How did you manage to do so using units that will make sense in your game engine? I guess I'm missing something, it's late and can't compute.
same question :P
Think I spend half my time working out scales and adjusting them slightly where needed but ofc this could be me being a total dumb ass with UE4 but 1uu = 1cm is alot better than UE3's scaling.
I'm only making this scene. I am not concerned with making perfectly modular pieces for an entire game level to be assembled later. Therefore I can model everything at an arbitrary scale and as long as everything is at a good scale relative to all the other parts of this scene I can avoid thinking about scale until the very end.
Once I have everything built out and ready to export, I can bring in a scale mesh of a human (or in my case, it will be the Unreal engine 4 character collision mesh) and scale the entire scene to match it. For example, the width and height of the doorway are likely to be close to what they are in real life, as is the bunk bed and chair.
Also, since the concept artist most likely started from a 3d block out, he has already done all the work in figuring out the relative scales of all the scene's parts. When I transferred his concept back into 3d it still holds up and looks correct. A lot of the work in figuring out the environment scale often falls on the concept artist, and rightfully so.
Hello all, this will be my first entrance at this kind of contest here at polycount, been following the site/forum at a long time but never had the time to enter any of the Challenges. Will try to post some WIP today.
Note: THank you ScottMichaelH for the references link
Computron.. that match perspective tool is fantastic. I'm a Maya user and am going to have to familiarize myself with max for that feature alone.
I'd be interested to know the focal length it gave you if you feel like sharing? Even though its a little too late for me I just roughly got it right and started modeling.
Blockout, considering modular approach
Quick atlas, ready to tear about and populate the room. Bit of a cheap approach but theres so many unique assets I wanted to cut corners where I could.
I don't have udk on this machine so this will have to do for now, ignore the material settings..
Its supposed to line up to the side you see on the 3/4 view. You'll have to eyeball it, it doesnt perfectly line up with the concept. I think its there just to illustrate a little better how the interior fits in the exterior.
Ok, I can see how the exterior profile fits, but the interior one still doesn't make sense. Maybe it be more clear once I get some more modelling done.
Dude, get manic time. It will record time spent in each program and even fetch the name of the file your working on inside the program. It can export to excel if you need.
Thanks a lot I'll be testing it.
I've jumped from the blockout to the biggest shapes, starting with the bed. Tried to do some quick subd modeling, I'll move on to the lockers next.
I've jumped from the blockout to the biggest shapes, starting with the bed. Tried to do some quick subd modeling
What methods are you using to make the bed sheets on the mattress? Did you simply move vertices around as anything else? I was thinking I would sculpt mine out or even drop a cloth simulation onto to it.
What methods are you using to make the bed sheets on the mattress? Did you simply move vertices around as anything else? I was thinking I would sculpt mine out or even drop a cloth simulation onto to it.
I made a 1-sided folded plane, tesselated it, added some wavyness with noise modifier, ffd 3x3x3 modifier to just tweak things around and dropped it on the mattress using cloth, with the Terrycloth preset.
After the simulation was done, I collapsed it, moved some verts around to fix errors and added a shell modifier to give it thickness. The two beds are the same, just mirrored sideways.
I did pretty much the same thing for the pillow, but I've added internal pressure on the cloth modifier it so it puffs up instead of collapsing into itself. Sure I could sculpt it, but would take much longer.
can i hang with all you cool kids in the blockout club?
I think a good place to start is to get all these smaller props all onto a single texture sheet. I'm going to try a highpoly sculpt to 3dcoat type of pipeline and see how that works. I'll probably take this into marmoset for final rendering. Wish me luck!
noob question, when creating a scene like this when there's a bunch of objects ... where do you guys usually start ?
Personally I think it's the best to start from the biggest shapes and work your way to the smallest ones. That way you prevent getting caught up doing silly detailing and reworking on things that don't affect the final scene as much, and makes it easier to spot proportion problems and whatnot.
A few years back when I was starting I'd usually tackle first what was more satisfying to work on, but that's usually not what's more efficient. Besides the planning issues this might cause, after I was done with the fun stuff there was a lot of not-so-fun work piled up, which would make my motivation dip a bit.
For example, I found it kinda boring to do this blockout, but I'm glad that I did since it already saved me quite some time trying to figure some things out. It's an exercise that really helps wrap your head around the scene.
Night #3. My biggest question at this point is about how much to model on those boxes on the shelf. I suppose you could model everything or nothing and save it for the texture. At such a small level, and if painted over, you probably wouldn't notice the normal much so I'm thinking of painting it. Then again some of those knobs may need to be protruding and an AO bake could give you a good starting point to the paint. This is where I struggle. I've modeled for years but when it comes to game design conventions I'm still dealing with gaps in my understanding.
Personally I think it's the best to start from the biggest shapes and work your way to the smallest ones. That way you prevent getting caught up doing silly detailing and reworking on things that don't affect the final scene as much, and makes it easier to spot proportion problems and whatnot.
A few years back when I was starting I'd usually tackle first what was more satisfying to work on, but that's usually not what's more efficient. Besides the planning issues this might cause, after I was done with the fun stuff there was a lot of not-so-fun work piled up, which would make my motivation dip a bit.
For example, I found it kinda boring to do this blockout, but I'm glad that I did since it already saved me quite some time trying to figure some things out. It's an exercise that really helps wrap your head around the scene.
thanks for the advice, before this challenge ... I challenge myself to learn a 3-month single room scene with planning thinking that it would be good to finish it but I ended to got bored in the mid process because just like yours I took the fun stuff first and less not-so-fun, I guess I got the wrong plans but I learned how to plan haha ...
Night #3. My biggest question at this point is about how much to model on those boxes on the shelf. I suppose you could model everything or nothing and save it for the texture. At such a small level, and if painted over, you probably wouldn't notice the normal much so I'm thinking of painting it. Then again some of those knobs may need to be protruding and an AO bake could give you a good starting point to the paint. This is where I struggle. I've modeled for years but when it comes to game design conventions I'm still dealing with gaps in my understanding.
Play some current games and notice how they deal with it! Try a first person game closed-space game and it should give you a good idea.
If you check TF2 you'll see that they have a bunch of knobs modeled in their computers, and that's a game from 2007. I usually just do some very quick shapes and reuse them around. You can get away with a handful of knobs and buttons and such, just put them into different layouts!
Hello everyone, I'm jumping in on the challenge with the fusion coil wooo
Edit:
I was looking for some size ref for the coil since I didn't know anything about it and found this. It's more or less the size of a barrel, at the beginning I was thinking it was big enough to fit one guy inside, hope this helps someone.
I've begun working on the prop, the concept is somewhat annoying as i only have one image where i can see a surface directly on. So i started out by fixing that in photoshop, and made this:
Its not perfect, but it works really well. And from that i continued to model this:
its probably as far as i'll get today, i'm really happy with the results so far, and tomorrow will consist of detailing the outer shell after i've added the top.
The inner parts are going to be tricky though, and does anyone have a scale reference for this?
Nice way of planning. I haven't started yet but i was doing some similar planning on paper, if you draw a red line down the middle of the left and right thirds of your image, you have 1/4 of the model. Much easier to think about it this way when getting started i'm sure.
Now I just need some free time to clear up.
Keep going.
EDIT: Bet you you could make a tileable out of that, smack it on to a 1/4 octagon, Auto UV, bam. Done
Nice way of planning. I haven't started yet but i was doing some similar planning on paper, if you draw a red line down the middle of the left and right thirds of your image, you have 1/4 of the model. Much easier to think about it this way when getting started i'm sure.
Now I just need some free time to clear up.
Keep going.
EDIT: Bet you you could make a tileable out of that, smack it on to a 1/4 octagon, Auto UV, bam. Done
I might be able to do some of this this weekend possibly. POSSIBLY.
Play some current games and notice how they deal with it! Try a first person game closed-space game and it should give you a good idea.
If you check TF2 you'll see that they have a bunch of knobs modeled in their computers, and that's a game from 2007. I usually just do some very quick shapes and reuse them around. You can get away with a handful of knobs and buttons and such, just put them into different layouts!
So they made a second Tournament Fighters?! That took a while :poly129: (sarcasm) Yeah, I don't know what is TF2. This comment did convince me to go on Youtube and get seriously acquainted with what's current, which has not been me for some years now. I'm kinda old school.
Night #4. I decided to sculpt the bed. I realized I would want to sculpt on top of a cloth sim anyway, and this will give me a simpler normal bake as well. Hockey distracted me tonight, will try to get more done tomorrow. :poly009:
I've begun working on the prop, the concept is somewhat annoying as i only have one image where i can see a surface directly on. So i started out by fixing that in photoshop, and made this:
Its not perfect, but it works really well. And from that i continued to model this:
its probably as far as i'll get today, i'm really happy with the results so far, and tomorrow will consist of detailing the outer shell after i've added the top.
The inner parts are going to be tricky though, and does anyone have a scale reference for this?
Nice planing mate, about the highpoly...make those edges more soft...right now they are sooo tight and you will not see them when bake down into normal map...
So they made a second Tournament Fighters?! That took a while :poly129: (sarcasm) Yeah, I don't know what is TF2. This comment did convince me to go on Youtube and get seriously acquainted with what's current, which has not been me for some years now. I'm kinda old school.
Team Fortress 2, sorry! It's free to play on steam, i'd suggest you to check it out!
I've begun working on the prop, the concept is somewhat annoying as i only have one image where i can see a surface directly on. So i started out by fixing that in photoshop, and made this:
Its not perfect, but it works really well. And from that i continued to model this:
its probably as far as i'll get today, i'm really happy with the results so far, and tomorrow will consist of detailing the outer shell after i've added the top.
The inner parts are going to be tricky though, and does anyone have a scale reference for this?
A few things you guys need to do is add wayyyy more depth to the models, instead of making all the details on a flat plane, play with it a bit and don't be afraid to experiment. As others have said loosen up the edges a tad too:)
as for my progress, i've broken up the assets into logical texture groups. Then I've laid out a rough expectation of the uv's of the high poly shapes. I've just got to finish up the layout for the pipe and utility bits, then its off to start the normal maps.
as for my progress, i've broken up the assets into logical texture groups. Then I've laid out a rough expectation of the uv's of the high poly shapes. I've just got to finish up the layout for the pipe and utility bits, then its off to start the normal maps.
how do you plan texturing / unwrapping later on?
and how those different colors differ from each other?
Love everyone's work so far I have to make time to start mine, so no pics yet :P
Quick question about modeling process:
Do you guys go for the high poly model first?
In school I was taught to go for a "mid-poly", a "Mid-Poly" model as I was told was one you reduce to make low poly and add detail to make High poly. So instead of making a high poly and then reducing it, it adds an extra step, but at times it seems inefficient adding an extra step like that.
I'm already a dreadfully slow modeler and I'm looking for ways to speed things up a bit.
Thanks for the compliments guys
I'm trying to burn through this as quick as possible, don't want to waste time on those small props. Here are some more:
I don't know if what I'm doing counts as mid poly, but it's a bunch of blocky shapes with support edges and a turbosmooth modifier on top. Sometimes I take those as a starting point for the low poly, sometimes I use the smooth version for some shapes. In this case I just take the high poly and delete a bunch of loops using ctrl+backspace (I have a mouse button assigned to that), or I detach some pieces, or make them from scratch, whatever is quicker. Just start doing it and you'll adjust yourself with time!
@skyline5gtr: I'm taking them into Marmoset. Made a quickie scene, two directional lights, nothing fancy!
@Voxel Take what I say with a grain of salt & pepper, but I generally just go straight to the high poly, most of the time when I'm creating my high poly I will start out with a 'mid poly' but that step for me is more about blocking out the geometry and getting the shape and feel of the object rather than giving myself a starting point for a low poly. I feel like for hard surface work its fairly easy to produce a workable low poly object from scratch or simply reduce the number of polys on the high poly.
Replies
same question :P
Think I spend half my time working out scales and adjusting them slightly where needed but ofc this could be me being a total dumb ass with UE4 but 1uu = 1cm is alot better than UE3's scaling.
Once I have everything built out and ready to export, I can bring in a scale mesh of a human (or in my case, it will be the Unreal engine 4 character collision mesh) and scale the entire scene to match it. For example, the width and height of the doorway are likely to be close to what they are in real life, as is the bunk bed and chair.
Also, since the concept artist most likely started from a 3d block out, he has already done all the work in figuring out the relative scales of all the scene's parts. When I transferred his concept back into 3d it still holds up and looks correct. A lot of the work in figuring out the environment scale often falls on the concept artist, and rightfully so.
Here's the first couple nights worth. The pipes and wires are what concerned me the most so I'm starting with them.
Also I gathered several references in a Pinterest page if anyone is interested:
http://www.pinterest.com/scotthoneycutt9/interior-references/
Note: THank you ScottMichaelH for the references link
I'd be interested to know the focal length it gave you if you feel like sharing? Even though its a little too late for me I just roughly got it right and started modeling.
Blockout, considering modular approach
Quick atlas, ready to tear about and populate the room. Bit of a cheap approach but theres so many unique assets I wanted to cut corners where I could.
I don't have udk on this machine so this will have to do for now, ignore the material settings..
Thanks a lot
I've jumped from the blockout to the biggest shapes, starting with the bed. Tried to do some quick subd modeling, I'll move on to the lockers next.
What methods are you using to make the bed sheets on the mattress? Did you simply move vertices around as anything else? I was thinking I would sculpt mine out or even drop a cloth simulation onto to it.
I made a 1-sided folded plane, tesselated it, added some wavyness with noise modifier, ffd 3x3x3 modifier to just tweak things around and dropped it on the mattress using cloth, with the Terrycloth preset.
After the simulation was done, I collapsed it, moved some verts around to fix errors and added a shell modifier to give it thickness. The two beds are the same, just mirrored sideways.
I did pretty much the same thing for the pillow, but I've added internal pressure on the cloth modifier it so it puffs up instead of collapsing into itself. Sure I could sculpt it, but would take much longer.
(don't mind the missing edge on the middle door)
I think a good place to start is to get all these smaller props all onto a single texture sheet. I'm going to try a highpoly sculpt to 3dcoat type of pipeline and see how that works. I'll probably take this into marmoset for final rendering. Wish me luck!
Personally I think it's the best to start from the biggest shapes and work your way to the smallest ones. That way you prevent getting caught up doing silly detailing and reworking on things that don't affect the final scene as much, and makes it easier to spot proportion problems and whatnot.
A few years back when I was starting I'd usually tackle first what was more satisfying to work on, but that's usually not what's more efficient. Besides the planning issues this might cause, after I was done with the fun stuff there was a lot of not-so-fun work piled up, which would make my motivation dip a bit.
For example, I found it kinda boring to do this blockout, but I'm glad that I did since it already saved me quite some time trying to figure some things out. It's an exercise that really helps wrap your head around the scene.
Night #3. My biggest question at this point is about how much to model on those boxes on the shelf. I suppose you could model everything or nothing and save it for the texture. At such a small level, and if painted over, you probably wouldn't notice the normal much so I'm thinking of painting it. Then again some of those knobs may need to be protruding and an AO bake could give you a good starting point to the paint. This is where I struggle. I've modeled for years but when it comes to game design conventions I'm still dealing with gaps in my understanding.
thanks for the advice, before this challenge ... I challenge myself to learn a 3-month single room scene with planning thinking that it would be good to finish it but I ended to got bored in the mid process because just like yours I took the fun stuff first and less not-so-fun, I guess I got the wrong plans but I learned how to plan haha ...
Play some current games and notice how they deal with it! Try a first person game closed-space game and it should give you a good idea.
If you check TF2 you'll see that they have a bunch of knobs modeled in their computers, and that's a game from 2007. I usually just do some very quick shapes and reuse them around. You can get away with a handful of knobs and buttons and such, just put them into different layouts!
Edit:
I was looking for some size ref for the coil since I didn't know anything about it and found this. It's more or less the size of a barrel, at the beginning I was thinking it was big enough to fit one guy inside, hope this helps someone.
Made the interior faces of the lockers, moving on!
Its not perfect, but it works really well. And from that i continued to model this:
its probably as far as i'll get today, i'm really happy with the results so far, and tomorrow will consist of detailing the outer shell after i've added the top.
The inner parts are going to be tricky though, and does anyone have a scale reference for this?
Nice way of planning. I haven't started yet but i was doing some similar planning on paper, if you draw a red line down the middle of the left and right thirds of your image, you have 1/4 of the model. Much easier to think about it this way when getting started i'm sure.
Now I just need some free time to clear up.
Keep going.
EDIT: Bet you you could make a tileable out of that, smack it on to a 1/4 octagon, Auto UV, bam. Done
I might be able to do some of this this weekend possibly. POSSIBLY.
DO IT. :poly142:
For the big objects used the floor tiles as reference for the scale.
So they made a second Tournament Fighters?! That took a while :poly129: (sarcasm) Yeah, I don't know what is TF2. This comment did convince me to go on Youtube and get seriously acquainted with what's current, which has not been me for some years now. I'm kinda old school.
Night #4. I decided to sculpt the bed. I realized I would want to sculpt on top of a cloth sim anyway, and this will give me a simpler normal bake as well. Hockey distracted me tonight, will try to get more done tomorrow. :poly009:
Good idea and good start.
Nice planing mate, about the highpoly...make those edges more soft...right now they are sooo tight and you will not see them when bake down into normal map...
Team Fortress 2, sorry! It's free to play on steam, i'd suggest you to check it out!
Done with the chair and moving on!
Pooosssssiibliddefinitely
Aw, haha thanks for the feedback didn't see that ... and I updated today, added wrinkles on the bedding and added another prop- lockers
A few things you guys need to do is add wayyyy more depth to the models, instead of making all the details on a flat plane, play with it a bit and don't be afraid to experiment. As others have said loosen up the edges a tad too:)
By the way, I found this for the fusion coil.
http://wiki.polycount.net/Model%20Presentation
i'm sure marmoset is a quick and easy option as well.
as for my progress, i've broken up the assets into logical texture groups. Then I've laid out a rough expectation of the uv's of the high poly shapes. I've just got to finish up the layout for the pipe and utility bits, then its off to start the normal maps.
how do you plan texturing / unwrapping later on?
and how those different colors differ from each other?
Quick question about modeling process:
Do you guys go for the high poly model first?
In school I was taught to go for a "mid-poly", a "Mid-Poly" model as I was told was one you reduce to make low poly and add detail to make High poly. So instead of making a high poly and then reducing it, it adds an extra step, but at times it seems inefficient adding an extra step like that.
I'm already a dreadfully slow modeler and I'm looking for ways to speed things up a bit.
I'm trying to burn through this as quick as possible, don't want to waste time on those small props. Here are some more:
I don't know if what I'm doing counts as mid poly, but it's a bunch of blocky shapes with support edges and a turbosmooth modifier on top. Sometimes I take those as a starting point for the low poly, sometimes I use the smooth version for some shapes. In this case I just take the high poly and delete a bunch of loops using ctrl+backspace (I have a mouse button assigned to that), or I detach some pieces, or make them from scratch, whatever is quicker. Just start doing it and you'll adjust yourself with time!
@skyline5gtr: I'm taking them into Marmoset. Made a quickie scene, two directional lights, nothing fancy!
@Voxel Take what I say with a grain of salt & pepper, but I generally just go straight to the high poly, most of the time when I'm creating my high poly I will start out with a 'mid poly' but that step for me is more about blocking out the geometry and getting the shape and feel of the object rather than giving myself a starting point for a low poly. I feel like for hard surface work its fairly easy to produce a workable low poly object from scratch or simply reduce the number of polys on the high poly.