Subtraction engines...

[Stardog]

I have a simple answer for that one. If we're talking about purely subtractive, consider building a ladder using each method.

But I'm guessing purely subtractive editors are rare. UnrealEd has both add and subtract. The blue cube in my screenshot is an added brush like lego. You could duplicate it to add walls etc.

 

Subtractive mapping might foster creating a quick room / containing area, but that's where the advantage ends, and Additive pulls ahead fast at that point. Further, you can simulate the same with Additive by simply creating a brush and clicking 'Make Room.' There, you have one big sealed "air brush" to work within. In fact, I'm willing to bet you can do this faster in DR than DromEd or UnrealEd -- you drag a diagonal line, and click a button. Hugely easier than brush manipulation in the latter editors.

Nope, it's one click of Subtract in UnrealEd! It already has a brush placed.

 

For me there are no positives of additive engines, especially when you have trims along the bottom and top of every wall then want to add windows or resize things later.

 

CSG engines are intersection negative and positive brushes: which makes simple indoors areas ultra-fast. But it's expensive. And it's also error-prone when geometry gets too complicated. That's why games like Thief, Deadly Shadows, all liked using static meshes for detailed bits.

You're right there, but we should be using static meshes with this engine anyway. Is an easier import possible with DarkRadiant that would automatically write a basic material file for you and put your mesh/mat in the correct place? This is a timesaver in UnrealEd.

Edited October 27, 2009 by Stardog

[SneaksieDave]

I personally would never say subtractive is cleaner. Subtractive in my mind is a counter-intuitive hassle of using all sides of 3D shapes for the sake of containment (the room) and detail (the fountain, or whatever) at once. In Additive, you simply build what you want, WYSIWYG, instead of carving a bunch of holes which pollute the view (the subtractive brushes stay behind; they don't vanish after the cut, or the cut would vanish too). When construction workers go out and build a bridge across a river, they don't start with a giant block and cut a bridge out of it; they build the bridge up from pieces in empty space. Consider the same room (a simple box with a 'plus sign' in the middle) with each method. Which is more intuitive, more real-world? Simpler, and more flexible to modify?

 

(Edit: the outer line of the subtractive room shouldn't be there, assuming an infinite clay cube... but since we're allowing for geometry outside the Additive room, it's only fair to allow for geometry outside the Subtractive room, neverminding the mess that happens if the mapper now wants to surround the Subtractive room with more Subtractive rooms.)

 

I see it as containment vs. detail. If you desire Subtractive-style first-step containment in an Additive editor, simply drag the containment brush and click Make Room. Then you can get on with proper Additive building.

 

 

Edit:

But I'm guessing purely subtractive editors are rare.

I'd hope so. So if we're not talking purely Subtractive anymore, immediately, the case for Subtractive editing is watered down, because beyond the first click, the subtraction of the containment area, Additive takes over.

Nope, it's one click of Subtract in UnrealEd! It already has a brush placed.

Of the proper size...? Nope.

For me there are no positives of additive engines

Even though we've established that beyond Subtractive containment cuts, Additive takes over? That is, unless you rely on all positive detail to be meshes, which is very limiting cookie-cutter construction.

 

Do you want to edit the container easier, or the details easier? (Keeping in mind either is possible; it's just about which is easier). Do you want to build your map from a store of meshes, or unique brush detail? These are the pertinent questions, and establish my preference.

[Stardog]

Of the proper size...? Nope.

No, but the height of the dragged brush would also need changed also before you hit Make Room.

 

Even though we've established that beyond Subtractive containment cuts, Additive takes over? Unless you rely on all positive detail to be meshes, which is very limiting cookie-cutter construction.

But containment cuts are a huge part of every map. And not using meshes has limited how good looking D3/HL2 maps have been compared to UT maps. Too many people are making tiny little light switches out of 1 grid sized brushes. I do prefer meshes, although luckily D3 has patches which give some freedom inside the editor.

 

Do you want to edit the container easier, or the details easier? (Keeping in mind either is possible; it's just about which is easier). Do you want to build your map of meshes, or brush detail? These are the pertinent questions.

For me there's no reason not to use meshes if you have dynamic lighting/shadows because they no longer stand out like a sore thumb.

Edited October 27, 2009 by Stardog

[SneaksieDave]

No, but the height of the dragged brush would also need changed also before you hit Make Room.

Yep, I indicated that in my post on the previous page that in DR you,

1. drag a diagonal line, and then

2. click Make Room.

 

In UnrealEd it would be,

1. click Subtract

2. resize the subtraction brush (however that's done these days; back in Unreal 1's time, even this was a nasty chore of vertex manipulation).

[Mortem Desino]

For everyone's information: The new Unreal's subtraction is actually fake. The engine is still additive and you're carving out of "The big blue brush". Look at editing forums. "Fake-subtractive" makes lighting calculations slower because it's still an additive engine calculating light and shadow for "the big blue brush" that you just carved out of. Really good UnrealEd mappers will all tell you that Additive in Unreal is just your best bet for performance.

 

And Doom 3 doesn't use "static meshes" since Doom 3 is not Unreal Engine. And HL2 doesn't use "static meshes" since HL2 is not Unreal Engine. idTech4 has entities that behave in similar ways. Source has entities that behave in similar ways. But they are all called something different for a reason...

 

It really falls down to flexibility vs. simplicity for me. And I'm not calling simple "bad". You can easily prove that "drag one brush now there's a playable area" is fast. Think about this: what is the majority of Half-life 1? A series of indoor corridors, rooms, and rather simple outdoors areas. Hypothetical question: Why, then, doesn't HL1 use subtrative CSG geometry?

 

One last point. I'm a bit of an AutoCAD buff, and using the excuse "Subtractive can't leak" just sounds like an excuse to be sloppy and imprecise with your architecture.

[STiFU]

Man, some people seem to get really pissed by this topic... Nobody is saying that doom 3 or TDM sucks. It's just a general discussion, so please relax people. In the end, We love us all!!

 

[Wiseass-Mode]

Anyway, the word "static mesh" is no name of some sort, but a general term that can very well be applied to the corresponding elements in other engines, like "Models" in Id Tech 3. Whereas the "Entity" is a term too general, since it doesn't only describe static meshes, but also lights and stuff. So it is actually better to call them static meshes... :-) Just for clarification: In ID Tech 4, Static meshes and animated meshes belong to the group of models which belongs to the group of entities.

 

And well, that big blue brush thingy wasn't introduced until unreal engine 3. The previous engines were convex void engines, so subtractive, with additive support... (Not sure, whether what you were saying was limited to unreal engine 3)

[/Wiseass-Mode]

Edited October 27, 2009 by STiFU

[SneaksieDave]

Nah, no pissy here. Quite happy with TDM editing, love DR, and in fact it's the first mapping that's held my interest for so long.

[Stardog]

For everyone's information: The new Unreal's subtraction is actually fake. The engine is still additive and you're carving out of "The big blue brush". Look at editing forums. "Fake-subtractive" makes lighting calculations slower because it's still an additive engine calculating light and shadow for "the big blue brush" that you just carved out of. Really good UnrealEd mappers will all tell you that Additive in Unreal is just your best bet for performance.

You're right about the new Unreal, it defaults to Additive mode, but you can still use subtractions inside those additives. They're used all over the official UT3 maps.

 

It really falls down to flexibility vs. simplicity for me. And I'm not calling simple "bad". You can easily prove that "drag one brush now there's a playable area" is fast.

I'm still not seeing how UnrealEd-style subtractive isn't flexible...

 

Think about this: what is the majority of Half-life 1? A series of indoor corridors, rooms, and rather simple outdoors areas. Hypothetical question: Why, then, doesn't HL1 use subtrative CSG geometry?

It being based off of the Quake engine had something to do with it.

 

One last point. I'm a bit of an AutoCAD buff, and using the excuse "Subtractive can't leak" just sounds like an excuse to be sloppy and imprecise with your architecture.

Quite the opposite. Subtractive makes your map look much cleaner in top/side view allowing you to keep things much more in control without forgetting to select the odd brush that's floating around. I'm having to use layers like crazy in DR because there are so many brushes.

[SneaksieDave]

Perhaps demonstrate? I put a simple image for comparison above that shows the opposite is true:

 

Same result. Which is more intuitive? Which is 'cleaner'?

[Stardog]

Perhaps demonstrate? I put a simple image for comparison above that shows the opposite is true:

 

Same result. Which is more intuitive? Which is 'cleaner'?

The screenshot here shows a similar thing. http://forums.thedarkmod.com/topic/10064-subtraction-engines/page__view__findpost__p__198549

 

The door and room were subtracted from an additive. You could subtract a door/room from your additive cross.

[jdude]

In what way?

 

It's hard to explain, maybe it's my state of mind or something when working on maps with subtraction. However it feels like it's more difficult to do really detailed work within the editor like with a 1 .5 or .25 grid. With additive you just make your object filling in the spaces with subtractive I'm guessing you'd carve it out which I do not like at all and it gets really messy with a bunch of carves.

[STiFU]

Perhaps demonstrate? I put a simple image for comparison above that shows the opposite is true:

 

Same result. Which is more intuitive? Which is 'cleaner'?

Yeah, but that's a 100% subtractive approach. Combine both and you got it perfect... Subtract the room, add the cross. Done with four brushes, of which three are for the cross! This is the way to go. Definitely the cleanest approach and that's what I am talking about all the time...

Edited October 27, 2009 by STiFU

[SneaksieDave]

The screenshot here shows a similar thing.

Thanks. For a simple box room, no one will argue against the fact that it's easier to just subtract a brush. It's also easier to adjust the position of it, wall thickness, add a door and move it, etc. Not saying it's hard to do in Additive (it's not), just simpler in Subtractive. My shot shows an Additive build looking clean as well.

 

But my point has been mainly that that's where the advantage stops. What if you wanted to surround that room with more rooms, as if it were a building on a street? You'd need to plan and subtract the rest of all empty (and filled) space surrounding it. What a headache. What about something really simple: if you wanted beams in the room, what do you use? Addition (unless you're really a sucker for punishment and want to use Subtraction to carve around empty gaps and allow for beams). For all such positive manipulations, Addition pulls (far) ahead in ease, because you only have to grab the brush and move it. With Subtraction, any such a move would affect the empty space of the room, any intersected contents, and even the containing wall geometry. And all of this must be planned in advance. Because of that last fact, a purely Subtractive editor would be quite unwieldy -- Addition is needed to make up for the shortcomings of Subtraction. Subtraction specializes in building empty spaces. The opposite is not true for Addition -- the gamespace is empty already and you only need fill and seal it. Subtraction is not needed at all.

 

If it's argued from a purely "number of brushes" angle, subtractive can certainly be lower, as long as you don't try to make much detail. If you do, then the argument usually becomes a wash, both methods having roughly the same count. And that's part of the point, too -- they each specialize somewhat, but which is more flexible, intuitive, and world-realistic? Subtraction is best at creating the containing space. Addition is best at filling it. Subtraction alone doesn't fill it well, and such modern game editors instead depend on Addition to fill that need. Addition can contain space with a simple box and doesn't depend Subtraction at all. They're both useful for particular tasks, but the advantage goes firmly to Addition.

 

 

Edit: Yes, I am speaking about pure Subtraction (as stated). That's part of the point -- a method that can do it all versus a method that relies on its competition to get the job done. Want Subtraction in an Additive editor like DR? Use CSG. Not a great thing to do, but you can if you insist. Often there's no need for it, and it creates excessive geometry. If used carefully, it saves some brush clipping and that's about it.

[Tels]

People who debate about "subtractive vs. additive" always use the wrong examples. It doesn't matter (much) for cubes, but consider pyramids or more complex things.

 

Consider a churchtower top, consisting of two stacked pyramids (one with a more broad as base, the other as top). If you just want to make a solid tower, easy enough. However, it you want to make it hollow, it becomes quite impossible to build this with addition alone.

 

With subtraction, you just create two more pyramids, a bit smaller then the first ones, and subtract them, and presto! hollow church tower top. Even adding windows is as easy as creating two intersecting cubes that "cut out" the four windows on each side.

 

*That* is the power and beauty of subtraction, but unfortuntaly, people get so riles up on wether a cube is easier to subtract or add, that they completely neglet the real examples.

 

(Edit: Pure subtraction is also not a good idea, a mix is the way to go. Just saying)

[SneaksieDave]

People who debate about "subtractive vs. additive" always use the wrong examples.

Gee, thanks Tels.

 

I hope after I've said it what, four or more times, people understand that I'm talking purely subtractive or purely additive for the sake of argument, and to make the point that the two methods are not equally strong. Subtractive has a lot of very useful application (else no games and editors employing it would exist). At the same time, it can't effectively do the job without Addition helping out. If someone can't get beyond the hurdle of needing to fill empty space in games like this, then editing TDM won't be for them.

 

If you had to choose between the two pure methods only, the choice is fairly clear, I hope. Whatever technical reasons lie beyond my grasp and beneath the surface, Mr. Carmack picked Addition.

 

 

However, it you want to make it hollow, it becomes quite impossible to build this with addition alone.

Erm, no.

[Stardog]

What about something really simple: if you wanted beams in the room, what do you use? Addition (unless you're really a sucker for punishment and want to use Subtraction to carve around empty gaps and allow for beams). For all such positive manipulations, Addition pulls (far) ahead in ease, because you only have to grab the brush and move it.

 

If it's argued from a purely "number of brushes" angle, subtractive can certainly be lower, as long as you don't try to make much detail. If you do, then the argument usually becomes a wash, both methods having roughly the same count. And that's part of the point, too -- they each specialize somewhat, but which is more flexible, intuitive, and world-realistic? Subtraction is best at creating the containing space. Addition is best at filling it. Subtraction alone doesn't fill it well, and such modern game editors instead depend on Addition to fill that need. Addition can contain space with a simple box and doesn't depend Subtraction at all. They're both useful for particular tasks, but the advantage goes firmly to Addition.

You should always use additive brushes for things like beams. You'd never subtract 8 brushes to create a beam in the middle of a room.

 

I'm not arguing against additive geometry in itself, just additive only engines. I love adding loads of brushes, but not when I can't subtract from them. If the D3 engine allowed subtracting from additives I'd have an FM released by now.

[SneaksieDave]

You should always use additive brushes for things like beams. You'd never subtract 8 brushes to create a beam in the middle of a room.

We agree on that without doubt. And, truth be told I'd like some of the convenience Subtraction provides from time to time.

 

Adjusting to DarkRadiant will take some time for DromEd users, we can be sure of that. I'd hate to have potential TDM mappers turn away because they think Addition is limited or lacking in some way when it's not. Hopefully the tutorials we have up will help in this regard. And it's good we have this discussion going on here (hopefully it's not too heated? ), as I'm sure many will be wondering about the very same things.

[Baddcog]

 

Consider a churchtower top, consisting of two stacked pyramids (one with a more broad as base, the other as top). If you just want to make a solid tower, easy enough. However, it you want to make it hollow, it becomes quite impossible to build this with addition alone.

 

With subtraction, you just create two more pyramids, a bit smaller then the first ones, and subtract them, and presto! hollow church tower top. Even adding windows is as easy as creating two intersecting cubes that "cut out" the four windows on each side.

 

*That* is the power and beauty of subtraction, but unfortuntaly, people get so riles up on wether a cube is easier to subtract or add, that they completely neglet the real examples.

 

(Edit: Pure subtraction is also not a good idea, a mix is the way to go. Just saying)

 

Yeah, easy enough to make a tower roof that way.

 

Also quite easy with addition only (In Radiant - not Dromed), you need more brushes but with vert editing you can get very specific how those brushes interact. Would take 8 brushes instead of 4. But you could also make curved rooftops with patches.

 

See each has their own abilities. Dromd is easy to make a rough simple roof that way, but just as easy to make in Radiant (2 brushes, clone rotate 3 times) and can be more complex and look better.

 

I think the real underlying debate here is Radiant vrs. Dromed. Those are the engines in question, talking subtractive engines is pointless as Dromed is the only true one any of us would use. And most engines anymore are additive.

 

So the real question is:

 

Basic blocky terrain with Dromed or much more detailed and realistic terrain with Radiant (if you want to go beyond blocky terrain)?

 

All said and done, if you don't use the benefits of subtraction in Dromed and start with one big air space you might as well just start in an additive engine and one big open space.

As I said before, the ONLY benefit of subtractive to start is NO LEAKS.

[Stardog]

I think the real underlying debate here is Radiant vrs. Dromed. Those are the engines in question, talking subtractive engines is pointless as Dromed is the only true one any of us would use. And most engines anymore are additive.

I'm talking about UnrealEd vs Radiant more than anything, personally. I've never used Dromed.

 

 

All said and done, if you don't use the benefits of subtraction in Dromed and start with one big air space you might as well just start in an additive engine and one big open space.

But then you can't subtract from the additives And we begin again...

Edited October 27, 2009 by Stardog

[Mortem Desino]

Yeah. I wish idTech4 sometimes did have the convenience of an air brush. But no such thing exists in Doom 3.

 

CSG Subtract button is not an air brush. Don't treat it that way, and your map will be happier. The only time I've used it is when I'm carving out details to turn into func_static.

or when I need to precisely place a hole or a depression in a wall (for holding a torch or placing a doorframe)

[Weasel]

Then again, I loved playing with "Legos" when I was a little kid, and the additive geometry mindset fits nicely with that toy.

http://www.youtube.c...h?v=yTuba531Fp8

(Lol that youtube code has "Tuba" in it!)

 

Hey, that's my baby! What a nice surprise to be in here reading about the Dark Mod and find a link to my own thing. I really should finish it some time...