PDA

View Full Version : IKboost Question



DaveLEWIS
02-15-2009, 10:55 AM
Hi,

Just a quick IKboost question if I may:)

Does anyone know how to put IKboost on a rig that has already been set up with the more traditional methods, without completely starting from scratch?

DaveLEWIS

SplineGod
02-15-2009, 11:54 AM
Just apply IKB to your character directly or to the root of your rigs heirarchy. I used to use blended rigs but no longer do that because you lose alot of the power of IKB provides such as manipulating poses, motion clips etc.
How your rig is already setup will effect how IKB may or may not work with it.

DaveLEWIS
02-15-2009, 12:17 PM
Hi Larry,

Sadly, I tried this, but still no go. I am using LW 9.6. When I load up the model from scratch, convert skelegons, select the model and hit IKboostTool, it works no problemo, but when I try to do this to an existing rig, I find no response. Have you found this to happen?

DaveLEWIS

SplineGod
02-15-2009, 02:52 PM
Again it depends on the rig.
I just tried it on a rig that has standard IK and it works in 9.6.
I parented the character to a master null as well as the IKGoals.
I applied IKB to the master null and it works.

VFXCast
02-16-2009, 09:42 AM
Dave, Larry also has about 3 hours of free IK Booster tutorials online at VFXCast.com. They are among some of our most viewed videos.

Thanks Larry

GraphXs
02-16-2009, 10:11 PM
Does it work with Joints? IKB is a great tool, hope Newtek will eventually give us something similar in CORE. :thumbsup: By the way, thanks for all the great info on IK Boost Larry!

SplineGod
02-16-2009, 10:24 PM
IKBoost works on any heirarchy including joints. You could use it on a car or other type of rig. Doing so gives you all the other advantages of IKB such as saving and reusing poses and motion clips, dynamics etc.

Dirk
02-17-2009, 06:14 AM
Hi Larry,

Sadly, I tried this, but still no go. I am using LW 9.6. When I load up the model from scratch, convert skelegons, select the model and hit IKboostTool, it works no problemo, but when I try to do this to an existing rig, I find no response. Have you found this to happen?

DaveLEWIS


What happens if You remove IK and modifiers from the bones that You want to control with IKB, and apply IKB after that?

Another option would be to bake the motions (if there is an animation), remove the modifiers, etc., and apply IKB after that.

SplineGod
02-17-2009, 08:11 AM
IKB will only effect those channels that do not have standard IK, expressions or plugins applied.

kishigo
02-17-2009, 05:13 PM
IKBoost works on any heirarchy including joints. You could use it on a car or other type of rig. Doing so gives you all the other advantages of IKB such as saving and reusing poses and motion clips, dynamics etc.

IKB works with joints? I was told joints were incompatible with IKB. That's good to know.

SplineGod
02-17-2009, 05:35 PM
Works for me. Setup a null and draw some joints. Apply IKB to the null. It should work.

wsantiago
02-18-2009, 02:51 AM
Is IKBoost going to be included in Core? This is some of the info that would be good to know before March 31st.

mav3rick
02-18-2009, 04:06 AM
Is IKBoost going to be included in Core? This is some of the info that would be good to know before March 31st.

no.... lw core will have ik booster on steroids means new CA stuff in lw core should be somethin lot better than standard tools you have in lw

SplineGod
02-18-2009, 08:09 AM
Im not aware that Newtek has accounced any sort of CA tools for core. I would be very surprised if core had any sort of CA tools beyond the standard ones LW has now. If core doesnt have at least the minimum functionality of IKB Ill continue to use LW9.x

kishigo
02-18-2009, 12:46 PM
I agree. I hope there's at least some sort of IKB functionality in Core. I've come to really like that tool. As a hobbyist, it's difficult for me to get my head around the more complex rigs. I prefer to just move nodes around and load and save poses and motions.

Kevin

SplineGod
02-18-2009, 01:23 PM
Hard to say. Ive seen little evidence that anyone at NT really knows whats in IKB or what it can do. Hopefully strong input from ppl now may help influence what goes in.
As far as complex rigs go I agree. Many of the shows Ive worked on dont have the time to create and modify complex rigs. IKB allows you to animate in the shortest period of time. Afterall its about animating in the end. If NT is serious about preserving those parts of LW that make it great for doing things quickly then they need to take a serious look at IKB and keep the strong things, fix the weakest parts and take it further. Ironically the closest thing Ive seen to IKB is in Blender.

DaveLEWIS
02-19-2009, 01:48 PM
Nice thread:) I have to say that I like IKB very much and hope that it doesn't disapear after 9.6.

DaveLEWIS

geothefaust
02-20-2009, 12:11 AM
Hey Larry, I haven't been using MSN much at all. Otherwise I'd just ask you on there. I'm curious though, you're a great rigger and animator, among other things. I think you of all people, would be great to see having some sort of influence over the direction of CORE. Do you plan on joining HC or are you one or their "key account users"? If you gave input, we could very well have tools just like IKB, or even IKB on steroids.

Just curious. :)

omeone
02-20-2009, 03:47 AM
Nice thread:) I have to say that I like IKB very much and hope that it doesn't disapear after 9.6.

DaveLEWIS

I think it is unlikely, IKB development stopped a long time ago. But that's not a problem - there's nothing stopping me from having 8.5 installed with anyother LW install (which I do anyway). IKB was built in such a way to make it very easy to move data between characters / scenes / versions. I would expect the CORE approach to CA to be one new unified system and homogenous with 'BIG 4' workflows. Disappointing if it turns out to be true, but not the end of the world. They won't go back and 'break' anything we already have :)

Surrealist.
02-20-2009, 05:04 AM
Yeah the IKB in Blender is the Auto IK function. It is built right into the app. It works very much like IKB in so far as dragging stuff around but Blender has a good deal of pose functions (copy paste etc) built in when you are using the Pose Mode which you do for animation. The nice thing here is that you also have all those functions for standard IK in Blender. You don't have to have Auto IK. I think IKB has some very nice functions, some I'd like to see in Blender too.

After working with it and seeing the limitations of both IKB and Auto IK it finally hit me what the problem with both systems is. It must have to to with the math involved. I tried similar movements with both systems and found that for instance keeping a node such as a hand or foot locked produced nearly the exact effect in IKB and Auto IK. That familiar looping or wobble. In IKB the "solution" is Binding and Baking. I quote solution because, though I have spent hours on the great tutorials at www.IKBoost.com and days even weeks working - full time by the way - with rigging and animating in an R&D mode specifically aimed at mastering IKB to find much to my disappointment that neither binding nor baking would give me an acceptable quality I was looking for in my animation.

The problem with this I can only guess. And my complete speculation is that the writer of IKB knew that due to the complex math involved it was virtually impossible and when he saw that Binding and Baking gave rather stiff (some say amateurish) results for prolonged serious work he dropped it. Because basically what you are asking IKB to do is calculate all of these rotations in FK based on pulling the nodes around in a IKB control mode. And this system and approach seems to be flawed as a solution for various levels of animation. (which is why in Blender I suspect it is not implemented) Good for some things but in the end not a complete replacement for the standard IK system. It took a long time working with it to finally come to that conclusion. And the reason as I say is only my guess. But what we do know is that the development stopped. And I think it stopped because it was a dead end. Just my opinion and I do understand all of its uses and yes, I have seen very very nice animation with it up to a certain point. Just that it did not work for me.

So what I wound up doing is using IKB with joints but in a different way.

I began to look at IKB as two separate things. 1) The actual "IK" system and 2) The controls and other tools you get with it. You can use these controls in IKB regardless of whether or not you use the IKB "IK".

What I did was I set most of the controls to "Move". This made it basically the same thing as moving a null. But what that also did was take the "IK" out of IKB for most of my rig. And since I liked the way the arms moved with IKB I left those in as IKB (controllers set to Rotate). But for the legs, pelvis, spine and head I used targeting primarily and the controllers (pole targets etc) were all mostly based on movement, with some set to an FK rotation control.

This allowed me to animate in IKB and make a rig that worked for me because I found myself gravitating towards a workflow that had me setting up IK stops at set points of the rig anyway so that free movement of the whole rig aspect of it I did not mind missing. Plus it did not require baking or binding the feet and still gave me all of the animation controls and tools I found useful and so much better in IKB. Just being able to click and drag a controller in IKB is so much better than standard Layout controls to me.

For the hands if I then needed to place a hand I would have a rig for that shot that was specifically designed for it. But I never got that far and have been using Blender lately.

So my hope for LightWave would be that they look at the functionality of IKB and Blender or any other packages in so far as the other things such as saving and copying poses and all of the other cool things in IKB including much better ways to control the animation process and put that into the main part of the program as standard functions for all things including lights camera etc.

As for the IK in IKB I think it is a dead end road. I see that as a technical fact and I could be wrong, but it is my observation. It just seems like if it would work, it would have been done.

The real shame is that they never did anything to improve layout across the boards in the same way.

Here are the animations I did with this rig set up:

http://www.vimeo.com/2510941

And by the way I do not recommend this rig. It has some very nice features such as being able to turn on the dime very easliy but had too many other problems. I am not rigging master nor any IKB expert. I am only giving my opinion of it from my experience.

The reason I dumped the idea was that joints were too strange. Seems like they were not worked out fully as of 9.5. Since then I have been learning Blender and not using LightWave but I do hope they listen to users and improve the animation side and most of all make things actually work.

SplineGod
02-20-2009, 08:56 AM
I agree with omeone. I use iKB all the time with 9.6. Others Ive been working with love using iKB once they get their hands on it. Weve used it quite successfully on several TV shows. If we had to spend the time rigging the usual way and tweaking/tuning those rigs we wouldnt get much done.
I dont even use the baking or binding that much to be honest and they work fine when used properly. I generally do those as a finalizing stage and only if the camera will see those parts of the character.

Geofaust,
Jays mentioned that NT wants to keep what makes LW what it is in Core. To me thats the speed and ease of use in getting things done which are very essential things especially in TV production work.
IKB fulfills that bill beautifully. Unfortunately I feel that theres really nobody at NT who has a good understanding of what IKB does or can do. The documentation on it makes this painfully obvious as well. This is based on various comments made about it by people from NT. Ive asked before what aspects of IKB they felt were worth preserving and never have gotten a response. I think theres a drive at NT to make LW, at least at first, into a Maya or XSI clone simply to make it "industry standard". So I expect that whatever CA tools they develop will be more like those in other apps rather then something innovative/fast/easy. I hope im wrong but Ive seen little to convince me otherwise.

The focus initially in Core is apparently going to be modeling. If thats the case Core when release will either have no CA tools or at the very worst what we have now but probably minus IKB.
I dont expect the first release of Core to actually hit Q4 simply based on past experience. When it does theres the issue of stability since traditionally most releases of LW have been pretty shakey. On top of all this theres the time required to actually make Core bulletproof enough to work reliably in production. Since Core is also a new app theres also the issue of simply convincing studios and end users to adopt it. The key thing I hear from coworkers is that the further NT gets away from that central LW ethic of being able to get things done quickly and easily the less likely people will move over to it.

As Colin pointed out before. NTs had IKB for how long? Theyve had the opportunity over the years to fix any issues, improve weaknesses, build up its strengths and so on. Even now people I work with who dont know LW are impressed with IKB because of its versatility, simplicity and speed. Its been an opportunity lost iMO but could be one again if the parts and principles that make it great are added into Core. In the end I dont care about IKB per se being in Core. I just want those aspects that make it powerful added into Core.

I personally have a vested interest in LW...I make a living off it. Ive worked at alot of studios and on a lot of freelance projects. Its my central app. Its a real pain having to move over to another app. If core doesnt have the things I need in it then Ill stay with the current version of LW until moving to something else that does do what I need.

ericsmith
02-20-2009, 09:03 AM
After working with it and seeing the limitations of both IKB and Auto IK it finally hit me what the problem with both systems is. It must have to to with the math involved. I tried similar movements with both systems and found that for instance keeping a node such as a hand or foot locked produced nearly the exact effect in IKB and Auto IK. That familiar looping or wobble. In IKB the "solution" is Binding and Baking. I quote solution because, though I have spent hours on the great tutorials at www.IKBoost.com and days even weeks working - full time by the way - with rigging and animating in an R&D mode specifically aimed at mastering IKB to find much to my disappointment that neither binding nor baking would give me an acceptable quality I was looking for in my animation.

The problem with this I can only guess. And my complete speculation is that the writer of IKB knew that due to the complex math involved it was virtually impossible and when he saw that Binding and Baking gave rather stiff (some say amateurish) results for prolonged serious work he dropped it. Because basically what you are asking IKB to do is calculate all of these rotations in FK based on pulling the nodes around in a IKB control mode. And this system and approach seems to be flawed as a solution for various levels of animation. (which is why in Blender I suspect it is not implemented) Good for some things but in the end not a complete replacement for the standard IK system. It took a long time working with it to finally come to that conclusion. And the reason as I say is only my guess. But what we do know is that the development stopped. And I think it stopped because it was a dead end. Just my opinion and I do understand all of its uses and yes, I have seen very very nice animation with it up to a certain point. Just that it did not work for me.

No! It can't be true!

It must be that... that you just don't have any clue what it's capable of! Yeah, that's it. You can't judge it unless you've used it for years in production!

In all seriousness, though, it sounds like you've wrapped your head around the underlying issues.

Ultimately, what I'd like to see in CORE (from the standpoint of IKB, at least) is some good FK manipulation tools ala Messiah. When it comes to all of the pose/motion loading and saving, that should be part of the foundation, not just tied to a CA tool.

Eric

Surrealist.
02-20-2009, 12:06 PM
Exactly. Just as it is in Blender.

I think that is the main drawback of IKB that it has to be some sort of plug in.

I don't think the point about functionality in a production environment has to be driven home anymore. At least not for me. And you know Eric, Larry, I was there in that other - more or less recent thread - the whole way. All I was saying at that point was I wanted to explore it myself and see. I wanted the information not to be discouraged by naysayers.

I still would encourage people to explore it. I do think if people use it like Larry has for quick production it is good. And I think people should look into it and explore it for themselves to see what it can do for them rather than just write it off.

But for me I think I did what I could. I am in production, but I also have the luxury of doing R&D as I am the boss. So I don't have weekly deadlines over my head. What I do have is a need for things to be at a certain level of quality and I think I more than satisfied any criticism that perhaps I was too quick to judge. I mean I literally was up all hours doing nothing but R&D, study, practice and test after test after test for a couple of weeks straight. I did learn how to bake and bind very well actually. I figured out all of the various pitfalls and workarounds to that end as well. - Thank you Colin!

So in the end I'd say as an animation system in itself, it is best left behind. That is my opinion. As an inspiration for functionality it should be studied. It is quite brilliant. But those things can and should be in the core functionality in the app. That would make it useful all around.

As for binding and baking I would say that it is a brilliant idea, but it is simply not up to par with where I think they want to go with Core. And unfortunately for some shots - not all - but for some shots, you'll need it if you are using IKB. This is why I devised a hybrid rig.

I learned a lot by studying IKB and I do recommend it for many reasons not the least of which it will raise the awareness in the LW community especially Hard Core people to influence the outcome of the CA system.

Also I think Collin's work on it is brilliant and his animation examples prove that you can do very nice work with it. So the last thing I would want to do is discount any of that because of my personal experiences.

ncr100
02-21-2009, 09:03 PM
...

Jays mentioned that NT wants to keep what makes LW what it is in Core. To me thats the speed and ease of use in getting things done which are very essential things especially in TV production work.

IKB fulfills that bill beautifully. Unfortunately I feel that theres really nobody at NT who has a good understanding of what IKB does or can do. The documentation on it makes this painfully obvious as well. This is based on various comments made about it by people from NT. Ive asked before what aspects of IKB they felt were worth preserving and never have gotten a response. I think theres a drive at NT to make LW, at least at first, into a Maya or XSI clone simply to make it "industry standard". So I expect that whatever CA tools they develop will be more like those in other apps rather then something innovative/fast/easy. I hope im wrong but Ive seen little to convince me otherwise.

...
Since Core is also a new app theres also the issue of simply convincing studios and end users to adopt it. The key thing I hear from coworkers is that the further NT gets away from that central LW ethic of being able to get things done quickly and easily the less likely people will move over to it.

As Colin pointed out before. NTs had IKB for how long? Theyve had the opportunity over the years to fix any issues, improve weaknesses, build up its strengths and so on. Even now people I work with who dont know LW are impressed with IKB because of its versatility, simplicity and speed. Its been an opportunity lost iMO but could be one again if the parts and principles that make it great are added into Core. In the end I dont care about IKB per se being in Core. I just want those aspects that make it powerful added into Core.

....


THIS is the kind of feature flow people pay for. If another free app offers the basic feature but not the flow then it's not got the same value. Nor worth paying for.

Splinegod, I hope NT is on top of replicating or evolving these flows in CORE. They do have the sourcecode, after all! ;)

SplineGod
02-21-2009, 09:58 PM
It doesnt matter to me that much about the actual IKB code. It was written at least as far back as LW7 or so. What Im more concerned about is its functionality and workflow. If this functionality and workflow can be replicated and improved upon in core LW would have a very fast, easy to use animation system.
Simply replicating the rigging and animating paradigms used in other apps isnt innovating. I do agree that some industry standardization should be done but having both isnt mutually exclusive. My concern about NT replicating the functionality of IKB is that Im not convinced anyone theres put the time into reallypoking around with it to get a good understanding of what it does.
I view IKB vs standard ways of rigging like I do layers vs nodes.
Layers are fast and easy and are what I end up using most of the time because of schedules etc. Nodes are great when you have the time and need the power. We have both and both should continue to be preserved.