Icons of Andria/ My first animation!
-
Kerro Perro
- Posts: 306
- Joined: 20 Mar 2011, 01:01
- Location: Leiden, the Netherlands
- Contact:
Re: Icons of Andria/ My first animation!
Ah thanks, yeah i was wondering that sitting behind my desk today: texture Andria or model the other assets? I was tempted to keep modelling so i think i'll do that!
Re: Icons of Andria/ My first animation!
Thats right Kerro, hit up all the assets you need, say in the groove. It'll take less time with each one you do, and you'll keep building on the those skills. When all are done, get read for the next "texturing groove". And if at that point you need to make model adjustments - it'll be so easy.Kerro Perro wrote:Ah thanks, yeah i was wondering that sitting behind my desk today: texture Andria or model the other assets? I was tempted to keep modelling so i think i'll do that!
Look forward to the next clay.
-
RichardCulver
- Posts: 34
- Joined: 07 Jan 2012, 18:21
Re: Icons of Andria/ My first animation!
It is looking good. It is OK to experiment with workflow. And you learn your lessons.
I would only caution that you are boxing your self into a corner technically if you fire through all of one stage of an asset. I agree with the reasoning stated. However I also have learned the hard way that you can find yourself repeating mistakes that you had not anticipated. Then you can look back at all of your assets and realize they will all have to be changed. That is a PIA. And I have been there.
Doing characters is a very technical process from one end of the pipeline to the other. And it is probably the most challenging and technically interdependent of all the assets. Each phase feeds into the other. Modeling, UV mapping, envelop painting, rigging and deformation. Then finally any dynamics that have to be added, such as cloth. Not to mention facial animation. You have to take all of these things into consideration before you finalize a single character. It needs to be rigged up, deformed, put through motion tests and all of that before you even get to texturing because you may find you'll change the poly flow or even find geometry that does not deform well at all or find you have to solve exactly how you will deform it or change it or simply omit it all together or go into a redesign.
Your modeling and design is fantastic. But I can also tell you you have put geometry in parts of the body that are the most difficult to get deforming properly even on a smooth mesh.
I would strongly suggest taking a supplied model or make a stripped down human form just to experiment with the entire process from start to finish and make notes as to what were the areas that caused difficulty and what you had to do to change geometry, bone positions, poly flow and so on in order to make it work for you. Then go back and visit your first chracter and take her through the same steps. And see what happens. Solve the problems that come up, take notes and apply that to the rest of your characters.
Here is a fantastic sight that will get you going in the right direction for connecting the modeling and poly flow process with bone articulation:
http://www.hippydrome.com/
I once took his images and modeled an entire character with exactly the same poly flow to simply study the process. You can not really learn to much about this. It is a constant process you perfect.
But the last thing you need to do is paint yourself into a corner technically when approaching the modeling stage.
Unfortunately, 3D animation is a highly technical subject that has to be approached from the view of the entire pipeline from start to finish. It will save you lots of work and frustration in the long run if you approach it that way.
I would only caution that you are boxing your self into a corner technically if you fire through all of one stage of an asset. I agree with the reasoning stated. However I also have learned the hard way that you can find yourself repeating mistakes that you had not anticipated. Then you can look back at all of your assets and realize they will all have to be changed. That is a PIA. And I have been there.
Doing characters is a very technical process from one end of the pipeline to the other. And it is probably the most challenging and technically interdependent of all the assets. Each phase feeds into the other. Modeling, UV mapping, envelop painting, rigging and deformation. Then finally any dynamics that have to be added, such as cloth. Not to mention facial animation. You have to take all of these things into consideration before you finalize a single character. It needs to be rigged up, deformed, put through motion tests and all of that before you even get to texturing because you may find you'll change the poly flow or even find geometry that does not deform well at all or find you have to solve exactly how you will deform it or change it or simply omit it all together or go into a redesign.
Your modeling and design is fantastic. But I can also tell you you have put geometry in parts of the body that are the most difficult to get deforming properly even on a smooth mesh.
I would strongly suggest taking a supplied model or make a stripped down human form just to experiment with the entire process from start to finish and make notes as to what were the areas that caused difficulty and what you had to do to change geometry, bone positions, poly flow and so on in order to make it work for you. Then go back and visit your first chracter and take her through the same steps. And see what happens. Solve the problems that come up, take notes and apply that to the rest of your characters.
Here is a fantastic sight that will get you going in the right direction for connecting the modeling and poly flow process with bone articulation:
http://www.hippydrome.com/
I once took his images and modeled an entire character with exactly the same poly flow to simply study the process. You can not really learn to much about this. It is a constant process you perfect.
But the last thing you need to do is paint yourself into a corner technically when approaching the modeling stage.
Unfortunately, 3D animation is a highly technical subject that has to be approached from the view of the entire pipeline from start to finish. It will save you lots of work and frustration in the long run if you approach it that way.
-
Kerro Perro
- Posts: 306
- Joined: 20 Mar 2011, 01:01
- Location: Leiden, the Netherlands
- Contact:
Re: Icons of Andria/ My first animation!
Hi Richard, first off thanks for taking an interest and sharing your insights! 
Then an embarrassing confession: i did not model her body, it's a Species body with tweaking (mostly on T&A) with a Daz head on it. I did not mean to give the impression that i did more than i did, i just forgot about it - i did mention on it DA i think...
But so hence why i said i took waaay to long, nearly all that time was spent on her "extra bits" and nearly every one of those is a third iteration or more. But again that was my crappy workflow, i have now streamlined it and kicked out Z-Brush completely, even though i think it's awesomely cool for this project it's just a time waster. Might use it for some terrain stuff further down the road, or i might just do that in Mudbox as i texture paint in that anyway.
Anyhoo i've got a modelling workflow that works for me now: do a rough sketch > make basic model > take screencap > sketch over that > refine modelling > take screencap etc. Feels natural and allows me stay close to the comic book feel i'm going for:
Currently working on the ship where most of the action takes place, decided to leave the characters for a bit because as you rightly predict rigging them is causing problems:
The thingies on her back ( i don't know what they are ;) ) were deforming very badly when i posed her for the turntable. Now i managed to fix it somewhat with "weight" painting for that certain pose but yeah when i tested some movements it became very problematic. For the turntable she just was two meshes: her body/skin and all the other bits merged and GATOR'd to get the Species envelope - quick and dirty. The body/skin mesh is pretty well weighted by Species and the few (typical) bad areas of knees, thighs, elbows etc i am hoping to cure with Adam Sale's ICE shape correction techniques: https://vimeo.com/album/2167864/video/27562602
But for the bits i'm not certain, a lot of them deformed quite well but for those things on her back and a few others i'm thinking about just "pinning" them to her body with point cluster nulls. Though i must say the few tests i did with that were not really promising...
But anyway yeah i really see your point, i will keep plugging away with the non-characters assets for now, but am going to rig the ship up as soon as i finish modelling. For that i'm planning very simple ICE driven rigs with wich i do have some experience - unlike character rigging at wich i'm a noob.
Frankly i'm not not sure about a "test character" - actually sounds like a fun part but maybe i just should push forward with the character i already have and deal with the problem as they arise? Gonna roll that around in my head for a while...
Thanks again!
Then an embarrassing confession: i did not model her body, it's a Species body with tweaking (mostly on T&A) with a Daz head on it. I did not mean to give the impression that i did more than i did, i just forgot about it - i did mention on it DA i think...
But so hence why i said i took waaay to long, nearly all that time was spent on her "extra bits" and nearly every one of those is a third iteration or more. But again that was my crappy workflow, i have now streamlined it and kicked out Z-Brush completely, even though i think it's awesomely cool for this project it's just a time waster. Might use it for some terrain stuff further down the road, or i might just do that in Mudbox as i texture paint in that anyway.
Anyhoo i've got a modelling workflow that works for me now: do a rough sketch > make basic model > take screencap > sketch over that > refine modelling > take screencap etc. Feels natural and allows me stay close to the comic book feel i'm going for:
But for the bits i'm not certain, a lot of them deformed quite well but for those things on her back and a few others i'm thinking about just "pinning" them to her body with point cluster nulls. Though i must say the few tests i did with that were not really promising...
But anyway yeah i really see your point, i will keep plugging away with the non-characters assets for now, but am going to rig the ship up as soon as i finish modelling. For that i'm planning very simple ICE driven rigs with wich i do have some experience - unlike character rigging at wich i'm a noob.
Frankly i'm not not sure about a "test character" - actually sounds like a fun part but maybe i just should push forward with the character i already have and deal with the problem as they arise? Gonna roll that around in my head for a while...
Thanks again!
-
RichardCulver
- Posts: 34
- Joined: 07 Jan 2012, 18:21
Re: Icons of Andria/ My first animation!
Yeah that is indeed one I roll around all the time. I mean there are times I feel like I could do that for a year straight and not get anything done on my project.
I have been working on a personal project that I started back in 2005. Granted, a lot of life lived in between. A lot of frustration, giving up, several times, overwhelm, you name it. But I have persisted. And I think in the end - that is as of my thinking now - I have learned the hard way once again to simply take the time to learn how do it right. That you can not rush it in that sense.
I can not advocate, you know, what you should do. But just think first about separating education/training from production. From the point of view of isolating particular items that need addressing. And that there is also the concept of knowing something so well you don't have to think about it. So that when you are in production you are simply attacking particular problems that you already have a good idea how to solve. But anytime you come against something that you are hitting a block on, realize that it has to step backward into a training situation. And for that it is always better to use similar assets are designed to be used only to learn a particular task and bot burdened by the complexity you find in production.
Then once you learn that task, test it out with all of the other things it has to work with in production. In your particular case not in another scenario.
So in the case of a human character, the reason I suggest this is because this is how I like to approach problems. For example:
What is the best way to rig shoulders in general? (Not my character but just learn how the shoulder should work etc.)
What is the best way to do Enveloping? (Again without the stress of the character I have become attached to etc.)
After going through a entire list of things that need to be solved - generically - then I can go back and have a good judgement on what it is I am working on.
I can then make very analytical decisions about production problems. I can say, "well I was doing this right" and "this over here I had been approaching fundamentally wrong. Now I how how to change it."
And in this way I don't feel locked into the asset I have committed to and try to fit square pegs into round holes so to speak. Not trying to force things into my limitied vew, but rather the other way around. First just expand my perception to the way things should work. At least to have that in place first.
From there there are then always problems that are unique. But I always try and approach them the same way.
Now I do have a suggestion for your appendages on the character. And a practical application of the theory I have been talking about.
And that is vector displacement maps. You can do those in Mudbox. If you first just take a simple portion of a character. Like an arm or a leg. And try and do some similar sculpting to what you will do with your character. You can actually sculpt the appendage separately and turn it into a vector displacement map. then painted on the character. These can be imported into SI.
Then what you'd have is a character that was deforming on the base mesh. And the appendages are vector displacement maps that are tessellated at render time. In theory it might actually work. But you'd have to do testing. And then once you find something that works then apply that to the rest of the pipeline.
I have been working on a personal project that I started back in 2005. Granted, a lot of life lived in between. A lot of frustration, giving up, several times, overwhelm, you name it. But I have persisted. And I think in the end - that is as of my thinking now - I have learned the hard way once again to simply take the time to learn how do it right. That you can not rush it in that sense.
I can not advocate, you know, what you should do. But just think first about separating education/training from production. From the point of view of isolating particular items that need addressing. And that there is also the concept of knowing something so well you don't have to think about it. So that when you are in production you are simply attacking particular problems that you already have a good idea how to solve. But anytime you come against something that you are hitting a block on, realize that it has to step backward into a training situation. And for that it is always better to use similar assets are designed to be used only to learn a particular task and bot burdened by the complexity you find in production.
Then once you learn that task, test it out with all of the other things it has to work with in production. In your particular case not in another scenario.
So in the case of a human character, the reason I suggest this is because this is how I like to approach problems. For example:
What is the best way to rig shoulders in general? (Not my character but just learn how the shoulder should work etc.)
What is the best way to do Enveloping? (Again without the stress of the character I have become attached to etc.)
After going through a entire list of things that need to be solved - generically - then I can go back and have a good judgement on what it is I am working on.
I can then make very analytical decisions about production problems. I can say, "well I was doing this right" and "this over here I had been approaching fundamentally wrong. Now I how how to change it."
And in this way I don't feel locked into the asset I have committed to and try to fit square pegs into round holes so to speak. Not trying to force things into my limitied vew, but rather the other way around. First just expand my perception to the way things should work. At least to have that in place first.
From there there are then always problems that are unique. But I always try and approach them the same way.
Now I do have a suggestion for your appendages on the character. And a practical application of the theory I have been talking about.
And that is vector displacement maps. You can do those in Mudbox. If you first just take a simple portion of a character. Like an arm or a leg. And try and do some similar sculpting to what you will do with your character. You can actually sculpt the appendage separately and turn it into a vector displacement map. then painted on the character. These can be imported into SI.
Then what you'd have is a character that was deforming on the base mesh. And the appendages are vector displacement maps that are tessellated at render time. In theory it might actually work. But you'd have to do testing. And then once you find something that works then apply that to the rest of the pipeline.
-
Kerro Perro
- Posts: 306
- Joined: 20 Mar 2011, 01:01
- Location: Leiden, the Netherlands
- Contact:
Re: Icons of Andria/ My first animation!
Oh on the "learning to do it right" front i totally hear ya! Luckily even though i'm doing this animation for a band since it's just free collaboration there is also no rush at all. I've gone through a similiar thing with my drawing where my desire to create overshot my actual skill, but now after years of practice it is catching up. I see no difference with this except that animation requires a lot more problem solving and a different kind of patience - if that makes sense ;)
As i said i need to roll the character issue in my head but i just want yell yipee about one thing: till now i've not been able to use displacement or normal maps for this thing because they do not play nice with the toon ink lens. And i love the toon ink lens sooo much! But after your post i quickly tried out VDM because i wondered if the different way they get processed (through wizardry and fairy dust for all i know) would make a difference... Turns out it does!
The toon ink lens is rendering the shape perfectly now! You do not want to know the weeks i spent trying to make regular displacement work...
So thanks for that, you've just given me another big tool to work with! Dunno if it'll solve the "back thingies" problem, they might still deform in ways i don't want but it's definitely something i will try out on other characters!
As i said i need to roll the character issue in my head but i just want yell yipee about one thing: till now i've not been able to use displacement or normal maps for this thing because they do not play nice with the toon ink lens. And i love the toon ink lens sooo much! But after your post i quickly tried out VDM because i wondered if the different way they get processed (through wizardry and fairy dust for all i know) would make a difference... Turns out it does!
The toon ink lens is rendering the shape perfectly now! You do not want to know the weeks i spent trying to make regular displacement work...
So thanks for that, you've just given me another big tool to work with! Dunno if it'll solve the "back thingies" problem, they might still deform in ways i don't want but it's definitely something i will try out on other characters!
-
RichardCulver
- Posts: 34
- Joined: 07 Jan 2012, 18:21
Re: Icons of Andria/ My first animation!
lol! I love it when the voodoo behind the curtain works out in out favor for a change!
And I hear you about the art thing. Yep it is the same in my view. Just a little more technical - in a different way anyway - as you said.
Glad yo got something working positive.
Look forward to more updates. Looking real good so far as I said.
And I hear you about the art thing. Yep it is the same in my view. Just a little more technical - in a different way anyway - as you said.
Glad yo got something working positive.
Look forward to more updates. Looking real good so far as I said.
Who is online
Users browsing this forum: No registered users and 20 guests