I'm looking for some ideas to finish this off.  I've added some of the decorative swirls and stuff in an effort to give it some flair before texturing, but I really would like to do something with either the seat, or the interior of the dome.  My polycount so far is decent.  I'm right around 27,000 polygons after smoothing, which isn't bad at all.  My self-imposed budget here is right around 50,000.  So I have room for something fancy.

I don't know, I was thinking of maybe some "dragon" feet or something on the front flat parts of the bench, or maybe a dragon head at the top of the dome?  Then again, I don't wanna spend too much time on this thing, as I've got some other projects planned, and with school and all, my time is really pressed.  Any ideas that can be done fairly quickly?

What abot a couple of bird baths either side of or just in front of the seat.Could be built in or on a plinth.

Thanks, Ace10.  I might actually try that.  I went ahead today and totally started from scratch on the dome part.  I'll keep the swirly designs for the front faces of the dome, but I really didn't like the ornament at the top there, and since it was modeled into the piece, I couldn't easily change it without some issues.  So I'm starting that part over, and will just add like a lion head or some ornament to the top center, which should look nice.

The bird baths are a great idea.  I'm going to add them as separate elements.  This is turning into a rather complex unit.

I've decided to re-model this object from the ground up, as the adjustments I had in mind late in the development of the previous version just weren't suited to the topology I had created. So I blocked out an entirely new model, starting with a simple sphere, and working down to the base from there. It's all one piece so far, but I did save some of the design elements from the previous version to use in this one, if I decide to use them.

This time around, I decided to start out with some relatively basic beveling, and edge chamfering.  My plan here is to build it up with some really complex extrusion and beveling work, just as a personal challenge, to see how detailed I could get it as a single geometry. I'm keeping the poly budget at 50k for this one.  I doubt I'll come anywhere near that, unless I really go crazy with very small details, but I'll prob just save most of that for the textures.  Lots of work still do do yet, but I'm definitely keeping the idea of the bird baths, and maybe expand on that a little as well.

Dimensions and scale are about the same as the original.  It's a pretty big object.

I like that dome on the second version a lot better.

Bird bath, I like. What I'm not keen is that massive pole at the top!

Quote - Bird bath, I like. What I'm not keen is that massive pole at the top!

How would you remedy that?  I need the pole sphere to create the dome, because of how I'm going to run the masonry blocks on there, using chamfered edges.  A quad sphere with no poles won't allow the flow of topology I'll need.

See this example:

http://greatwenlondon.files.wordpress.com/2011/04/sdc11929.jpg

Instead of the scroll at the top,would it look a little more in keeping if you included a key stone  at the top ?.Dont know if you call it that in the US.It is the load beaing stone that keeps the arch together.Just my oppinion but I think the scrolls on the front are just a bit to much,but I am enjoying watching you.

Oops sorry ignore my last post,just saw the photo in your example,but even there you can see the key stone which is still a little proud from the arch.

Good point again, Ace10.  I missed the keystone, and it's a must-add for "structural integrity".  Definitely going to incorporate that.  I think Airflamesred was referring to the "pole' of edges created by the use of a pole sphere for the dome.  However, i think I need to use the sphere in this case, because I intend to duplicate the look of the "stone seams" you see in the photo example of a similar structure.  The flow of edges on a sphere are perfect for duplicating that look, whereas a quad sphere wouldn't have the tapered flow of edges to create that look.

I'll update it later tonight with another screencap...

Well, I was partially joking, Lux, knowing your fondness for the quad, but Ace10 has a good point with the keystone.

I'd be tempted to split the mesh into seperate stones. That said, I'm getting lazy with 3d atm.

@Ace10 Are you in Derbyshire?

Anyway, I started blocking out a keystone now.  Don't mind the Ngon here, it's only temporary while I extrude the stone.  I'll probably re-attach the scroll to the top after I finish this part.

Keystone with scroll attached.  Working now on adding in a lot more detailed bevels and chamfering, then I'll start on the birdbaths, which I plan to make quite ornate.

Quote - I'd be tempted to split the mesh into seperate stones. That said, I'm getting lazy with 3d atm.

Mark, just for fun, I was playing around with a little free maxscript tool I came across a while ago, which I think addresses BOTH of your quoted assertions.  It's a Wallbuilder, which works to create walls from individual blocks automatically along any spline shape, and has fully adjustable parameters for the position, rotation, and placement of all the blocks.  It really is fast and easy to create individual blocked walls, etc.  I was originally thinking I might impliment this script to create the BASE of the alcove, but this totally takes away any ability to create recessed beveling or ornate designs in the mesh.  Looks great, though, if you want simple bricks or blocks for a building base.  Much better than say, a displacement or bump map effect.

Not going to utilize it for this project after all, but it's a really instant solution, if you're feeling lazy. ;)

Luxxeon, that's NOT a keystone.  Not in the architectural sense of a dome.  What you have modelled there is simply a header block at the top of a dome arch.  A keystone, in dome architecture, would be a slab of stone which is at the very center of the dome, and holds together the next two segments.  That block you added isn't that.

If you look at your reference (or the inspiration I should say, since you aren't modelling that exact structure here), you can see the bond areas between each stone in the dome.  You modelled them yourself in your FIRST version of this.  Well, at the very top, in the center, where the verticle bond ENDS, that is your keystone.  You did it right in your first version.  You can still do it in this one if you want to, and maybe you were planning on it anyway, but just letting you know, that block wouldn't be a keystone.

Unless you add the running bonds to indicate that there are masonry units involved in creation of this dome, which would meet at the apex, or keystone, then what you have here is not a true arcuate dome.  It's what we call a "false dome".

LuxXeon sorry if I mislead you about the key stone,I always thought you needed the key stone to take the strain as it were over the top of the archway.Like it though,much better without the scrolls.

I think this is looking more achitecturally correct, Lux, and would be easier on the UVs.

I just added some more details to the mesh.  I beveled out some stones along the cornice part of the dome; there's also a pendant now inserted into the top of the dome, under the scroll (this is separate geometry).  I also modeled in two buttress blocks in front of the bench area where the bird baths will eventually sit; I inset some areas around the base, mostly for visual affect, nothing structural.  Then I also inset two deep bays in the facade of the base on either side, as you can see in the screencap, where I'll also add vase-like geometry, or possibly small columns for even more detail.

Hope you guys like it so far.  I'm continuing on now with the birdbaths, which I think will add a striking visual impact to the structure overall.  Well, at least I hope they will.

I'm doing really well so far on my originally projected polygon budget of 50k.  So far, I'm now at just 8430 total polygons.  This model will not be all quads; obviously I had no choice but to add some tris here to get the shapes I needed, but the ratio is still heavily in favor of quad topology.  I have exactly 8268 (98%) quads, and only 162 (2%) triangles here so far.  No NGONS, of course.  I'd rather poke out my eyes with jagged shards of glass than to leave even one ngon in a finished mesh.

Anyway, the polycount will get exponentially larger, especially once I add some chamfering to the outer edges, and continue with more geometry additions, but I think keeping this under 50k shouldn't be an issue at all, and should allow for a very manageable, yet detailed piece in the end.

Birdbaths are in progress.  I'll update a WIP shot of those tomorrow at some point.  Gonna get some sleep now; early class tomorrow.

Quote - Front view, showing the new inlaid pillars to the facade bays at either side.  I might doctor these up a little with some additional beveling if I feel like getting crazy, but I think they look fine as simple pillars for now.

Birdbaths are in progress.  I'll update a WIP shot of those tomorrow at some point.  Gonna get some sleep now; early class tomorrow.

Not bad!  I like the idea of the columns.  They really add a fresh look to it.  I also like how you built plinths for the eventual birdbaths in there.  I was trying to imagine how you would include those.

I can see you're not following any architectural style for this.  More of a fantasy build. It's looking good though.  Definitely much different direction than I thought you were going..

So I gathered up a few reference images from Google, and decided on doing some kind of horse-like beast head at either side of the entrance, facing each other.  As you can see in the screencap I posted above, my choice of reference for this was a chess "knight".  I've already begun outlining and filling in the profile of this addition with quads.  I'm not sure, at this point, if I'll use subdivision modeling to create it, or just keep it semi-polygonal, as if roughly carved from the stone.  I'll probably decide that for sure later this evening, as the modeling progresses.

Anyway, any thoughts on this change of events is welcome, but I'm pretty sure this is where I'm going with it now.  Again, the birdbaths were a great idea!  If I hadn't already put the columns in those recessed facade bays, I would have probably gone with them.  I think these heads will add some real complexity to the overall structure, as well as a fantasy appeal.  Will see.

The horse heads look good,but just a thought ?.Are you staying with the stone seating,or will it be wooden,bench like.I think the wood ,would look a lot better( no pun intended).Looking really smart though

Good idea.  I was gonna suggest maybe gargoyles or foo dogs for those spots, but the horses should look cool..

You think this change will impact your poly count goal?

I don't have eyes in the eyesockets yet, obviously.  I have a plan to model those as separate geometry, so I can put some extra detail into them.  There's still a chance I might subdivide this for smoothing, but I'm not sure it will be necessary, as I've modeled in enough subdivisions here to make it smooth enough, especially for a stone carving. Comments and suggestions welcome, as always.

Impressive.  I think the base of the head could be squared off some to match the plinth beneath it, but it's lookin real good.  Very slick modelling overall, Lux.

Fun model.  I got to test some organic techniques combined with hard surface here.

I think the best route to unwrap this thing is to do it in pieces.  Currently, most of the alcove is one large, manifold geometry.  I'll break it down into logical pieces, remove unnecessary faces (like the polys beneath the base that will never be seen), and unwrap each one individually for texturing.

Hope you guys like it; I'm fairly pleased with it, especially the horse heads.  Organic modeling is still a fairly ambiguous modeling genre to me, so this project afforded me a great opportunity to really dig into some organic modeling techniques, and combine them with familiar hard-surface stylings.  Can't get more fun than that!

Looks great Lux

I agree, it looks great.  I think 5 hrs is a little too long for a model like this, because that doesn't include the UV and texturing yet.  But considering you did the horse head polygon by polygon, I could understand why.  I would have used box modeling, or subdivision methods to get there a lot faster, but your results are first-class.

Just think, if you ever decide to model a full chess set, you will already have a really great knight piece ready to go.  Bonus!

Fastest method would have been sculpting then retopology. You could have sculpted the form in less than an hour and started your final retopology right after, leaving the rest of your time for the UV/texture portion. It's fast becoming the omre popular way to create content.

Quote - Fastest method would have been sculpting then retopology. You could have sculpted the form in less than an hour and started your final retopology right after, leaving the rest of your time for the UV/texture portion. It's fast becoming the omre popular way to create content.

That's exactly what I ended up doing.  The 5 hours was the entire alcove model, from start to finish (except for unwrapping and texturing). 

As for the horse heads, I originally planned to use edge extrusion over a photo, as I show in the screencaps on the previous page.  I'm very comfortable with edge extrusion techniques, as I feel it provides the ultimate control over countouring and profile shaping.  Max has some tools to make edge/poly modeling very efficient, and in some cases, it's exactly the same techniques and tools used in retopology, minus the step of actually sculpting something first.

Ultimately, however, I wasn't satisfied with my reference material.  Didn't have enough ref photos to work with, so I achieved the final result of the horse head using Sculptris to rough out a quick sculpt in about 15 minutes, then imported that into 3dsmax, and used the native retopology tools to build a nice quad version of the mesh in about 20 minutes.  Then I just simply subdivided, and refined that shape a bit using standard subD techniques.  I finished the horse head in under an hour.

Sculpting and retopo are the way to go nowadays, but I disagree it's any faster than box modelling.  Mostly that's the technique used because you can get a lot more detail in the sculpting apps.  But most workflow for that starts with modelling the base figure first in a standard 3D app anyways, then detailing and refining in the sculpting app.  I don't see how that process is faster than straight box modelling.  The end result may be more detailed, but faster?  Nah.  Someone with good skills, and the right package,  could box model a decent head in subD just as easy and fast as a guy sculpting one from a ball in Zbrush.  Gollum was all box modelling, and I think it still sets the standard for great CG characters to this day.  Nothin against using sculpting apps, just sayin.

I've got sculptris, but for the life of me, I have to have a base mesh or I'm going nowhere.

Quote - I've got sculptris, but for the life of me, I have to have a base mesh or I'm going nowhere.

Sculpting, and voxel sculpting software, is a different world from traditional modelling, and even though it's the hot thing right now, usually those who specialize in sculpting aren't the best modellers, and vice versa.  There's people who can do both very well, but usually it's one or the other.

I'm a hard surface and architectural modeller myself, but I can do organic if need be.  Sculpting is even a step removed from that.  Organic modelling, where I come from, consists of subD or nurbs modelling.  Sculpting and organic modelling are now interchangeable words in the industry, but they weren't always, and I don't consider them the same.

I'd actually prefer Mudbox, if it wasn't Autodesk, to Zbrush or Sculptris.

This model is now complete, and available for download in Freestuff and other resource locations.  Here's some shots of what the final wireframes look like, so you can preview the topology, and get an idea of the polygon distribution:

Front:

Side profile:

Back:

You can download the model here, if you're interested.

'tis fine work, once again Lux. I can see this being the background for many a naked, sword wielding women! Poser folks will love this

Very impressive textures.  You created a fine model here in the end.

Quote - 'tis fine work, once again Lux. I can see this being the background for many a naked, sword wielding women! Poser folks will love this

Funny.  I was thinking the same thing.