you have a leveled list that calls other leveled lists, this needs to be checked to maintain random results.

Chance None – This is the percentage chance that the list will not generate an item. In the example of LL0ClutterBones75 above, the Chance None field would be 25 (100-75=25). Be aware that this seems to have a disproportionate effect – Chance None 50 actually had a 75% or 80% chance of generating no items when tested on several subjects.

Now, the list of objects below Chance None:

Level – This is the (usually player) level at which a given item will be generated. In the list in the picture, this means you won’t see IronBoots until level 4, SteelBoots until level 8, and OrcishBoots until level 20.

Where things get tricky is that while containers are always based on the player’s level, NPC/monster items are not. The way that works is hazy, but appears to be that the item the NPC gets is based on either the NPC’s level or the player’s level, whichever is lower.

Count – Number of items generated. Usually 1, but sometimes more, especially in the case of gold.

Form Editor ID – The Editor ID of the item or leveled list.

If you right click on an item in this list, you’ll have two options – delete and new. Selecting delete will delete the item from the list, while selecting new will create a duplicate of that item in the list which can then be edited in its own right. To create an entirely new blank item, right click in empty white space and then select new.

To the right of the list is a dropdown and two boxes labeled Object, Level, and Count. If you’ve clicked on an item in the big list to the left, this is where you can edit it. This should be fairly self explanatory. Be aware that the Object dropdown is often quite slow. The best way of using it is to type the first few characters of the item name you want, then wait for it to catch up.

Alternately, you can drag items and leveled lists straight from the Object Window into the list. This is sometimes faster.

Preview Calculated Result – This is a preview function. It generates items from the list at Preview Level, Preview Count times. For example, in the list AdventurerWarriorBoots, setting the Preview Level to 10 would generate SteelBoots, and setting Preview Count to 10 would do so 10 times. Other lists are far more exciting than this.

Now, let’s move on to dissecting the different forms leveled lists can take, based on the two “Calculate” check boxes in the upper right. There are several forms this can take:

1. Neither box checked – Always returns the object closest to but below the player’s level (I say player’s, but it can be the NPC’s level – I talk about this under Level above).

2. Calculate from all levels <= PC’s level checked – As above, but can return random objects up to the value of the global variable iLevItemLevelDifferenceMax (default is 8) levels below the player’s level.

3. Calculate for each item in count – Makes sure, that for complex lists containing other leveled lists, that the same object cannot be returned twice. For other lists with multiple items at each level, has no effect in-game, but fixes a bug with a Preview Calculated Result.

4. Both boxes are checked – This combines both behaviors, and is used for complex lists.

This will all make more sense in a moment. Allow me to demonstrate:

Neither Box Checked:

This is the most basic sort of leveled list, demonstrated here by our friend AdventurerWarriorCuirass. It’s operation is very simple. Given the player’s (or NPC’s) level, it generates an item as close to but below that level as it can.

For example:

From levels 1-3, it will always return LeatherCuirass.
From levels 4-7, it will always return IronCuirass.
From levels 8-11, it will always return SteelCuirass.
From levels 12-15, it will always return ChainmailCuirass.
From levels 16-19, it will always return DwarvenCuirass.
From levels 20 and above, it will always return OrcishCuirass.

Simple, like I said.

Calculate from all levels <= PC’s level checked:

Now we’re getting more complicated. This list will return any item that falls within the range of (player level – 8) to (player level). For example:

At level 1: Always returns ArenaIronBow
At level 5: Could return ArenaIronBow, ArenaSteelBow, or ArenaSilverBow.
At level 10: Could return ArenaSteelBow, ArenaSilverBow, ArenaDwarvenBow, or ArenaElvenBow.
At level 20: Could return ArenaGlassBow, ArenaEbonyBow, or ArenaDaedricBow.

This is all capped at the highest level in the list. For example, the range for a level 25 player would normally 17 to 25 (25-8), and would always return ArenaDaedricBow. However, because the level of the highest level item in the list is 20, the actual range is 12 to 20 (20-8).

This is the basis for generating random leveled items that level with the player.

There are a couple of variations on this:

Variation 1:

Now, note our friend CGMythicDawnLootList here.

The idea behind this list is that it can return any item in the list. Level is irrelevant, and is always 1.

There’s also a variation on this concept:

Variation 2:

The way this list works is this. Given the player’s level, it generates an item as close to but below that level as it can. That’s like the neither box checked behavior. However, notice that at each level there are multiple items, and that Calculate for each item in count is checked. This means that at any given level, there’s a chance of generating any of the items at that level. For example:

At level 4, it could generate EnchIronDaggerEmber, EnchIronDaggerShiver, or EnchIronDaggerSpark (the entire level 4 set)
At level 6, it could generate EnchSteelDaggerEmber, EnchSteelDaggerShiver, EnchSteelDaggerSpark, or EnchSteelDaggerDraIntelligence (the entire level 5 set)

Both Boxes Checked:

This is the most complex type of leveled list. Looks bad, but it’s actually quite simple. To generate an item, the list can pick any item from (player level – 8) to (player level), AND can pick any item at that level. This is opposed to the plain Calculate from all levels <= PC’s level checked type list, because you can add multiple items at each level.

It works like this.

At level 1: 4 chances to pick WeapIronMace
At level 2: 1 chance to pick WeapSteelMace, 4 chances to pick WeapIronMace
At level 3: 2 chances to pick WeapSteelMace, 4 chances to pick WeapIronMace
At level 9: 4 chances to pick WeapIronMace, 4 chances to pick WeapSteelMace, 3 chances to pick WeapSilverMace, 3 chances to pick WeapDwarvenMace, and 1 chance to pick WeapElvenMace.
At level 20: 2 chances to pick WeapGlassMace, 1 chance to pick WeapEbonyMace, and 1 chance to pick WeapDaedricMace

And that’s how to read leveled lists. In Part 2, we’ll examine the AFK_Tweaks series of leveled lists, see how they fit together, and talk about how to edit them.

EDIT #1: Fixed erroneous thinking regarding Calculate for each item in count.

EDIT #2: Clarified behavior of Calculate from all levels <= PC’s level in cases where player level was greater than the highest level item in the list.