I found these chairs on reddit, and thought it would be fun to post it here. I am curious to how people would approach making these chairs, just to see how many different ways you could do it.
Some more images in this album. http://imgur.com/a/HA9RH#0
Note, I don't need help making the shape, so this is more of a challenge/fun thing.
Will you just make a tutorial series on sub-d modeling and sell it on 3D Motive or Eat 3D, please? Just do it. I'll buy it, scout's honor...
Honestly there is no secret. Perna is miles ahead of my modeling skills, but I think I am good enough to where I can workout any crazy shape. Once you understand the basics of sub-d modeling just keep challenging yourself to get better. Pick projects with fun and challenging shapes to make, and don't keep playing it safe with simple boxy stuff. Do little random modeling studies on crazy shapes like that chair I posted, or crazy cast iron machine parts. Its amazing how much you can learn by finding objects you have no clue how to model and just work your way though it.
Do you think you could make a video tutorial of your workflow for making a complex shape like that one you showed Computron how to make?
That would be the coolest thing ever.
Honestly there is no secret. Perna is miles ahead of my modeling skills, but I think I am good enough to where I can workout any crazy shape. Once you understand the basics of sub-d modeling just keep challenging yourself to get better. Pick projects with fun and challenging shapes to make, and don't keep playing it safe with simple boxy stuff. Do little random modeling studies on crazy shapes like that chair I posted, or crazy cast iron machine parts. Its amazing how much you can learn by finding objects you have no clue how to model and just work your way though it.
I get what you're saying but there's clearly something there to be taught, otherwise everyone would be busting out awesome subd and this thread wouldn't exist.
I get what you're saying but there's clearly something there to be taught, otherwise everyone would be busting out awesome subd and this thread wouldn't exist.
Yeah and that thing is 'experience'. Subd is not a rocket science .
Did a quick and dirty model and got a decent result. I have some pinching because I didn't use enough geo on the outside cylinder, if I doubled up those outside edges I would probably be fine. I made it by making the shape flat first, patching in the array of holes, and then deforming it around a super dense sphere.
Making it work to be rounded like that wasn't to bad. I am still baffled at how you could get that angled shape the chair has in the reference, without having to spend a few hours punching all those holes by hand through the primitive shape.
Purple is the actual shape, and red being the fastest way to do it.
Doesn't work, because then the pattern warps wrong. Thats what I tried first.
It quite worked for me; at the second try, I managed to obtain a good mesh, pretty close to the original;
My workflow was:
1) Global shape;
2) Negative model of the holes (perpendicular to the chair surface);
3) Boolean subtraction of the holes;
4) Clean up of the geometry (this was the worst part of the work);
5) Shell and meshsmooth. [/QUOTE]
i gave it a quick try,
the outer shape itself isn't much of a problem, i made 2 mistakes here, i ignored the circular shape of the inner cuts and i totally ignored the fact that on the centerline my inset won't work due to beeing mirrored so there is a shading error, which can easily be removed by removing the centerline
i should invest another 5 minutes to build it properly but oh well...
the whole thing is symmatrical on 2 axes, so i only build a quarter of the shape
then i used ffd to bring it into the outer shape
then i used 2 symmetry modifers to get the overall object
on top of that an edit polymodifier to bend the outer corner into the proper form and add the trim
shell on top, look at the override toggles, we will use those in the next step
edit poly on top, select the outer materialID and use the outline option in edit polygons to make the inner part of the holes bigger than the outer part
double smooth on top, i could chamfer but why would i? it's faster this way and totally sufficient for this kind of asset, if you care to sculpt it's even better that way, as chamfered edges are a bitch to sculpt on
This is great everyone! Complicated Modeling - "Challenges" are great. Love to see the workflows you people come up with!
Maybe some of you guys remember the one or other complicated part they had to model once and want to let the people have a try
It quite worked for me; at the second try, I managed to obtain a good mesh, pretty close to the original;
My workflow was:
1) Global shape;
2) Negative model of the holes (perpendicular to the chair surface);
3) Boolean subtraction of the holes;
4) Clean up of the geometry (this was the worst part of the work);
5) Shell and meshsmooth.
This is the issue I am talking about, the gap between the holes in the center stays the same across. It seams that you can only get one shape or the other, unless you want to spend like 2 hours manually cleaning up every hole punch in the correct pattern. The only thing I can think of doing is maybe trying to guess the offset of the curve the pattern would need to bend it into shape.
Neox is super close, but the roundness of the bowl is a little off I think. I think Perna has the right idea though with just sticking to whats important the design.
This is also a perfect example of what kind of accuracy is necessary, and what is good enough. Trying to be perfectly accurate is going to be a time sink.
@perna I didn't mean to troll, I'm just saying on that step 6 image, the refence image has a much flatter seat and back part, it isn't really rounded like that. But in the end, it doesn't really matter unless you are going to have a character sitting in it.
This is the issue I am talking about, the gap between the holes in the center stays the same across. It seams that you can only get one shape or the other, unless you want to spend like 2 hours manually cleaning up every hole punch in the correct pattern. The only thing I can think of doing is maybe trying to guess the offset of the curve the pattern would need to bend it into shape.
Neox is super close, but the roundness of the bowl is a little off I think. I think Perna has the right idea though with just sticking to whats important the design.
That's not the hardest part; having to work only on a quarter of the mesh (being it specular) you can tweak it;
I made a copy of the segment to use as reference; when you move the vertices of the holes to match the desired shape, you can move them on the y axis too, to match the reference mesh surface and mantain the curvature;
Yea that is true. I think you nailed it. I work in Maya so its a little clunky to get the bend and symmetry working at the same time.
On a side note would you guys be cool with me posting a little challenge each week? I think this is a lot of fun seeing how everybody approaches the shape differently. I have all kinds of random pictures of stuff that would be fun to sub-d model. Should I start a new thread somewhere for it? Or do it in here.
Interesting couple days in here. Chair is tempting, But I'll catch the next one.
I think you should definitely do a weekly challenge! personally, I'd like to see a thread started. So you can update the OP with the current object and you don't have to worry about staying on top of this thread and worry about finding/missing the challenge.
Maybe a different thread in tech talk? "FAQ: How u model dem shapes? Weekly challenge edition"
Yea, thats more like it.
If it's weekly, won't it have to be a pretty complex piece?
Maybe the thread should have some rules too where every participant in that weeks challenge should post steps describing their process, so it doesn't get clogged up with the standard "i dont have enough geo, add more geo" stuff...
Everyone could throw out their ideas for the elements that grab their attention and quickly write up their attempt or post a timelapse.
If its truly a challenge, you could time how long it takes you and post that as your record, see who has the quickest method.
at the end of each challenge/week we can gather the best methods for modeling each complex element in the piece as like a wiki post or something.
If it's weekly, won't it have to be a pretty complex piece?
Maybe the thread should have some rules too where every participant in that weeks challenge should post steps describing their process, so it doesn't get clogged up with the standard "i dont have enough geo, add more geo" stuff...
Everyone could throw out their ideas for the elements that grab their attention and quickly write up their attempt or post a timelapse.
If its truly a challenge, you could time how long it takes you and post that as your record, see who has the quickest method.
at the end of each challenge/week we can gather the best methods for modeling each complex element in the piece as like a wiki post or something.
IDK, just some ideas...
I don't think it needs to be a competition, just a fun way to challenge yourself to get better, and hopefully learn from others. Its fun talking about different ways to approach a shape with people, also to see what workflows people come up with. Even if people cannot model the shape that gets posted each week, they could still learn by seeing what other people do.
Plus is more about modeling a specific shape, not en entire complete model, so it shouldn't take anybody a ton of time to do each challenge. Which is a good thing, because then hopefully a lot of people will join in.
Guys, you know how this goes. If you try to plan it too much, make it too ambitious, or too complex, it will never happen. Just post a challenge with a small but technically difficult shape and see whether anyone is interested.
From experience I know there's going to be more people who are interested in weighing in on rules and regulations than there are who are actually going to bother modeling anything.
Haha, truth.
Get something simple going and see if anyone actually participates. Then grow it from there like Alex did with the monthly noob challenges...
I will post it every Sunday or monday, each week in P&P. I will update back here when I post the thread. If anybody has any pictures of awesome shapes we could put up then PM them to me please.
I will post it every Sunday or monday, each week in P&P. I will update back here when I post the thread. If anybody has any pictures of awesome shapes we could put up then PM them to me please.
I have wanted to do this and was thinking about adding it into a challenge too. I am really happy someone will. I would love to participate in this so much!
We even had a subforum called Workshop or something. And it only lasted a couple of months. People stop going there.
Best way is how we have it here. Lets keep going in this same thread, posting pictures and your attempts and not focus on creating more threads and more clutter. That way we will fill this thread with lots of info and have something to fallback into.
Well, I'm posting way too much lately, so why not keep it up. Made a couple mistakes here but the result is ok:
1-lowpoly cage ensures smooth curves
2-adjust subd and cylinder segments to line up as well as possible
3-boolean
4-join verts and space out some of them by using a script that aligns verts on a loop at equal distances. I believe there's a few scripts out there like this, might be in graphite tools too. Can do it manually by moving verts a bit with "constrain to edge" turned on
5- symmetry modifiers make the whole from the quarter, shell modifier gives thickness and meshSmooth rounds it off
Yeah, That's a Terrible ideas guy. Geez. Learn how to PC.
Seriously though, if the thread fails, it fails...What is one more thread lost in the depths?
I can see logic to keeping shape modeling info in one thread, For the sake of it being easy to find, But To be honest, It's not exactly a peice of cake to find the info in this thread. Searching two threads wouldn't be a huge stretch.
If the vets say leave the idea behind, perhaps you should yeild the warning. But who knows? maybe his thread is the one that pans out and becomes a regular thing. How many "how you model dem shapes" threads existed before this one was a sucess?
My point here is: I don't see the harm in *trying*, Even if 10000 threads of the same caliber failed before. Might as well go for 10001.
Yeah I'd rather see a 2nd thread here in tech talk, this thread is already huge and takes a while to go through. It's more related to this subsection than pimping and preview.
I am cool with it either way. I think having a dedicated thread makes it easier for more people to contribute. I can see how it could fail if we don't get enough people though. I have no problem with keeping the thread updated because I just need to keep the original post updated, and post a few pictures each week. The art jam thread works fine because the person that runs it does a good job.
The downside to having it in here is that is might wash out people's posts who actually need help with a shape they are modeling. I think it's kind of a catch 22.
So wait, what? You guys want to make another thread somewhere else where people model the same shape as a sub-d challenge?
So the same point as this thread?
And then, having two threads will somehow make things easier to find?
If you guys want to try and do a weekly sub-d challenge that is cool, just keep it in this thread. We've tried the semi-regular sub-d challenge thing with the workshop forum and it really did not work.
I'm not going to stop anyone if you guys feel like you need to make a new thread for this, but it just seems like a really bad idea to fragment the information.
From experience I know there's going to be more people who are interested in weighing in on rules and regulations than there are who are actually going to bother modeling anything.
Hello, i have 4 times trying to model this thing and i never have the right shape, anyone can help me with this?
I dont care too much if is the complete model but i only want to see how you guys will make the triangled hole and screw hole.
@The new thread idea: I wouldn't say the point of the thread is to get help on creating difficult shapes (Which I'm sure there would be plenty of) but rather to have fresh meat to chew on, On a regular basis. With the subject mater targeted at honing ones modeling skills. In said thread you wouldn't see the same Washer, Bolt, and "bumpy edged" cylinder every few weeks.
You would see new and interesting brain teasers to work out. A lot of shapes in this thread aren't so much hard, And challenging, as much as they are tedious and complicated. (if there's a difference lol). But maybe that's just me.
I think you should do it man If it sinks like a rock, So be it. You'll have wasted 10 minutes writing the OP. If not, it will be a fun little thread to follow.
@The new thread idea: I wouldn't say the point of the thread is to get help on creating difficult shapes (Which I'm sure there would be plenty of) but rather to have fresh meat to chew on, On a regular basis. With the subject mater targeted at honing ones modeling skills. In said thread you wouldn't see the same Washer, Bolt, and "bumpy edged" cylinder every few weeks.
You would see new and interesting brain teasers to work out. A lot of shapes in this thread aren't so much hard, And challenging, as much as they are tedious and complicated. (if there's a difference lol). But maybe that's just me.
I think you should do it man If it sinks like a rock, So be it. You'll have wasted 10 minutes writing the OP. If not, it will be a fun little thread to follow.
MrOneTwo, I don't understand what you mean by that statement. To me, it seems that you are saying that subd modeling is NOT a highly technical job, as evidenced by the continued discussion of tools, workflow, scripting, et al, as well as the fact that even with extensive experience, anyone who hasn't put time into studying theory will fall short. Since that can't possibly be what you mean, a clarification might be in order.
I think that sub-d modelling is technical job since you really have to understand the proccess. I do belive it's a crucial skill in this cg. I just think that it's knowledge + experience. Problem is that the knowledge is here. I feel like you have written posts about this thousand times. It's not like suddenly you started writing something new and ground breaking. It's still awesome stuff and something everyone should read. I just find it funny that there is this cycle when people look at subd modelling and become amazed again :P
Experience on the other hand is something people have to gather (ground breaking statement!). I'm not that experienced with crazy shapes in subd but every time I stumbled upon something weird I tried. Sometimes it takes some time but eventually I will get it.
Just as you previously wrote the importatnt part is knowing your software. I used 3ds max (discovering FDDs... good times!) now I'm using Modo. Those softwares have so many cool tools which people won't even try.
I don't want to make it some long read because I don't think I have anything new to add. I get this feeling that people are awaiting for some AMAZING PERNA modelling trick that will change the world while they post mostly the same problem here without even browsing through this thread.
I hope that I expressed what I think properly (english isn't my language). I really respect proper sub-d modelling as a skill. You can't cheat though. There is no shortcut. Learn it and try (mostly by yourself).
Yeah i agree this thread goes a bit round in circle. I'm pretty sure if you've read the 3279 posts of this thread, you have a good comprehension of the theory process. Now sure, the more examples, the better. Then experience is required to apply theory or adapt theory in an efficient way. Same goes for every discipline.
Now what i think we lack a bit is good tools and fast process.
Tried one... instead of breakfast... definately not perfect and the geometry is messy. Could have use less geo and still got good result (and smoother shape transition which would be closer to the ref).
Next time I will do it with a breakdown. Modo 701 about 30 minutes.
You can't cheat though. There is no shortcut. Learn it and try (mostly by yourself).
See, this is where I think you're very wrong. Learning from someone who already has the skill that you're trying to acquire IS a shortcut. Whenever perna or EQ or some other talented subd modeler drops a mini-tutorial in this thread, that's knowledge I gain that I didn't have before. All of that adds up, over time.
Yes, you should absolutely work stuff out on your own and learn as much as you can via experience. But advocating slogging it out in isolation is silly, IMO. Learning from your peers is an intelligent thing to do.
Replies
Some more images in this album. http://imgur.com/a/HA9RH#0
Note, I don't need help making the shape, so this is more of a challenge/fun thing.
Will you just make a tutorial series on sub-d modeling and sell it on 3D Motive or Eat 3D, please? Just do it.
Honestly there is no secret. Perna is miles ahead of my modeling skills, but I think I am good enough to where I can workout any crazy shape. Once you understand the basics of sub-d modeling just keep challenging yourself to get better. Pick projects with fun and challenging shapes to make, and don't keep playing it safe with simple boxy stuff. Do little random modeling studies on crazy shapes like that chair I posted, or crazy cast iron machine parts. Its amazing how much you can learn by finding objects you have no clue how to model and just work your way though it.
Lol I was planning it making it tomorrow after class
Hey Perna, I've been wondering for some time if you ever made that pro booleans tutorial you were talking about in a thread here: http://www.polycount.com/forum/showthread.php?t=106272&page=3&highlight=pro+booleans
Do you think you could make a video tutorial of your workflow for making a complex shape like that one you showed Computron how to make?
That would be the coolest thing ever.
Yeah and that thing is 'experience'. Subd is not a rocket science
Making it work to be rounded like that wasn't to bad. I am still baffled at how you could get that angled shape the chair has in the reference, without having to spend a few hours punching all those holes by hand through the primitive shape.
Purple is the actual shape, and red being the fastest way to do it.
That shape is a bit odd, I'd try to make a model that matches that shape without the holes first.
Doesn't work, because then the pattern warps wrong. Thats what I tried first.
Look at Perna's latest post. Full of great information.
It quite worked for me; at the second try, I managed to obtain a good mesh, pretty close to the original;
My workflow was:
1) Global shape;
2) Negative model of the holes (perpendicular to the chair surface);
3) Boolean subtraction of the holes;
4) Clean up of the geometry (this was the worst part of the work);
5) Shell and meshsmooth.
the outer shape itself isn't much of a problem, i made 2 mistakes here, i ignored the circular shape of the inner cuts and i totally ignored the fact that on the centerline my inset won't work due to beeing mirrored so there is a shading error, which can easily be removed by removing the centerline
i should invest another 5 minutes to build it properly but oh well...
http://steffenunger.com/advices/minitut/bowlseat.max
the whole thing is symmatrical on 2 axes, so i only build a quarter of the shape
then i used ffd to bring it into the outer shape
then i used 2 symmetry modifers to get the overall object
on top of that an edit polymodifier to bend the outer corner into the proper form and add the trim
shell on top, look at the override toggles, we will use those in the next step
edit poly on top, select the outer materialID and use the outline option in edit polygons to make the inner part of the holes bigger than the outer part
double smooth on top, i could chamfer but why would i? it's faster this way and totally sufficient for this kind of asset, if you care to sculpt it's even better that way, as chamfered edges are a bitch to sculpt on
Maybe some of you guys remember the one or other complicated part they had to model once and want to let the people have a try
This is the issue I am talking about, the gap between the holes in the center stays the same across. It seams that you can only get one shape or the other, unless you want to spend like 2 hours manually cleaning up every hole punch in the correct pattern. The only thing I can think of doing is maybe trying to guess the offset of the curve the pattern would need to bend it into shape.
Neox is super close, but the roundness of the bowl is a little off I think. I think Perna has the right idea though with just sticking to whats important the design.
maybe you should stop showing angles that would help after people interpreted the shape of the one angle you showed before
lol, the first post has the whole album of pictures from all different angles of the chair
@perna I didn't mean to troll, I'm just saying on that step 6 image, the refence image has a much flatter seat and back part, it isn't really rounded like that. But in the end, it doesn't really matter unless you are going to have a character sitting in it.
I made a copy of the segment to use as reference; when you move the vertices of the holes to match the desired shape, you can move them on the y axis too, to match the reference mesh surface and mantain the curvature;
On a side note would you guys be cool with me posting a little challenge each week? I think this is a lot of fun seeing how everybody approaches the shape differently. I have all kinds of random pictures of stuff that would be fun to sub-d model. Should I start a new thread somewhere for it? Or do it in here.
I think you should definitely do a weekly challenge! personally, I'd like to see a thread started. So you can update the OP with the current object and you don't have to worry about staying on top of this thread and worry about finding/missing the challenge.
Yea, thats more like it.
If it's weekly, won't it have to be a pretty complex piece?
Maybe the thread should have some rules too where every participant in that weeks challenge should post steps describing their process, so it doesn't get clogged up with the standard "i dont have enough geo, add more geo" stuff...
Everyone could throw out their ideas for the elements that grab their attention and quickly write up their attempt or post a timelapse.
If its truly a challenge, you could time how long it takes you and post that as your record, see who has the quickest method.
at the end of each challenge/week we can gather the best methods for modeling each complex element in the piece as like a wiki post or something.
IDK, just some ideas...
I don't think it needs to be a competition, just a fun way to challenge yourself to get better, and hopefully learn from others. Its fun talking about different ways to approach a shape with people, also to see what workflows people come up with. Even if people cannot model the shape that gets posted each week, they could still learn by seeing what other people do.
Plus is more about modeling a specific shape, not en entire complete model, so it shouldn't take anybody a ton of time to do each challenge. Which is a good thing, because then hopefully a lot of people will join in.
Doesn't make it less true.
Get something simple going and see if anyone actually participates. Then grow it from there like Alex did with the monthly noob challenges...
I have wanted to do this and was thinking about adding it into a challenge too. I am really happy someone will. I would love to participate in this so much!
We had something like this before.
We even had a subforum called Workshop or something. And it only lasted a couple of months. People stop going there.
Best way is how we have it here. Lets keep going in this same thread, posting pictures and your attempts and not focus on creating more threads and more clutter. That way we will fill this thread with lots of info and have something to fallback into.
Seriously though, if the thread fails, it fails...What is one more thread lost in the depths?
I can see logic to keeping shape modeling info in one thread, For the sake of it being easy to find, But To be honest, It's not exactly a peice of cake to find the info in this thread. Searching two threads wouldn't be a huge stretch.
If the vets say leave the idea behind, perhaps you should yeild the warning. But who knows? maybe his thread is the one that pans out and becomes a regular thing. How many "how you model dem shapes" threads existed before this one was a sucess?
My point here is: I don't see the harm in *trying*, Even if 10000 threads of the same caliber failed before. Might as well go for 10001.
The downside to having it in here is that is might wash out people's posts who actually need help with a shape they are modeling. I think it's kind of a catch 22.
So the same point as this thread?
And then, having two threads will somehow make things easier to find?
If you guys want to try and do a weekly sub-d challenge that is cool, just keep it in this thread. We've tried the semi-regular sub-d challenge thing with the workshop forum and it really did not work.
I'm not going to stop anyone if you guys feel like you need to make a new thread for this, but it just seems like a really bad idea to fragment the information.
lololol yes
I dont care too much if is the complete model but i only want to see how you guys will make the triangled hole and screw hole.
You would see new and interesting brain teasers to work out. A lot of shapes in this thread aren't so much hard, And challenging, as much as they are tedious and complicated. (if there's a difference lol). But maybe that's just me.
I think you should do it man
I made a quick list of some shapes that would be cool. http://imgur.com/a/8fG1d
http://imgur.com/a/0Qdxx
http://imgur.com/a/kND1U
I think that sub-d modelling is technical job since you really have to understand the proccess. I do belive it's a crucial skill in this cg. I just think that it's knowledge + experience. Problem is that the knowledge is here. I feel like you have written posts about this thousand times. It's not like suddenly you started writing something new and ground breaking. It's still awesome stuff and something everyone should read. I just find it funny that there is this cycle when people look at subd modelling and become amazed again :P
Experience on the other hand is something people have to gather (ground breaking statement!). I'm not that experienced with crazy shapes in subd but every time I stumbled upon something weird I tried. Sometimes it takes some time but eventually I will get it.
Just as you previously wrote the importatnt part is knowing your software. I used 3ds max (discovering FDDs... good times!) now I'm using Modo. Those softwares have so many cool tools which people won't even try.
I don't want to make it some long read because I don't think I have anything new to add. I get this feeling that people are awaiting for some AMAZING PERNA modelling trick that will change the world while they post mostly the same problem here without even browsing through this thread.
I hope that I expressed what I think properly (english isn't my language). I really respect proper sub-d modelling as a skill. You can't cheat though. There is no shortcut. Learn it and try (mostly by yourself).
Now what i think we lack a bit is good tools and fast process.
Next time I will do it with a breakdown. Modo 701 about 30 minutes.
Yes, you should absolutely work stuff out on your own and learn as much as you can via experience. But advocating slogging it out in isolation is silly, IMO. Learning from your peers is an intelligent thing to do.