Final Fantasy VIII - Balamb Garden
polycounter lvl 14
Hello Polycount! My next personal work will be based on Balamb Garden from Final Fantasy VIII. I have an emotional attachment to this game, its scenery and music so I'm going to try and have fun with this one. While most any environment from this game could be a greta piece, I feel that Balamb Garden sort of defines FFVIII and set the tone right away. Heres the concept art and some screenshots:
and the music:
http://youtu.be/JfdwMIiUork (first four minutes)
I always welcome comments as long as they are constructive. Some of my initial thoughts and goals:
1) 3ds max, photoshop, mud box and Unreal 4 are the software. I'm toying with the thought of the 30 trial of Marmoset, but if I do I would like to make good use of the 30 days to re-render a couple other props I have.
2) This will be my first time dealing with a large environment involving modular pieces and tiles in a game editor.
3) Modeling is my strong suit, material definition and presentation are what I'm more concerned about. Help me out here ... I won't call it done unless the materials and lighting are good.
4) I will need to make quick decisions about how closely to follow the concept art and/or the original game art. How big will it be? What colors? How detailed? How clean/dirty? What are the materials made out of?
Thanks
and the music:
http://youtu.be/JfdwMIiUork (first four minutes)
I always welcome comments as long as they are constructive. Some of my initial thoughts and goals:
1) 3ds max, photoshop, mud box and Unreal 4 are the software. I'm toying with the thought of the 30 trial of Marmoset, but if I do I would like to make good use of the 30 days to re-render a couple other props I have.
2) This will be my first time dealing with a large environment involving modular pieces and tiles in a game editor.
3) Modeling is my strong suit, material definition and presentation are what I'm more concerned about. Help me out here ... I won't call it done unless the materials and lighting are good.
4) I will need to make quick decisions about how closely to follow the concept art and/or the original game art. How big will it be? What colors? How detailed? How clean/dirty? What are the materials made out of?
Thanks
Replies
Good luck and have fun
good luck
Make sure you blockout everything first!
@selwyn369: good call
@gsokol: oh yeah.Let me know if the spacing seems off.
Carry on!
*watching*
Good blockout! Really wanna see how you tackle that curvy-ass exterior!
BTW you could import Squall, make him 1.8 m and compare scale with references. This helped me a lot.
Also this fmv is a good reference how roof and main entrance are made:
https://www.youtube.com/watch?v=jWrAZhemws4
+9999
Thanks everyone. I was happy to find a saved game on my PS3 copy standing right outside of Balamb Garden. This way I can quickly walk around having it on the TV while I'm blocking out.
Its crucial to get this part right. After getting many of the small pieces sized, I realized the circular hallway wasn't long enough (I was squeezing the benches in). It confused me at first until I realized it was the central pillar's radius being too small. I'm thinking my next step is to delete all but one of each small props along the arched path, remake the arched path to the correct scale, then re-size the central pillar and its ground, then place the small props in again. This is why you block out first! :poly009:
http://ffdream.com/downld-mapsff8_1.html
They are at native PSX resolution.
The main entrance path is huge in contrast with the other seven paths. This makes sense artistically with the size contrast, however it means the room is not a perfect circle. The in-game camera angles never explain it behind the main entrance. It worked for them. My choice is to either simply do as they did and never make the wall behind the main entrance (and obviously never point the camera back there in the folio shots), or to create a break in the circle as demonstrated below...
For making it in actual 3d i guess I'd go for the second option, and maybe even go as far as to make up something like a hallway or some kind of entrance, even if we don't know exactly what's there from the ingame shots.
Anyway you're making some seriously sweet progress on this and I enjoy watching people try to solve issues like these!
Next issues. Keep in mind this is my first time dealing with modular pieces snapping together. With the first image, I figured out how to evenly use edge connects on a big 128 sided cylinder to turn it into 8 snapping pieces. Then in the second image I realized I still need to curve the height planes out. Even after using normal constraints and vertex snapping, you can see a small gap remains. This means I can't edit the exterior edges at all after creating my module. I'm open to any and all suggestions, however my primary thought is to design around it. I've heard of modeling for UVs, modeling for normal maps, so it would follow to also model for modular re-use. In the bottom-right corner of the second image, I placed green dots where there could be ridges along the exterior indicating how it was constructed together in pieces. Maybe I should use this to my artist advantage?
Edit: After thinking about it more, I'm leaning toward the conclusion to stop worrying so much about the low poly. I have basic cylinders in place to define the spacial relationships between major objects, so from here I think I can start seriously modeling. Afterwords I will likely have to remodel the low polys, again they're just cylinders.
While continuing to slowly work my way around the basic shapes of the major environment chunks, it's good to be seriously modeling again :poly009:
I'm going to ask for advice/input on modeling the fish statue. How would you do this? If it isn't very obvious this image is sort of that moment where you are staring at something, moving slowly, researching refs, trying one thing, then another, and still aren't sure which direction to go in. I'm using 3ds Max and Mudbox. Should I make it all in 3ds? Should I sculpt? How much should I sculpt? Only scales? The basic shape? I'm still finding Mudbox' limit in terms of how much of a starting mesh you need to have ready before sculpting will "work". Should the fins by the head be connected or separate? How do you set up a mesh best for Mudbox to not struggle on some areas then others? I guess I'm torn between the choices of 1) "add more details because it can't move polys that far or effectively" and 2) "use a clean evenly spaced mesh with no details so you can sculpt it better".
Also the original game has two fish, with the tail forward and back. Would it seem like I wasn't trying hard enough if I only made one? Would it help that much to have both? Here's my Pinterest page with just a few refs on top:
http://www.pinterest.com/scottmichaelh/inorganic-references/
I'm considering that my hesitation could simply be that I can't see the statues in the game well. May I should let it go and follow some real life Japanese Fish Fountains more? Any and all thoughts welcome :poly142:
DO IT!!! :thumbup:
http://i.imgur.com/72gkzx4.png
//will start my own thread rather than hijack this one, unless Scott wants to do a sweet ass collab.
I'm game if you are?
The fish statue is going to take me a few days at least. Without the textures its a creature model so it will stretch me more than the rest of the scene in the modeling. I decided to take a break on it for a while and spent quite a bit of time on the circular blocks. It took me a long time but I eventually figured out how duplicate the pink ground and gray arches. I could duplicate in one direction but had a mental block about the center-front parts that needed to be mirrored.
I'm starting to wonder about how close to the original my high poly models should be.
Having a major creative block at the moment as to how to deal with the bridge, circular hallway ground, and the curved height of the structure coming together here. Any thoughts?
I can honestly say that's the first time I've exceeded the 10MB bandwidth for a month :poly142:
The rest should reset in a few days...
ref: http://www.ffdream.com/downld/images/decors/ff8/050.jpg
How is the cloth sculpt? Does it read as cloth? Keep in mind that it is a prop meant to be seen from a distance.
and then for the hanging cloth, either straight up sculpt it, or use marvelous designer to simulate it.
Thanks. Good call on Marvelous Designer. Today I managed to find it, figure out what it was, watch tuts multiple times, download the 15-day trial, try it, struggle, try 3ds Max cloth, give up on that and go back to Marvelous Designer, watch tuts again, figure it out and get both pieces done in Marvelous Designer. I think this result is much better.
I can sculpt this into infinity at this point without a second opinion and never stop so here's my best effort. I've learned a lot about Mudbox, the importance of scale, the usefulness of retopologizing, etc. I'm not sure about the scales.
Edit: I think the main issue I'm having is I can't decide what it's made out of. I can't stress enough that "I can't tell what it's made out of" is the clear winner for my "most common critique" award. From anyone's understanding of the game or common sense, I'm thinking either brass or stone ... not sure what it should be. The graphics of that day made everything look like plastic and the water reflection is the main thing I see on them.
In both cases i'd suggest hardening the edges up a bit and getting some nice strong definition in there. in the case of metal, some areas will be more curved (like where the fins join on to the body for example) which is a biproduct of the casting process.
here's my progress, going a bit slow as i only do art on weekends now, but getting there. need to do a lot of work on the face, it's not quite there at the moment.
Welp. That escalated quickly. You've been making awesome progress, Scott, seriously keep it up. Remember that someone had to sit down and sculpt this statue for casting, likely, on top of what I mentioned in Lunchcrunch. A hearty +1 to everything almighty_gir said, too.
Gir, looking great so far. Looking forward to seeing more progress!
Also, subbed.
Thanks. Here's the next sculpt (Mudbox cavity and AO turned on). Is it readable as a stone sculpture? Does it look better with harder edges or did I not take it far enough? Too far? How about the scales? How far should I take a prop such as this? I want this piece to scream, "Give this artist a job!" but it is a prop in a large room. Any additional thoughts are welcome at any time. I'm thinking I will likely NOT make the interior of the elevator or the details on the front entrance (seen only in the Zell FMV). My reasoning is to balance "time/energy spent" with "what is needed", and I am leaning toward the portfolio shots being of the original game camera angles to really sell what it is and please fans. However, I would want to run such an idea past the forum first before committing for sure.
@Bagelhero: I was hoping you'd show up
@Almighty_Gir: Looking good as expected. :thumbup:
Maybe these links could give you an idea of were to go with the scales.
Minor update. Do you feel I should have to make the interior of the elevator? Do you feel it would be good enough to make the door a solid color and the blue "tube" above it either see through or simply have enough refraction or whatnot to hide the inside?
At this point I believe I'm primarily done with the high poly. Any final thoughts or warnings. Feel free to stop me if you feel you should. I can't really see much to change (Although I did simplify the arches over the exit sides ...didn't like the original ones). I get sort of a Star Trek The Next Generation feel from this place. I don't think throwing small details in for details sake is the key. Again any comments welcome.
http://www.polycount.com/forum/showthread.php?p=2138815#post2138815
Here's some spam for a couple tech threads I've started relating to this project. The second one is especially difficult for me. Any advice for someone struggling to tile circular modules is appreciated. I've seen a couples tuts on modules like Minos', however if these were box shapes, it would be easy. They aren't ... they're cylinders with extensions. I've struggled to figure this out since day 1 on this project. Thanks
Make a dummy helper in the center of the model and connect the meshes to that?
I thought of that , tried it multiple times ... it ends up very slightly off. I'm beginning to think more and more that I just need to stop over thinking it, make a giant object, then figure it out later for the colors.
As in you don't know how to perfectly match corresponding vertexes of one mesh to the other.
Have you tried snap to vertex?
Sorry if im not much help but im kinda having a hard time understanding what the real problem is lol.
Snap vertex would give a slight variation in the placement as well. The issue isn't that I can't put a scene together on 3ds Max. Obviously I can simply grab something and weld/collapse/bridge it in. The issue is trying to create a perfect module that is one piece that could be duplicated in UE4 as 1/6th on the whole circular walkway to maximize efficiency. In other words, other than the front entrance piece, the rest is one piece that would have six instances in UE4. In order for that to work I would need a pivot set in the correct place to simply rotate a bit, and its sitting perfectly in place next to the other one without the slightest hint of overlap or space between. If I moved the vertices on the edge of the piece to accommodate, the instance would be ruined.
At this point I'm most interested in how someone else would handle this type of situation as it seems that I wasn't doing it the way others would.
This is one way you can make modules of that walkway.
The ones labeled 1 are the same 2 is just 1 flipped 3 is unique.
Instead of cutting the walkways heading to the sides in half cut them off the circle as a whole piece and repeat it 7 times.
Obviously to make it work as modular pieces in udk you will have to do a some coordinate gymnastics with the individual pieces but im sure its not that crazy hard to figure it out.
I thought about doing it that, my change to this plan was based on leaving the exit path clean for easier texturing. The Unreal "gymnastics" is what I'm concerned about. At the moment I have the floor combined as one object with all vertices collapsed together.
Clone the circle piece, delete half of it and then delete half of the half.
You're left with a quarter circle.
Split that as indicated with 1 and 3
Invert 1 on the z axis and you have 2, snap that into place
rotate 1 90 degrees snap that into place
Copy the whole thing invert on the x axis.
You should have a whole circle. xD
Of course im assuming everything was made to fit in 45degree arcs.
Finally in Unreal! That's quite a relief :poly003:
However, as you can see in the 3rd pic the walkway is too narrow and this is after I painstakingly placed the circular modules in a circle by hand. You can see on some how I modeled in some grooves and overlapping end pieces to deal with the issues I was asking about in regards to lining up the modules "perfectly". Thanks
I'm very happy to have fixed the issue without "too much" trouble.
Fortunately Unreal's scaling of multiple objects at once works well. I also thinned down the dividers between the walkway and the inner pool. The plants aren't in yet as they will be a new challenge for me. I plan to deal with them later after the bulk of things are in place. Next I need to set up a second UV channel for everything. I also need to begin deciding on what the materials are. This will be a clean environment with tiling materials so I will create materials in isolation ... something I've also never done before.