I am taking my final major class and I am starting my indep study project this summer to get a bit of a jump on it. HOPEFULLY I will get the project 80% done [that is my goal, realistic or not] before I meet with my prof for the first class. The reason I want to get most of it done is because my prof isn't qualified at all to help me with the project. He teaches photoshop and digital photography and admitted to me he wouldn't be able to help me with my 3d indep study project at all so it's all on me. He approved my idea for the project which is:
a medieval scene complete with a castle, various people, props and horses [he wants me to animate them which gives me a panic attack just thinking about that]. The big picture is a castle scene with people doing stuff, like buying cloth or spices and horses pulling carts. I'm GUESSING he thinks in the very end it will be a spin around video demo thing.
Authenticity is a big thing so no steam punk sort of stuff. I'm not sure how anal he wants me to be, I can email him and find out. I mean I'd love to make stuff up in like the spirit of the world but not make it crazy arty like WOW you know?
I need to learn over the summer proper unwrapping, texturing and normal map generation. I don't know what generation of tech he is expecting but obviously it can't look like DOOM 1. I'm thinking like the Return to Castlewolfenstein, that sort of dark look to things. I mean I don't know if he wants me to try and get a spec and AO and cavity maps on everything. Realistically I have bitten off way too much than I can chew but throughout my college career I have always done that and succeeded, despite that you people repeatedly tell me I'm a complete moron.
Anyways I am gathering proper references and trying to plan out in my head what the scene will look like. I want to start blocking out things tonight after I move out of the dorms today.
Wish me luck and I will be posting more in the future with this thread.
I am going to look for some tutorial dvds now. Like Eat3d, DigitalTutors or something. If I F up my indep study, I cannot re-take it and really then I have to change my major which is like super major you know? lol
It sounds to me like you have given yourself way, way too much work. Creating a large building, props, and a few characters is months of full time work for an industry veteran. I would really try to scale it back if I were you. Rigging and animating characters is completely out of the question if you haven't done it before, you're talking weeks-months just to learn how to do it.
If at all possible think more along the lines of a small scene, a medieval weapons shop or something as you have more experience making weapons. One character(shopkeeper) at the absolute max. Modeling, uving, baking, texturing, lighting and setting up even a small scene like this will eat up a lot of time, especially if you've never finished a scene like this before. Make it a weapons shop right outside of a castle(not to stray to far from the original plan), so you can have a partial castle wall represented behind the shop, but not the massive amount of work(months) that it would take you to build and texture a full castle.
How does it need to be presented? Do you need to walk around it/demo it as if its a "level"? A fully working(even just so you can walk around in it) level can be a massive amount of work. If at all possible concentrate only on what needs to be visible for presentation, creating a "facade" instead of a level.
It sounds like neither you nor your teacher understand the amount of work involved in this, so I would really step back and re-evaluate it. Its better to bring up the fact that this is totally unrealistic to accomplish now, than 2 weeks before it is due.
The first thing you will need to do is make a FAST and basic blockout of the entire scene, all the buildings, props, characters etc need to blocked out with basic modeling(primitives etc). From there you will need to make a list of all assets, and write down a time estimate for each require asset. Be realistic with yourself when coming up with time estimates, pencil in MORE time than you think it reasonably would take to account for any problems that may occur, pencil in a TON of time for things you've never done before. To pull something like this off you will have to diligently stick to a plan.
If your teacher admits that he isn't qualified to help with the 3D aspect, then he probably isn't qualified to estimate the time-sink that all of that is.
Yeah I'm listening
doing rest of castle block out now
plan on the castle, a wall around it that contains a few simple merchant shops and a few milling people.
maybe a a big tree or two
that's my goal at least
watching a bunch of DT videos in my free time basically trying to spend most of my free time learning this stuff till my brain falls asleep and then I start over again
Nice small castle! Just one brotip: use the Photoshop CameraRAW plugin for remove noise from render pictures. But i nothing say, if you keep noise the part of design
There's a much better way to go about building the castle. Try making it modular with a tiling brick texture. You can make each tower a separate piece of geo, along with that front wall and all those towers and what not and just build it out of those pieces in UDK. It will save you a lot of time. Don't make the castle one single piece of geometry
Yeah I'm listening
doing rest of castle block out now
I fail to see this when the advise given was to start with something smaller, like a single shop and merchant instead of building the whole thing...be realistic, specially since you're learning as you go.
This is already going down hill by the way you are building the castle...
concept artists, prop artists (doing small props), character artist, texture artist, level layout, level detail....
But going modular can save a ton of work. 1 character and 3 heads...1 hairpiece, 1 helmet, etc...
I'd to the same with the castle pieces. Make one tower, bake it or whatever you decide (tile textures, etc...)
But you said you were going to try and stay Authentic. You need to look at some real castle refs. While the tower might look pretty cool I don't think it looks like an authentic design.
Not trying to sound like an asshole - but how can you have 1.2k posts on this site and not know what UDK is and not know what modularity is?
I think you should go to the wiki and have a decent read through all the environment modelling info before you spend any more time on this castle.
Ok I skipped ahead and am reading UDK materials from the site
seems complicated
and watching this video
think it'll be a good reference guide?
[ame="http://www.youtube.com/watch?v=Q7MJTYLePF8"]Maya to ZBrush to UDK - Part 25 - Bringing in the mesh and setting up the material. - YouTube[/ame]
Looking at your ref you could get away with having your spire mostly unqiue (still using smaller modular chunks) and the dome. The other buildings however is going to be wayy too much to do for a unique environment and you will 100% need to plan it out ahead if you are ever going to finish it. Also you will need to sculpt a terrain.
Is the actual scene going to be composed like it is in the ref? if you are going for exactly the same camera angle you could easily cut the tri count of what you are already doing and put it on very small tex sheets if it really is going to be seen from a distance like this, but I get the feeling it is not?
Either way good luck but I would strongly consider what other posters have said and go for something much smaller and modular based. Dont run before you can walk.
A really simple explanation of modularity is just using one piece as many times as possible.
For an ai you can use one body, but have pieces of armor to make many ai from it.
For a building you can make a wall section. Then make a copy and add a window. Since the first wall is already uv mapped and textured you save time. Then make another and add a door.... Then you can slap them together in any combination.
That's very much how most game editing is done.
UDK is unreal Engine Dev kit. It is a very indepth engine (I haven't used that one yet myself). But it is basically a simple 3d editor (you make the walls of the castle and ground, the main play area, then you slap on models to decorate. door frames, window frames, stairs...) and then it will have scripting tools etc for making a playable game.
Not great for making models, but really good for base terrain and game stuff. Models make it look pretty.
I'm going to try to bake one of my towers. I added bricks to it, a window frame, roof tiles etc and just having it in Maya all HP makes the program lag a lot and also crash from time to time.
I have also made other pieces like a curved bridge, other walls, etc
Focus on baking the towers and bridge if I get time for the bridge
Set up my tower in my room and install Maya, UDK and Marmoset and get serious
why udk and marmoset? pick one and stick with it.
and the low fps in maya can be rectified using a modular approach. One HP piece, One LP piece, once the bake is done discard the HP and assemble the LP pieces.
Probably in regards to building in UDK should include the final render in UDK. Rendering the final in marm after all that would be like throwing it away.
I'm more concerned this is out of your scope. Through this thread there seems to be a lack of baking knowledge of basic props, even smoothing groups, yet tackling a whole castle with no knowledge of modularity, realtime art, UDK, etc. You need about a year's worth of practice before such a huge scene is accessible to you.
ok nevermind
doing good so far i think
beyond i still can't properly unwrap and object in maya...
got about 2 gigs of udk tut videos
pretty helpful
a few questions later i will ask but i think i made good progress today
pics of my bridge in a bit. imported to udk and all that but shows up screwy with the white and purple texture on it.
i saw a bunch of the material editor videos but i never saw him select the object and assign the material to it in the videos, he was always applying the materials to a primitive shape like a cube or sphere. i'd love to know after i import a mesh how i assign a material to it. ??
ok good I will try that
I have made up a simple texture for my bridge but dunno how it will fit on the object in UDK. Give it a go
also have to make a dirty normal from my texture, I have that tutorial around here somewhere
post up pics of Maya bridge later on today
also registered on UDK epic forums. take a while for them to OK my account but then I'll be getting it done right
playing in UDK some more I realize I need to pretty much redo all my work in Maya and really do the modular approach
it took me about 3 days thinking about it to realize what I need to do to make this project come together
instead of questioning WHY this and WHY that I need to just work within the conventions already established
I have successfully unwrapped, normal mapped and applied diffuses to several wall pieces in maya and hope to
redo the meshes in maya in a modular fashion and then assemble them in UDK
things are coming together quickly now
it tiles I just have 4 planes there and i duplicated them
I think he meant it doesn't tile *properly*. You can tile planes side-to-side infinitely but it is still going to have seams if the texture doesn't tile properly at the edges.
I also question the tiling (sorry), but the 8th brick down on the left side is half a red brick. 8th on the right side is mortar. and I don't see a matching brick in that row (unless it's off the right side somewhere).
In which case I'd just got it sloppy presentation as that none can tell that it does in fact tile. Which imo would be the point of showing it.
You can clearly see that some of the bricks have a very different size from all the others, and the colors don't flow nicely across the seams either. It doesn't tile.
edit: hah, Skillmister, you thought what I thought while I thought up those thoughts.
So it's not perfect? No big deal
That's the least of my concerns to be honest. I know the normals, diffuse and specs work and look good in UDK so that's all I'm worried about. The modeling is not bothersome, mostly working in UDK is.
I still need to watch an a55load more of tutorials. Heck, I don't even know how to duplicate objects.
Replies
Normal map made in PS from the diffuse didn't turn out all that good so I'm thinking I'll do it the traditional way.
http://www.polycount.com/forum/showthread.php?p=1524512#post1524512
oh man.
a medieval scene complete with a castle, various people, props and horses [he wants me to animate them which gives me a panic attack just thinking about that]. The big picture is a castle scene with people doing stuff, like buying cloth or spices and horses pulling carts. I'm GUESSING he thinks in the very end it will be a spin around video demo thing.
Authenticity is a big thing so no steam punk sort of stuff. I'm not sure how anal he wants me to be, I can email him and find out. I mean I'd love to make stuff up in like the spirit of the world but not make it crazy arty like WOW you know?
I need to learn over the summer proper unwrapping, texturing and normal map generation. I don't know what generation of tech he is expecting but obviously it can't look like DOOM 1. I'm thinking like the Return to Castlewolfenstein, that sort of dark look to things. I mean I don't know if he wants me to try and get a spec and AO and cavity maps on everything. Realistically I have bitten off way too much than I can chew but throughout my college career I have always done that and succeeded, despite that you people repeatedly tell me I'm a complete moron.
Anyways I am gathering proper references and trying to plan out in my head what the scene will look like. I want to start blocking out things tonight after I move out of the dorms today.
Wish me luck and I will be posting more in the future with this thread.
I am going to look for some tutorial dvds now. Like Eat3d, DigitalTutors or something. If I F up my indep study, I cannot re-take it and really then I have to change my major which is like super major you know? lol
If at all possible think more along the lines of a small scene, a medieval weapons shop or something as you have more experience making weapons. One character(shopkeeper) at the absolute max. Modeling, uving, baking, texturing, lighting and setting up even a small scene like this will eat up a lot of time, especially if you've never finished a scene like this before. Make it a weapons shop right outside of a castle(not to stray to far from the original plan), so you can have a partial castle wall represented behind the shop, but not the massive amount of work(months) that it would take you to build and texture a full castle.
How does it need to be presented? Do you need to walk around it/demo it as if its a "level"? A fully working(even just so you can walk around in it) level can be a massive amount of work. If at all possible concentrate only on what needs to be visible for presentation, creating a "facade" instead of a level.
It sounds like neither you nor your teacher understand the amount of work involved in this, so I would really step back and re-evaluate it. Its better to bring up the fact that this is totally unrealistic to accomplish now, than 2 weeks before it is due.
The first thing you will need to do is make a FAST and basic blockout of the entire scene, all the buildings, props, characters etc need to blocked out with basic modeling(primitives etc). From there you will need to make a list of all assets, and write down a time estimate for each require asset. Be realistic with yourself when coming up with time estimates, pencil in MORE time than you think it reasonably would take to account for any problems that may occur, pencil in a TON of time for things you've never done before. To pull something like this off you will have to diligently stick to a plan.
Don't be stubborn and follow the tips this guys are giving you...my two cents
doing rest of castle block out now
plan on the castle, a wall around it that contains a few simple merchant shops and a few milling people.
maybe a a big tree or two
that's my goal at least
watching a bunch of DT videos in my free time basically trying to spend most of my free time learning this stuff till my brain falls asleep and then I start over again
Edit: really read the advice that EQ posted. It's great advice.
Yea, like Sukotto said, thats just not a good way to build it...unless its smaller scale...like a toy model or something.
Thats thinking more like a prop artist than an environment artist...you can't uniquely unwrap/bake/texture everything.
You would probably want to use modular pieces, maybe make high poly pieces for the bricks and whatnot...or just use tiled textures.
Honestly, making a solid looking wall piece and ignoring the rest of the castle while focusing on your shops might be the best thing you can do.
I fail to see this when the advise given was to start with something smaller, like a single shop and merchant instead of building the whole thing...be realistic, specially since you're learning as you go.
This is already going down hill by the way you are building the castle...
If it was game development you're talking a team.
concept artists, prop artists (doing small props), character artist, texture artist, level layout, level detail....
But going modular can save a ton of work. 1 character and 3 heads...1 hairpiece, 1 helmet, etc...
I'd to the same with the castle pieces. Make one tower, bake it or whatever you decide (tile textures, etc...)
But you said you were going to try and stay Authentic. You need to look at some real castle refs. While the tower might look pretty cool I don't think it looks like an authentic design.
I'm not using UDK, never used it, don't even really know what it is. afaik it's a render engine for scenes. Somehow related to Unreal game series...
Working on blocking out the scene, maya is crashing on me. Figuring on using my tower from now on. Not sure if it will help a ton but I hope so.
one of my ref images as for where I want things placed. Obviously less buildings on the hillside and no huge tower
I think you should go to the wiki and have a decent read through all the environment modelling info before you spend any more time on this castle.
seems complicated
and watching this video
think it'll be a good reference guide?
[ame="
http://wiki.polycount.com/CategoryEnvironmentModularity?highlight=%28\bCategoryEnvironment\b%29
Looking at your ref you could get away with having your spire mostly unqiue (still using smaller modular chunks) and the dome. The other buildings however is going to be wayy too much to do for a unique environment and you will 100% need to plan it out ahead if you are ever going to finish it. Also you will need to sculpt a terrain.
Is the actual scene going to be composed like it is in the ref? if you are going for exactly the same camera angle you could easily cut the tri count of what you are already doing and put it on very small tex sheets if it really is going to be seen from a distance like this, but I get the feeling it is not?
Either way good luck but I would strongly consider what other posters have said and go for something much smaller and modular based. Dont run before you can walk.
For an ai you can use one body, but have pieces of armor to make many ai from it.
For a building you can make a wall section. Then make a copy and add a window. Since the first wall is already uv mapped and textured you save time. Then make another and add a door.... Then you can slap them together in any combination.
That's very much how most game editing is done.
UDK is unreal Engine Dev kit. It is a very indepth engine (I haven't used that one yet myself). But it is basically a simple 3d editor (you make the walls of the castle and ground, the main play area, then you slap on models to decorate. door frames, window frames, stairs...) and then it will have scripting tools etc for making a playable game.
Not great for making models, but really good for base terrain and game stuff. Models make it look pretty.
I have also made other pieces like a curved bridge, other walls, etc
Focus on baking the towers and bridge if I get time for the bridge
Set up my tower in my room and install Maya, UDK and Marmoset and get serious
and the low fps in maya can be rectified using a modular approach. One HP piece, One LP piece, once the bake is done discard the HP and assemble the LP pieces.
I was thinking of UDK to set up my model and assign textures, normals etc
but do rendering and final product display in marmo
unless that makes no sense?
I'm more concerned this is out of your scope. Through this thread there seems to be a lack of baking knowledge of basic props, even smoothing groups, yet tackling a whole castle with no knowledge of modularity, realtime art, UDK, etc. You need about a year's worth of practice before such a huge scene is accessible to you.
http://www.polycount.com/forum/showpost.php?p=1593543&postcount=6098
Do my lows need to be that way to work in UDK or no?
i tried to install udk but it says it needs an internet connection???
doing good so far i think
beyond i still can't properly unwrap and object in maya...
got about 2 gigs of udk tut videos
pretty helpful
a few questions later i will ask but i think i made good progress today
pics of my bridge in a bit. imported to udk and all that but shows up screwy with the white and purple texture on it.
i saw a bunch of the material editor videos but i never saw him select the object and assign the material to it in the videos, he was always applying the materials to a primitive shape like a cube or sphere. i'd love to know after i import a mesh how i assign a material to it. ??
I have made up a simple texture for my bridge but dunno how it will fit on the object in UDK. Give it a go
also have to make a dirty normal from my texture, I have that tutorial around here somewhere
post up pics of Maya bridge later on today
also registered on UDK epic forums. take a while for them to OK my account but then I'll be getting it done right
EDIT Bridge Pics added
it took me about 3 days thinking about it to realize what I need to do to make this project come together
instead of questioning WHY this and WHY that I need to just work within the conventions already established
I have successfully unwrapped, normal mapped and applied diffuses to several wall pieces in maya and hope to
redo the meshes in maya in a modular fashion and then assemble them in UDK
things are coming together quickly now
I plan on duplicating these inside UDK and use them as walls
I am re-building the castle now in a modular way
hope to show more soon
But I dont think a medieval castle will have colored bricks like the one you're working on now and it's a little uniformed.
IMO, i think it should look abit like this
I think he meant it doesn't tile *properly*. You can tile planes side-to-side infinitely but it is still going to have seams if the texture doesn't tile properly at the edges.
In which case I'd just got it sloppy presentation as that none can tell that it does in fact tile. Which imo would be the point of showing it.
Also creepy avatar man....
edit.....beaten to the punch not once, but twice (that will teach me for looking up how to spell seamlessly
No it doesn't. Even if i could be bothered to line them up properly..
You can clearly see that some of the bricks have a very different size from all the others, and the colors don't flow nicely across the seams either. It doesn't tile.
edit: hah, Skillmister, you thought what I thought while I thought up those thoughts.
That's the least of my concerns to be honest. I know the normals, diffuse and specs work and look good in UDK so that's all I'm worried about. The modeling is not bothersome, mostly working in UDK is.
I still need to watch an a55load more of tutorials. Heck, I don't even know how to duplicate objects.
Also it'd be better to redo it entirely in the right style. Those bricks are pretty modern for a castle.
I have cobblestones with generated normals, spec, etc
Thinking about HOW to make the modular castle. After I get done thinking I'll re-create the castle.