Yeah, this seems to help out with the cloth simulation... but I just see a lot of terrible clothing coming out of it unless style and fit are considered, which is a whole skill set in itself. Just like any good modeling problem, you'll still have to gain some level of understanding before it looks good. Mom pants.
I guess it's arguing semantics - I'm learning how to be an actual craftsman and it's a very different mindset and process. It's more akin to being a programmer or tech-artist, in fact most of the CNC guys I'm studying under are ex IBM engineers.
Hehe, nice app and I'm sure it will find it's uses, but to me sculpting is lot more enjoyable than making those paper-cut things and stitching them together.
Also, you rarely get to make characters with casual shirts and pair of jeans in videogames. You usually have jacked up guys with bulletproof vests, 40 pockets and belts strapped all over. I guess setting up that complex sim would be quite tedious and time-consuming.
Pretty much what we thought. We really looked into it to cut down production time for final assets, not prototyping.
Also adding drapery on top of complex chars - not smooth bodies - e.g. drapery on top of armor - still causes clipping issues and other problems which you still have to manually tweak, which can be just as time consuming as doing it in ZB right away. It's not really a "sure time saver". It depends a lot on what you want to do if it's faster I think.
Another thing where it is useful are environments - e.g. drapery, table cloths, curtains. Env artists don't like doing this, assuming they know ZB at all, and then it's something they have no patience or experience. But the price tag is a bit high to use the software just for that. But it's much faster than setting stuff up in nCloth.
About the Zbrush issue - the main thing is the whole export - import stuff that's not really scriptable into an easy 1 (or a very few) click solution. It's a lot of annoying clicking and for a complex character you better get everything right the first time, or it's even more clicking...
Imho it's not really fully there yet where I'd like to see this app. But it definitely has a lot of potential. Maybe someone should tell those guys that there's a market not just in fashion design but in games.
I'm going to necro this thread rather than start a new one. And JEEZUS CRIPES was I bored reading thru it. But I used this recently and it is the shit.
Give up on your misconception that you are an artist - you are a craftsman. Someone tells you to build a boat, shelf, table. You build it. You are skilled at recognizing the quality of that construction
An artist creates something provocative from nothing. While you are MAKING ANOTHER PERSONS IDEA you are a craftsman.
Look at the hand plane - it makes wood flat. You can do it by hand yes. But a machine can do that just as well 1000 times as fast.
If it made it all shitty and pockmarked your expertise in the craft would recognize that.
My job is to make a character? Making a pattern and running a cloth sim that drapes accurately better than I can sculpt it yields a better result. I'm going to DO THAT instead.
So here we go:
30 seconds of sim vs 4+ hours of sculpting.
UVs like whoa. Sculpting seams, adding loops in zbrush. Endless possibilities.
1 of the tenants I live by:
"I want to create as fast as I can think."
Um, how the bloody hell did that take you 4 hours? I can put a naked character under bed-sheets and sculpt every fold by hand, and it wouldn't take me 4 hours.
I like the fact that something like that can be cut down to 30 seconds and be used as a base or generic base for generic characters, it's cool, but c'mon, be fair now, that body right there will take at most 30-60 minutes top to sculpt, while referencing a few pictures, or less if we're going to skip hyper-detail parts, like the inside of the clothing.
Did you try creating custom brushes in application like ZB with Alphas and Gravity features enabled? Those do about 50% of the work.
ace i'd like to see you create those wrinkles in 30-60 minutes from scratch in any sculpting app at this level of detail and quality. I#m with MM here but also see the worth of it in production, i also don't see it as cheating, or not creating your own, you still have to come up with _working_ sewing patterns, for the designs people tend to do for games, its easy to sculpt stuff to look like you want it, its not so easy to actually make it work in simulation - its a different kind of work, more technical, but its not like you can get great results for free or without any postwork, including partially resculpting to fit the clients need. Often enough i just simulate and drag stuff around for reference not to actually use it
i love MD for creating a base for parts of cloth. its not like you can use it 100% all the time, there are a few big issues with it. The first thing is someting that i'd call fold memory of cloth. folds stay in fabric, its not like if you stand long enough the wrinkles will be removed entirely from gravity. Thats something i didn't see in any cloth sim, the longer ypu calculate the smoother everything gets, unless you manually drag and pin it. the ideal way of doing it, would be posing the character and then moving it back to neutral state. But as it is right now, after getting back to neutral the cloth will move to a neutral state as well.
The other is, i have yet to find convincing settings for thick cloth, even denim is something thats super hard to achieve, most of the times the wrinkles are too fine, too soft, too generic to look believable.
teriyakis example is perfect, whats that pant made of? silk?
Marvelous Designer has been around a while now and I've seen it being used more and more in VFX/Post-Production facilities, one its own, or in conjunction with any existing tech.
I know that Passion Pictures used MD for their BBC Olympic titles - http://vimeo.com/45119481
It's true (as others have said) that Maya (nCloth), Softimage (Syflex) and Max (solver?) all have very competent cloth dynamics and solvers. And essentially this is what MD is, another solver. However it does appear to have some advantages when it comes to setting clothing up for simulation, and I hear its simulations are fast as well. Simulations (of any kind) can take a long time to calculate, several hours in fact, so any way to reduces this would be of interest to some people.
They way it uses the actual proper cuts of the cloth, is very good and can allow for some very good accuracy.
When it comes to games though (and film), they key will be whether there's an SDK/API for people to tap into, to allow better interop with pipelines. Games would probably need a version of the solver capable of being implemented into the respective game engine tech. However, some games might not want realtime cloth, or at least not have it chewing up performance. It can be expensive. Currently, I've seen studios use a mixture of techniques for cloth, some build controllers into the animation rigs, so their animators have more control. Others us something like Nvidias PhysX cloth (which is pretty good), or even a mix of both hand animated and realtime solver. I guess the main thing is if, when and how you want cloth in your game, that will decide the potential solution.
I don't really think it's a question of one or the other, just like you don't usually hear people saying definitively 'sculpting' or 'modeling', they both have specific functions.
Honestly sculpting folds for flowy things is rarely ideal, you often see doughy results when people try to sculpt things like capes, it's not accurate to the physical folding of actual material and it can't really be used dynamically. For simple shirt creasing, details that might be baked down, you don't need to go through the hassle of literally constructing the clothing if you don't need it to exist as actual clothing.
For armor and more structured things I think that it'd be crazy to make them in MD. For capes and drapey clothing, for flowing skirts with folds and ruffles, dear lord just use MD, you can literally sow on frilly pieces in seconds and they can simulate as actual geometry.
On some new level with this type of tech being available and making a wider appearance. I IMO, believe most artists (character ones in particular) are threatened by the idea of having something like this take over what was self-taught by ones self over a period of time to perfection (or tried).
I tried using it at work, and it takes time learning. So I put it aside until I master it. But I have to say as a character artist. It's a great nifty tool, and no. This will not cut away production, or the workload for those who are looking at each part of their work, like a golden coin of fortune... It would certainly be nice to have more time refining other things on characters than cloth. When its a constant view point of a characters detail. Though production schedules only allow an artist to do so much in a specific given amount of time. So my feelings are quite neutral about software tools like this.
But... it.. didn't take me 4 hours... I think there's a VS you missed reading.
Personally - I could probably scratch sculpt this in dynamesh in 2-3 hours. But I said 4 as an average. If you're faster - good on you. Imo everything takes 4 hours haha.
My focus is the UVs - adding group loops in zbrush with this base is brilliant - and sculpting on each panel of cloth independent of the other is ideal to get a more realistic result.
Ah OK, my bad, I guess it a case of "Lost in Translation" deal on my behalf, apologies mate.
ace i'd like to see you create those wrinkles in 30-60 minutes from scratch in any sculpting app at this level of detail and quality. I#m with MM here but also see the worth of it in production, i also don't see it as cheating, or not creating your own, you still have to come up with _working_ sewing patterns, for the designs people tend to do for games, its easy to sculpt stuff to look like you want it, its not so easy to actually make it work in simulation - its a different kind of work, more technical, but its not like you can get great results for free or without any postwork, including partially resculpting to fit the clients need. Often enough i just simulate and drag stuff around for reference not to actually use it
i love MD for creating a base for parts of cloth. its not like you can use it 100% all the time, there are a few big issues with it. The first thing is someting that i'd call fold memory of cloth. folds stay in fabric, its not like if you stand long enough the wrinkles will be removed entirely from gravity. Thats something i didn't see in any cloth sim, the longer ypu calculate the smoother everything gets, unless you manually drag and pin it. the ideal way of doing it, would be posing the character and then moving it back to neutral state. But as it is right now, after getting back to neutral the cloth will move to a neutral state as well.
The other is, i have yet to find convincing settings for thick cloth, even denim is something thats super hard to achieve, most of the times the wrinkles are too fine, too soft, too generic to look believable.
teriyakis example is perfect, whats that pant made of? silk?
As I said, I have brushes setup for that (in my case, ZB is doing 90% of the work with Gravity, Alphas, DepthMasking, etc) so yes, it is easy on my end, I'm not going to spend 4 hours with the standard brushes in ZB, that's beyond insane.
I wasn't saying it was cheating, I'm saying it's great, if you're making a Sci-Fi or even fantasy clothes, you can use this has a base, or if you're creating generic character, this would be an easy quick solution to speed up the pipeline. No matter how you look at this, this tool will be useful (unless you're making naked characters).
My problem was, the previous post (for me) came off as abit of 1 sided argument, kinda like how someone would tell me Marmoset is better then UDK in showing off a model when it isn't, both of them do the same thing at the end of the day if you know what you're doing, only that one is a little more 'open' to the users skillset, but they lacked the knowledge for the latter, hence making excuses on their own skillset.
If I had to use it in a production environment, sure why not. Gets the job done faster.
But at home, different story. Training and enjoyment (not that I don't get SOME elements of that from work) take over completely and I would find it more fulfilling to do the job myself and would learn a lot from it too.
I guess it comes to the fact that at home, you have complete control over your product (as opposed to producers, designers, etc demanding change/input) and in a sense (at least I do) more pride over the final outcome. Doing a lot of the hard graft by hand seems to be a part of that.
However, job = deadlines = tight balance between quality and efficiency.
Aw yiss! You know literally every time I have to make cloth (very little), see people making cloth, or great cloth sims I think of this thread and wonder if anyone got use out of this program! It looked fantastic! Great to know it's viable.
One thing I think is great about this over sculpting is it looks pretty nondestructive, right? If you want to move a blanket or thicken up a material in zbrush, it's going to look very dodgy compared to just being able to fiddle with the settings you need.
Aw yiss! You know literally every time I have to make cloth (very little), see people making cloth, or great cloth sims I think of this thread and wonder if anyone got use out of this program! It looked fantastic! Great to know it's viable.
One thing I think is great about this over sculpting is it looks pretty nondestructive, right? If you want to move a blanket or thicken up a material in zbrush, it's going to look very dodgy compared to just being able to fiddle with the settings you need.
Yeah it really does free you to experiment constantly with the design, you can change the properties of the fabric on the fly, can literally change the design and sowing patterns and watch the design change in real time.
To the people saying this is like cheating, that's comparable to saying designing and making clothing in real life with real fabric and real physics is cheating.
This is an entirely different ballgame to sculpting; that doesn't make it a bad thing, doesn't make it like poser either. Nothing is already completed, nothing is made or designed for you when you start up MD. You have to design the clothing as you would in real life, cut out pieces of fabric with specific measurements and tapering, sew them together in specific ways, use all of the techniques a seamstress would (I've had to learn darting and all that jazz). You still definitely need to know what you're doing, it's not some clothing easy button, just uses different skills to sculpting, and honestly that actually makes a lot of sense.
I had a go with it a couple of years ago and was impressed. Took my own character model into it, made trousers for him and got the result out and into zbrush for some smoothing and tweaks and baked maps.
I thought it was great, but it's also pretty expensive.
Here some stuff from my vip.
I think its good time saver for doing base meshes for sculpting in Z, but not a z-killer. No need to think about topology, just tailoring.
And its extremly helpfull when you need some military-like equipment with lot of straps(wich composed and layered) interacting with cloth and each other.
Problems whith popping easy to solve. Reapetable stuff can be done just for free.
Definetly need to try for everybody, no matter good at sculpting or begginer as i.
Edit: One more thing. If you want folds that "works not only in one pose" you can use "pins" that are......pins. You can create support points and define manually where folds must seat, like you can do with your fingers in real life. But you still need sculpt "memory folds" for believeble look.
Replies
Pretty much what we thought. We really looked into it to cut down production time for final assets, not prototyping.
Also adding drapery on top of complex chars - not smooth bodies - e.g. drapery on top of armor - still causes clipping issues and other problems which you still have to manually tweak, which can be just as time consuming as doing it in ZB right away. It's not really a "sure time saver". It depends a lot on what you want to do if it's faster I think.
Another thing where it is useful are environments - e.g. drapery, table cloths, curtains. Env artists don't like doing this, assuming they know ZB at all, and then it's something they have no patience or experience. But the price tag is a bit high to use the software just for that. But it's much faster than setting stuff up in nCloth.
About the Zbrush issue - the main thing is the whole export - import stuff that's not really scriptable into an easy 1 (or a very few) click solution. It's a lot of annoying clicking and for a complex character you better get everything right the first time, or it's even more clicking...
Imho it's not really fully there yet where I'd like to see this app. But it definitely has a lot of potential. Maybe someone should tell those guys that there's a market not just in fashion design but in games.
Um, how the bloody hell did that take you 4 hours? I can put a naked character under bed-sheets and sculpt every fold by hand, and it wouldn't take me 4 hours.
I like the fact that something like that can be cut down to 30 seconds and be used as a base or generic base for generic characters, it's cool, but c'mon, be fair now, that body right there will take at most 30-60 minutes top to sculpt, while referencing a few pictures, or less if we're going to skip hyper-detail parts, like the inside of the clothing.
Did you try creating custom brushes in application like ZB with Alphas and Gravity features enabled? Those do about 50% of the work.
i love MD for creating a base for parts of cloth. its not like you can use it 100% all the time, there are a few big issues with it. The first thing is someting that i'd call fold memory of cloth. folds stay in fabric, its not like if you stand long enough the wrinkles will be removed entirely from gravity. Thats something i didn't see in any cloth sim, the longer ypu calculate the smoother everything gets, unless you manually drag and pin it. the ideal way of doing it, would be posing the character and then moving it back to neutral state. But as it is right now, after getting back to neutral the cloth will move to a neutral state as well.
The other is, i have yet to find convincing settings for thick cloth, even denim is something thats super hard to achieve, most of the times the wrinkles are too fine, too soft, too generic to look believable.
teriyakis example is perfect, whats that pant made of? silk?
I know that Passion Pictures used MD for their BBC Olympic titles - http://vimeo.com/45119481
It's true (as others have said) that Maya (nCloth), Softimage (Syflex) and Max (solver?) all have very competent cloth dynamics and solvers. And essentially this is what MD is, another solver. However it does appear to have some advantages when it comes to setting clothing up for simulation, and I hear its simulations are fast as well. Simulations (of any kind) can take a long time to calculate, several hours in fact, so any way to reduces this would be of interest to some people.
They way it uses the actual proper cuts of the cloth, is very good and can allow for some very good accuracy.
When it comes to games though (and film), they key will be whether there's an SDK/API for people to tap into, to allow better interop with pipelines. Games would probably need a version of the solver capable of being implemented into the respective game engine tech. However, some games might not want realtime cloth, or at least not have it chewing up performance. It can be expensive. Currently, I've seen studios use a mixture of techniques for cloth, some build controllers into the animation rigs, so their animators have more control. Others us something like Nvidias PhysX cloth (which is pretty good), or even a mix of both hand animated and realtime solver. I guess the main thing is if, when and how you want cloth in your game, that will decide the potential solution.
Honestly sculpting folds for flowy things is rarely ideal, you often see doughy results when people try to sculpt things like capes, it's not accurate to the physical folding of actual material and it can't really be used dynamically. For simple shirt creasing, details that might be baked down, you don't need to go through the hassle of literally constructing the clothing if you don't need it to exist as actual clothing.
For armor and more structured things I think that it'd be crazy to make them in MD. For capes and drapey clothing, for flowing skirts with folds and ruffles, dear lord just use MD, you can literally sow on frilly pieces in seconds and they can simulate as actual geometry.
I tried using it at work, and it takes time learning. So I put it aside until I master it. But I have to say as a character artist. It's a great nifty tool, and no. This will not cut away production, or the workload for those who are looking at each part of their work, like a golden coin of fortune... It would certainly be nice to have more time refining other things on characters than cloth. When its a constant view point of a characters detail. Though production schedules only allow an artist to do so much in a specific given amount of time. So my feelings are quite neutral about software tools like this.
As I said, I have brushes setup for that (in my case, ZB is doing 90% of the work with Gravity, Alphas, DepthMasking, etc) so yes, it is easy on my end, I'm not going to spend 4 hours with the standard brushes in ZB, that's beyond insane.
I wasn't saying it was cheating, I'm saying it's great, if you're making a Sci-Fi or even fantasy clothes, you can use this has a base, or if you're creating generic character, this would be an easy quick solution to speed up the pipeline. No matter how you look at this, this tool will be useful (unless you're making naked characters).
My problem was, the previous post (for me) came off as abit of 1 sided argument, kinda like how someone would tell me Marmoset is better then UDK in showing off a model when it isn't, both of them do the same thing at the end of the day if you know what you're doing, only that one is a little more 'open' to the users skillset, but they lacked the knowledge for the latter, hence making excuses on their own skillset.
But at home, different story. Training and enjoyment (not that I don't get SOME elements of that from work) take over completely and I would find it more fulfilling to do the job myself and would learn a lot from it too.
I guess it comes to the fact that at home, you have complete control over your product (as opposed to producers, designers, etc demanding change/input) and in a sense (at least I do) more pride over the final outcome. Doing a lot of the hard graft by hand seems to be a part of that.
However, job = deadlines = tight balance between quality and efficiency.
One thing I think is great about this over sculpting is it looks pretty nondestructive, right? If you want to move a blanket or thicken up a material in zbrush, it's going to look very dodgy compared to just being able to fiddle with the settings you need.
haha! Damn, I open just one thread today and I see my own face! Forgot posting that! :poly124:
Yeah it really does free you to experiment constantly with the design, you can change the properties of the fabric on the fly, can literally change the design and sowing patterns and watch the design change in real time.
To the people saying this is like cheating, that's comparable to saying designing and making clothing in real life with real fabric and real physics is cheating.
This is an entirely different ballgame to sculpting; that doesn't make it a bad thing, doesn't make it like poser either. Nothing is already completed, nothing is made or designed for you when you start up MD. You have to design the clothing as you would in real life, cut out pieces of fabric with specific measurements and tapering, sew them together in specific ways, use all of the techniques a seamstress would (I've had to learn darting and all that jazz). You still definitely need to know what you're doing, it's not some clothing easy button, just uses different skills to sculpting, and honestly that actually makes a lot of sense.
But theyve been using stuff like this in movies forever and its yet to creep into the game workflow
I thought it was great, but it's also pretty expensive.
I think its good time saver for doing base meshes for sculpting in Z, but not a z-killer. No need to think about topology, just tailoring.
And its extremly helpfull when you need some military-like equipment with lot of straps(wich composed and layered) interacting with cloth and each other.
Problems whith popping easy to solve. Reapetable stuff can be done just for free.
Definetly need to try for everybody, no matter good at sculpting or begginer as i.
Edit: One more thing. If you want folds that "works not only in one pose" you can use "pins" that are......pins. You can create support points and define manually where folds must seat, like you can do with your fingers in real life. But you still need sculpt "memory folds" for believeble look.