PDA

View Full Version : I don't understand the LW road-map...



mikkelen
08-14-2013, 11:57 AM
LW 11 & LW 11.5 adds bullet physics and soft body dynamics. Great! However, hinges, ragdolls and more features needs to be added to make it ready for simulations in any serious VFX-pipeline, but nothing new in LW 11.6… why?

3DGFXStudios
08-14-2013, 12:09 PM
I also do not understand this. 11.6 is still in beta but when it comes out I hope they've finally added hinges and constraints.

The mesh filter is fixed in 11.6 :D.... that's something ;)

Ryan Roye
08-14-2013, 12:46 PM
LW 11 & LW 11.5 adds bullet physics and soft body dynamics. Great! However, hinges, ragdolls and more features needs to be added to make it ready for simulations in any serious VFX-pipeline, but nothing new in LW 11.6… why?

You can simulate ragdoll physics by pairing IKBooster with other dynamic motion (IKBooster takes care of the rotations, and normal dynamics takes care of actually moving the character). The disadvantages are that it can be slow to calculate and is certainly not as fluid or user friendly as ragdoll methods in other programs.

I too would like to see a more user friendly version of ragdoll physics put into LW though, as the IKBooster method is certainly not the most efficient.

tcoursey
08-14-2013, 01:13 PM
I think Rob made it quite clear at Siggraph that there attention and his personally has been on Chronoscuplt. He took it to NewTek many years ago and he is now able to bring it to us through resources at his disposal. (my opinion and guess based on comments at SG).

No doubt Chronoscupt is amazing and cutting edge. And personally I wouldn't care about having ragdolls or hinges (what are hinges...) But a clear road map would be helpful.

Is LW headed towards a unified app? Is OGL going to be updated in LW (chronosculpt like hydra whatever it was...) Etc..etc...

They can't of course tell us everything and a timeline because for one, that leaves no marketing or trade secrets (you don't want competitors to beat you to the punch). So what we ask for is almost impossible...

But I found it interesting there was no mention about 12, future developments etc... Just two products (one very pre-release [nevron motion]) that were ADD-On's and a basic bug / small feature release of 11.6

hrgiger
08-14-2013, 02:11 PM
No doubt Chronoscupt is amazing and cutting edge. And personally I wouldn't care about having ragdolls or hinges (what are hinges...) But a clear road map would be helpful.

Hinges are pretty much what they sound like. A hinge constraint between dynamic objects.


Is LW headed towards a unified app? Is OGL going to be updated in LW (chronosculpt like hydra whatever it was...) Etc..etc...

LW OGL is not the issue. Its an outdated geometry handler. Hydra is the new geometry engine inside of Chronosculpt. It has been suggested that this geometry engine will eventually make its way into LightWave.

As far as whether LW is headed towards a unified app... Nobody that knows is talking. It has been hinted at but that's not much to go on as we all know how things can change around here. It would be ludicrous to think they would not be headed in that direction but no progress is evident from LW 11 any more then it was from LW10. So we just have to wait and see.

tcoursey
08-14-2013, 02:15 PM
LW OGL is not the issue. Its an outdated geometry handler. Hydra is the new geometry engine inside of Chronosculpt. It has been suggested that this geometry engine will eventually make its way into LightWave.


So that seems promising I guess. However I would bet that since the geometry engine probably is tied to EVERYTHING it's quite an overhaul. We'll patiently wait and see.

hrgiger
08-14-2013, 02:49 PM
So that seems promising I guess. However I would bet that since the geometry engine probably is tied to EVERYTHING it's quite an overhaul. We'll patiently wait and see.

Actually, and I'm just sort of paraphrasing from the developer of Chronosculpt, the idea of the Hydra engine (and the rest of chronosculpt for that matter) is that it is a plug-in completely seperated from the rest of the application. The beauty being is that everything can be taken out seperately and completely replaced if desired without breaking anything else in the application. Of course putting it into LightWave is going to take quite a bit more doing then just dropping it in there since LW has its existing internal structure which has to be considered. Another advantage is that Hydra can be tested and improved in CS while not breaking anything in LW while the new technology is developed.

mikkelen
08-14-2013, 04:45 PM
I'm actually quite disappointed in LW11.6. I bought LW 11.5 for my in-house VFX-group, and really looked forward to using LightWave as our main 3d-application. However, I soon noticed that the Bullet system implementation was very basic. I guessed that finalising the implementation, by adding basic features such as hinges and constraints would be a top priority. Unfortunately it's a completely neglected next-step in LW 11.6. For me this is a show-stopper for LightWave, and we are looking for alternatives… I really thought that LightWave focused on virtual production, simulation and strong modelling tools these days… :(

pinkmouse
08-14-2013, 05:22 PM
Yeah, it would certainly be nice if Bulllet in LW was boosted to match some of the other implementations out there. But it would be even better if it was integrated with other tools like particles and flocking as well. At the moment, it just feels like a beta release.

prometheus
08-15-2013, 11:55 AM
I'm actually quite disappointed in LW11.6. I bought LW 11.5 for my in-house VFX-group, and really looked forward to using LightWave as our main 3d-application. However, I soon noticed that the Bullet system implementation was very basic. I guessed that finalising the implementation, by adding basic features such as hinges and constraints would be a top priority. Unfortunately it's a completely neglected next-step in LW 11.6. For me this is a show-stopper for LightWave, and we are looking for alternatives… I really thought that LightWave focused on virtual production, simulation and strong modelling tools these days… :(

I had a little of that disappointed lw 11.6 factor too, this time around I think it was one of the weakest updates for me personally, that is not to say that lw 12 might be awesome of course.
Lw 11.5 one of the best I think.

And yes I would gladely see better support for better wayīs to fix surfaces, points, or weightmaps etc and also hinges and constraints etc Inside of Lightwave bullet dynamics rather than chronosculpt.
Now ..this isnīt the way it turned out for this update..maybe it will later on, I still feel that setting up dynamics and a lot of other stuff is very fast and easy to do ..compared to letīs say houdini, houdini is more advanced though..but not as easy
to get going with.

I had hoped for particle improvements and intergration with bullet on par what the lux team recently did with modo and particles and recoil, and of topic a bit..hypervoxels improvements as they did in modo too.
on the other hand Ivé heard complaints on the luxology forums about recoil itself not being on par with Lightwave bullet, but thatīs rumours and not my own findings.

I just hope the lightwave dev team finds a good balance of working with the new chronosculpt and nevron motion as well as Lightwave, after all it will be a question of time and resources put in to those three elements where there
from the beginning was one element mostly..I would say. the time and resources taken from the dev team to develop chronosculpt and nevron motion shouldnīt have a negative impact on the lightwave road map..that is what worries me
a little, and since I didnīt see so much of updates in lw 11.6 ..that is what I thought, then again...It is just a smaller update, have to keep my calm and wait to see what Lw12 brings...if it still shows lack of development, I will start to scratch my
head a little and go huhmm..

But we will see...meanwhile I will keep an eye on where it is today on the road and not tomorrow, thatīs another day:)

Michael

VonBon
08-15-2013, 12:25 PM
:stop: 11.6 was "FREE"

some people just don't know how to be Appreciative.

prometheus
08-15-2013, 12:37 PM
:stop: 11.6 was "FREE"

some people just don't know how to be Appreciative.

yes..of course it was free, but thereīs been better updates even though being free...has nothing to to with appreciation and knowing that the lw team does a hard work behind the scenes, Im sure they do..and
im certain that in one of the following updates we will see further great updates, itīs
merely considering the facts of what impact this 11.5 update has for oneself, and of course somewhat the expectiations.
Wouldnīt call it to be non appreciative.

Michael

JohnMarchant
08-15-2013, 01:25 PM
I suppose you look at it differently if you just upgraded from 10.1. I decided after the core debacle that i was not going to update straight away until i could see how well 11 and 11.5 were received.

I waited some time and scoured the forums for specifically talk of the new features in LW and how well they have been implemented and how stable the were. I saw allot of positive talk especially about 11.5. So i had decided to wait to see what Sig2013 brought but had already decided i would upgrade.

I'm glad i did, yes there are many things that still need to be done, but after Core, im happy that NT picked themselves up and carried on.

So for me im very happy, however ive had to get Dan Ablan's videos and the plethora on youtube to get me up to speed. I was woefully behind the curve.

hrgiger
08-15-2013, 06:05 PM
I thought that the additions in 11.6 were fine and I think overall the LW11 cycle has been a good one.

At the same time, I am not expecting the kind of changes I would like to see in a point upgrade like 11.6. I'm very much waiting to see what kind of changes will be done for LW12 and I'm not talking about new features. Its about time LW3DG put a little focus on legacy issues.

Ryan Roye
08-15-2013, 06:17 PM
As long as audio scrubbing gets fixed at some point, count me a 100% satisfied customer.

CaptainMarlowe
08-16-2013, 12:01 AM
I thought that the additions in 11.6 were fine and I think overall the LW11 cycle has been a good one.

At the same time, I am not expecting the kind of changes I would like to see in a point upgrade like 11.6. I'm very much waiting to see what kind of changes will be done for LW12 and I'm not talking about new features. Its about time LW3DG put a little focus on legacy issues.

+1

JohnMarchant
08-16-2013, 12:26 AM
Yes agreed audio scrubbing and i really would like to see some more modeler TLC. Modeler i feel has been left behind a little bit. Yes there have been a few changes and stuff but i feel it has not really kept up with other 3d packages. I really would like to see the whole package tightened up and squash more bugs than any devastating new features.

tayotain2
08-16-2013, 03:52 AM
http://forums.newtek.com/showthread.php?136585-ChronoSculpt-from-LW-Group/page18

Good stuff from newbie3d. Lot to read also on next page. Made me very happy!

Cageman
08-16-2013, 04:08 AM
As long as audio scrubbing gets fixed at some point, count me a 100% satisfied customer.

Whats wrong with audio scrubbing?

mikkelen
08-16-2013, 06:29 AM
:stop: 11.6 was "FREE"

some people just don't know how to be Appreciative.

I would rather pay a hundred bucks or so to get a more flexible and feature rich bullet implementation…

I sincerly belive that Newtek should consider adding hinges and constraint to LW 11.6. It would help thremendously to stay with LW 11.x while we await LW 12, instead of moving to a different platform.

BTW: The one thing I really appreciate in LW 11.6 is multithreaded surface baker, however, that should have been implemented a long, long time ago… but I'm still very appreciative of this!

tcoursey
08-16-2013, 06:57 AM
They just mis marked it as 11.6 It would have all been fine if they called it what it was...11.5.2 And announced the LW12 feature list! We'd all been happy. (at least if the 12 list was a hit!)

mikkelen
08-16-2013, 08:33 AM
They just mis marked it as 11.6 It would have all been fine if they called it what it was...11.5.2 And announced the LW12 feature list! We'd all been happy. (at least if the 12 list was a hit!)

I agree, LW 11.6 would be a perfect LW 11.5.2. Progression in the Bullet area should be something every LW version should offer until it's on par with competing applications…

VonBon
08-16-2013, 08:41 AM
They just mis marked it as 11.6 It would have all been fine if they called it what it was...11.5.2 And announced the LW12 feature list! We'd all been happy. (at least if the 12 list was a hit!)

:question: uh.... whats the difference from "11.6 = FREE" and "11.5.2 = FREE"?

mikkelen
08-16-2013, 09:12 AM
:question: uh.... whats the difference from "11.6 = FREE" and "11.5.2 = FREE"?

Communication, it tells us something about ambitions...

jasonwestmas
08-16-2013, 09:20 AM
I'll be mostly watching Chronosculpt and fancy schpancy rendering additions. That's my roadmap.

dwburman
08-16-2013, 09:42 AM
They just mis marked it as 11.6 It would have all been fine if they called it what it was...11.5.2 And announced the LW12 feature list! We'd all been happy. (at least if the 12 list was a hit!)

I disagree. In.this cycle 11.0.x and 11.5.x have been mostly bug fixes with little functionality added. The difference between 11.0.2 and 11.0.3 was the ability to run dongle-free. 11.6 doesn't add a lot of big new features, but it adds several smallish ones and it's more than a bug fix/licensing change.

Of course, it does seem like adding a bit more into the Bullet implementation would be within the scope of a .1 update and I'd like to see that happen too.

:)

Ryan Roye
08-16-2013, 12:01 PM
Whats wrong with audio scrubbing?

In 9.6 and earlier, audio scrubbing worked without distortion or echoing when either using the mouse or arrow keys to move around the timeline.

I believe the problem started with version 10 and has yet to be fixed. Either that, or there is something fundementally different about the way LW 11 handles audio, and it is causing some audio devices to go haywire. Unfortunately our systems over here all have the same audio devices, so I can't test it without borrowing someone else's compy.

hrgiger
08-16-2013, 05:26 PM
Yeah I just tested a few diferent audio files and one doesn't even play the whole clip in the timeline and both were distorted so there's no way I could match that audio to anything happening in the scene, the scrubbing doesn't sound at all like the original audio file.

dwburman
08-16-2013, 06:17 PM
Audio seems to work here. It's not smooth, but I can scrub. I tried to compare it to LW9.6.1, but I couldn't get any sound out of LightWave at all then, so I can't compare the way they are different.

For scrubbing, it seems that fractional keyframes has some impact on the smoothness of it. The Fixed Frequency option gives you the skipping CD effect. I'm kind of used to that now from working with NLE video apps that use that technique.

It seems like they need some smoothing operation for dragging the playhead when scrubbing audio. It jumps around a lot from too slow to too fast.

I tested with an audio clip I saved specifically for the test, converting an audio clip from something I did earlier today. I saved it as an 8-bit, 22050Hz, mono wav file.

Davewriter
08-16-2013, 08:45 PM
So we have the fact that Siggraph only happens once a year. For those of us lucky enough to go, the FIRST thing we saw walking in the door was the LW booth. Making it probably the most expensive piece of real estate at the show. And for a show where the house is charging $150.00 + per quad box just for power, that is saying a lot. 11.6 showed up for us to download only a couple of hours before the doors opened, and I'm a guessing there were more than a few all nighters being pulled just to allow that to happen.

Okay, so the main thrust may have been to use this opportunity to show off two new products, but if nothing else we got a free upgrade. There were certainly demos there that I had never seen before, so there must be new stuff there somewhere.
Is it perfect or does it cover everyone's Christmas wish list? Grumbling hints at maybe no. But there is still a lot of time left in this year and I would be willing to bet that there will be some fine tuning done and things added by time this comes out of beta.

We tend to be a lot like the charactors from Brain Divided. Only in this cas half of us want all the new stuff... right now! Even while it is still dripping from a programer's brain pan - while the other half only want to see it when it is 100% crash tested.

jasonwestmas
08-16-2013, 09:23 PM
What new stuff are you talking about Dave? There's new stuff being added in LW all the time.

geo_n
08-16-2013, 10:51 PM
I'm pretty happy with lw 11.6 considering the price. Its solid and stable which is hard to do for a new release.

Snosrap
08-16-2013, 11:06 PM
The new color picker, faster VPR while moving geometry and raycast node = great free upgrade!

- - - Updated - - -

While I do agree that Bullet is a little light on features maybe there are underlying reasons why hinges etc have not been added - like the developer was assigned new tasks. :) Look at how long Hurley has been messing around with his physics thing - NT surely can't afford to have one guy totally consumed by such an effort.

Snosrap
08-16-2013, 11:09 PM
Maybe NT's roadmap has been revealed. Build a half-way decent app and let the 3rd parties round it out with more robust features.

geo_n
08-17-2013, 01:10 AM
Probably no roadmap. Just add tools that are outright useable and requested by lw based studios doing vfx.

Cageman
08-17-2013, 01:45 PM
For scrubbing, it seems that fractional keyframes has some impact on the smoothness of it. The Fixed Frequency option gives you the skipping CD effect.

Yeah... you should allways use Fixed Frequency. This is the standard way how Maya and Motionbuilder deals with sound. When I scrub back and forth, I can see/hear the sync for a lipsync animation, as an example.

Ryan Roye
08-17-2013, 01:50 PM
Yeah... you should allways use Fixed Frequency. This is the standard way how Maya and Motionbuilder deals with sound. When I scrub back and forth, I can see/hear the sync for a lipsync animation, as an example.

Agreed, though fixed frequency still yields no fewer than 3 echos (which often pass over multiple syllables in speech) with every press of the left/right buttons, and mouse scrubbing is inaccurate.

I have tried every possible audio format/setting in my files that I can think of and they all yield the same result. That narrows down the issue to either:

1) a certain type of sound device. I am using a generic realtek onboard audio. This would explain why a lot of people are getting it, and why others are not.

2) Lightwave audio handling (which unfortunately the user is powerless to fix).

mikkelen
08-17-2013, 04:08 PM
Conclusion - we would like:
- More Bullet features, i.e. constraints and hinges
- Audio scrubbing that works (if it works, it needs documentation)

May this become a reality before 11.6 goes gold?

jburford
08-17-2013, 05:39 PM
On my System, it is better with the Fixed Frequency turned Off. It is just too odd with it on and echos, bobbling of sound while scrubbing.

Using an M-Audio Delta 1010LT Card.

- - - Updated - - -


Conclusion - we would like:
- More Bullet features, i.e. constraints and hinges
- Audio scrubbing that works (if it works, it needs documentation)

May this become a reality before 11.6 goes gold?


NO! Don't look for it.

hrgiger
08-17-2013, 06:36 PM
I'm guessing that 11.6 will see pretty much bug fixes only before it goes gold. I think Nevron and Chronosculpt are more likely to see some added functionality before the end of the year. Maybe a Linux version for Chronosculpt, better Alembic support, stuff like that.

mikkelen
08-18-2013, 04:04 AM
I'm guessing that 11.6 will see pretty much bug fixes only before it goes gold. I think Nevron and Chronosculpt are more likely to see some added functionality before the end of the year. Maybe a Linux version for Chronosculpt, better Alembic support, stuff like that.

I guess you are right, unfortunately LightWave is not usable for my company until hinges and constraints are a part of the Bullet integration. I took it for granted that such functionality would exist in LW 11.5 - thats why I licensed it. If it's not added in 11.6, I really hope we'll see a Bullet focused update 11.7 rather soonish. I'm sure LW 12 will be great, as LW 11.5 is great, but they should take their time, and make sure that 11.x is as usable as it can be in the mean time.

jasonwestmas
08-18-2013, 12:57 PM
I'm not sure what the bullet hinges do that you can't do with other tools but constraints in LW bullet are handled using weight maps.

mikkelen
08-18-2013, 01:18 PM
I'm not sure what the bullet hinges do that you can't do with other tools but constraints in LW bullet are handled using weight maps.

How? It seems like bullet simulation is either on or off, and it's not possible to give different part of the object different properties. How can you make a safety net for instance with Bullet in LightWave; a net that are connected between steel pipes, that are again connected to a wall.

Skonk
08-18-2013, 01:27 PM
How? It seems like bullet simulation is either on or off, and it's not possible to give different part of the object different properties. How can you make a safety net for instance with Bullet in LightWave; a net that are connected between steel pipes, that are again connected to a wall.

Using a weight map, as Jason said in the quote in your post.

mikkelen
08-18-2013, 01:57 PM
Using a weight map, as Jason said in the quote in your post.

How? Is it possible to layer soft-bodies?

dwburman
08-18-2013, 02:21 PM
a little more detail...

make a weight map with the areas that you want to be most controlled by bullet at 100%. Back in Layout, in your Bullet properties for the object set as a deforming body, set the "mesh filter" to your weight map.

I don't know about layering, but I'm not exactly sure what you mean by that.

I put together a little test

http://youtu.be/yDsF5CvmmdA

Dialing in the parameters takes quite a bit of trial and error, and I had to set the weight/density of the ball object to a pretty low weight to keep it from going through the net.

The frame is there for decoration. It's really not needed for the dynamics.

mikkelen
08-18-2013, 04:26 PM
Your test looks like mine, so this I understand how to do :) But my net is mounted to a wall, with only two steel pipes stretching out in the free air, with the net between them. When a heavy mass lands in this net, the steel pipes should also start to bend, not like the net itself, but like metal.

dwburman
08-18-2013, 05:34 PM
Your test looks like mine, so this I understand how to do :) But my net is mounted to a wall, with only two steel pipes stretching out in the free air, with the net between them. When a heavy mass lands in this net, the steel pipes should also start to bend, not like the net itself, but like metal.

Could try varying degrees of weight map values (metal being like 10% instead of 100%), but I don't know off hand if that works or even if bullet sees gradations in the weight values. I haven't experimented with that. You might also be able to fake bending of the bars with bones or a morph.

mikkelen
08-18-2013, 05:46 PM
but I don't know off hand if that works or even if bullet sees gradations in the weight values.

It's either on or off. Thank you for your time anyways, LW3DG - we need hinges and constraints! :)

jasonwestmas
08-18-2013, 06:24 PM
There's a feature requester on fogbugz. They may not see your needs on this thread.

prometheus
08-18-2013, 06:27 PM
To me the hard surface and softbody combination and interaction still have a lot left too desire, I donīt think weightmap assignment on hardbody would be the proper way to go, using surface assignment might be.
Faking with bones and morphs we can do already, but thatīs not what we need, underlying surface structure affecting softbody and vice versa would be great, if possible.

Ik booster and bone dynamics works as a workaround too...
http://vimeo.com/26327972
how bullet softbody/hardbody is working with bones I really donīt know.

cloth fx was quite nice in some areas for deforming wakes and dent impacts, when I work with bullet I canīt replicate such results...yet, and it worked with particles too, but thatīs a different story.

prometheus
08-18-2013, 06:42 PM
Maybe I need to check the manual more closely, Ivé just discovered that the softbody actually sticks and holds a hardbody if it is close by or intersecting the hardbody part.

jasonwestmas
08-18-2013, 06:43 PM
Well I do know that bullet fully reacts to any geometry that is being driven by bones and IK. That was greatly improved in 11.5.

Dodgy
08-18-2013, 09:15 PM
Do you have an example of that Prom? As far as I know, the only way to make a hardFX interact with a Cloth/SoftFx is putting CollisionFX on the clothFX/SoftFX object as well..

mikkelen
08-19-2013, 03:45 AM
It would be nice if LW3DG could chip in… I'm about to pull the trigger on a dozen Modo or LightWave seats. The problem with Modo is that I don't know how to use it, and I'm not going to learn it, so I cannot step in if my co-workers need extra help. The problem with LightWave is the marginal Bullet tool-set.

jasonwestmas
08-19-2013, 07:30 AM
It would be nice if LW3DG could chip in… I'm about to pull the trigger on a dozen Modo or LightWave seats. The problem with Modo is that I don't know how to use it, and I'm not going to learn it, so I cannot step in if my co-workers need extra help. The problem with LightWave is the marginal Bullet tool-set.

Actually, the easiest answer to your problem is LWBullet and Chronosculpt. Just wish there was a demo of CS for you.

mikkelen
08-19-2013, 07:56 AM
Actually, the easiest answer to your problem is LWBullet and Chronosculpt. Just wish there was a demo of CS for you.

Why would I want to paint a simulation, and not animated it - if simulation is not possible in a given software?

jasonwestmas
08-19-2013, 08:02 AM
Why would I want to paint a simulation, and not animated it - if simulation is not possible in a given software?

The idea is to sculpt on top of a cached sim to make it look exactly the way you want without re-simming a hundred times.

mikkelen
08-19-2013, 08:25 AM
The idea is to sculpt on top of a cached sim to make it look exactly the way you want without re-simming a hundred times.

But in the given scenario, LightWave is not able to deliver the basic simulation… so Chronosculpt would be of no use.

jasonwestmas
08-19-2013, 08:35 AM
But in the given scenario, LightWave is not able to deliver the basic simulation… so Chronosculpt would be of no use.

I disagree. But even so, you don't necessarily have to simulate every vertex for CS to be useful. You can simulate sections and then complete the look in CS.

hrgiger
08-19-2013, 10:21 AM
im sorry mikkelen im on mobile right now so its hard to read the whole thread but did you already mention what kinds of things you want to use bullet for so we can better understand what parts of bullet you feel are lacking? and to be clear i know it still has a ways to go im just trying to figure out what kinds of effects you want to achieve with it.

Emmanuel
08-19-2013, 10:21 AM
Maybe NT's roadmap has been revealed. Build a half-way decent app and let the 3rd parties round it out with more robust features.

No, thats 3ds max.

mikkelen
08-19-2013, 10:35 AM
I disagree.

Then you don't understand the scenario we've been discussing.

jasonwestmas
08-19-2013, 10:45 AM
Then you don't understand the scenario we've been discussing.

I doubt it. Of course maybe you didn't tell us the whole story.

mikkelen
08-19-2013, 11:02 AM
I doubt it. Of course maybe you didn't tell us the whole story.

Read again, others seems to have understood the scenario.

jasonwestmas
08-19-2013, 11:09 AM
Read again, others seems to have understood the scenario.

I didn't say I misunderstood. You're looking for a way to bend metal while a net deforms.

You'll get better results if you start a thread somewhere else and layout everything you are doing in LW. It's clear you are looking for answers, no? Or would you rather just say it can't be done.

mikkelen
08-19-2013, 11:15 AM
You'll get better results if you start a thread somewhere else and layout everything you are doing in LW. It's clear you are looking for answers, no? Or would you rather just say it can't be done.

I can animate it with bones and IK's, but thats really counterproductive, this can be simulated in other packages. I used this scenario as an example.

jwiede
08-19-2013, 01:04 PM
Look guys, like it or not, there are a bunch of things that cannot currently be simulated in LW's Bullet due to the absence of hinges/connectors, motors, and so forth, all of which are fairly trivial to do in fully-implemented Bullet integrations. Try putting an active suspension on a cart/vehicle, if you want a "real-world" example. That's a very common usage scenario for me (and quite doable in C4D, modo, etc. in their Bullet integrations), yet completely unworkable in LW's Bullet.

Weightmaps do not solve many of those issues, the issues are about how different objects' interactions are constrained (hence constraints) in 3space, and most such relationships cannot be described using just a scalar weightmap of Bullet's action or of a characteristic (for example, consider the behavior a "one-axis slider" connector between two dynamics-driven objects). Those primitives are part of Bullet's "toolset" for a reason, they're common and necessary to describe all sorts of real-world dynamics situations that Bullet users required. It's not like they were added to Bullet (nor integrated into so many other Bullet integrations) "just for laffs".

VonBon
08-19-2013, 02:40 PM
awe :grumpy:, yall gone make me look into Bullet more closely.

Is there a way to use layers to help achieve what your looking for?

jasonwestmas
08-19-2013, 03:33 PM
I don't think anyone here is saying LW's dynamics are anywhere near perfect or able to compete totally with other implementations.

pinkmouse
08-19-2013, 03:36 PM
I recently spent a good few hours inspired by the "Blow up a Car" thread, using Bullet to create a car door deforming after impact by an object and, frankly, got nowhere. Then I saw this video (http://vimeo.com/24733760) using C4D, and was shocked at how straightforward it would have been with a proper implementation. And no, Weight Maps didn't help. ;)

116503

mikkelen
08-19-2013, 03:36 PM
…and then we are getting back to the topic, that I don't understand the LightWave road-map, they've stared to implement a good dynamics system. Hard-bodies in LW11, soft-bodies in LW11.5, though very basic, but LW11.6 marks an end to the progress… thats rather confusing and discomforting...

jasonwestmas
08-19-2013, 03:50 PM
…and then we are getting back to the topic, that I don't understand the LightWave road-map, they've stared to implement a good dynamics system. Hard-bodies in LW11, soft-bodies in LW11.5, though very basic, but LW11.6 marks an end to the progress… thats rather confusing and discomforting...

I never buy software if it doesn' t do what I want it to do right now. So if you're not comfortable now, then my advice is to use something else.

I mean who knows why NT does what they do. "Unforseen complications" has been LW classic's past. It does a lot of things wonderfully but it did those things very early in the game. That has been the nature of using this software especially in an animation and dynamics context.

mikkelen
08-19-2013, 04:21 PM
So if you're not comfortable now, then my advice is to use something else.

Yes, you are right. Modo next.

jasonwestmas
08-19-2013, 04:25 PM
Yes, you are right. Modo next.

You can always render your sims back in Lightwave when 11.6 gets alembic. ;)

jasonwestmas
08-19-2013, 04:30 PM
Oh and just to be fair, Modo has its fair share of half baked tools too, so LW won't feel too lonely. :)

Surrealist.
08-19-2013, 05:09 PM
I am not surprised at all about LightWave development. They are just improvising based on a general idea of where they are going and it is clear that the main theme is generating money with features that can be released now. Bullet improvements would be great. If they are interested in LW Bullet to be taken seriously, they'll do it.

But as Jason said, get LightWave for what it can do now. Not what is promised and that is true with anything.

The best dynamics solutions in my opinion are in ICE and Houdini if that is what you need. Even XSI dynamics has hinges that work and easy to set up. ICE is out of control and Houdini is on another plane of existence. To me those are the solutions to be looking at. Modo is still too full of promise in its current state for me. I prefer things that work now.

geo_n
08-19-2013, 07:05 PM
Oh and just to be fair, Modo has its fair share of half baked tools too, so LW won't feel too lonely. :)

I heard from modo users and reading through modo forum that modo 6 and 7 are so buggy and always crashing, too. I'm only learning it right now so less crashy.
Lightwave has been really stable for me since lw 9 and that's something really critical during crunch time.

Snosrap
08-19-2013, 07:42 PM
…and then we are getting back to the topic, that I don't understand the LightWave road-map, they've stared to implement a good dynamics system. Hard-bodies in LW11, soft-bodies in LW11.5, though very basic, but LW11.6 marks an end to the progress… thats rather confusing and discomforting...I think it's just a matter of manpower - the team is small. Just look at how long Hurley has been working on his IBounce/Advance Placement thingy- it takes a long time to develop this stuff and the guy who wrote the LW Bullet implementation may also be responsible for a dozen or more aspects of LW. Plus throw in the fact that they are trying to develop a new code base at the same time and every new feature that they build most likely needs to work with the current code base as well as whats being developed - it probably can get pretty crazy.

jasonwestmas
08-19-2013, 08:58 PM
I heard from modo users and reading through modo forum that modo 6 and 7 are so buggy and always crashing, too. I'm only learning it right now so less crashy.
Lightwave has been really stable for me since lw 9 and that's something really critical during crunch time.

Yes, I find LW to be one of those predictable pieces of software. I don't often find myself wondering why something isn't working I can usually trouble shoot it quickly. One of the big reasons why LW people keep using it I think. Crashes are usually predictable as well, less random I find.

pinkmouse
08-20-2013, 12:42 AM
Would be nice if the developer involved in Bullet could pop in here and tell us what's happening, like David did in the CS thread.

Hail
08-20-2013, 03:18 AM
Would be nice if the developer involved in Bullet could pop in here and tell us what's happening, like David did in the CS thread.

+1!

mikkelen
08-20-2013, 07:38 AM
Would be nice if the developer involved in Bullet could pop in here and tell us what's happening, like David did in the CS thread.

+1

prometheus
08-20-2013, 09:29 AM
yes..that is worrying from the Modo perspective, from Lightwave 11.03 and up itīs been quite stable for me, Lw 11.6 might be a little more buggier but havenīt used it in that extent that I can say for sure.
I do know that it often crashes when working with VPR and procedural texturing, mostly when switching between procedural layers and also changing fractal typeīs in some cases, and very very often with texture displacement layers.

Ivé also got some Freezing of the system when I tweak some bullet parameters on linear damping, at 100% or more and some others of those parameters, seems to start more complex calculations and then I canīt escape the simulation or stop it at all except for
going to task manager and brute force stop it.
Might be considered quite small issues ..and at the same time frustrating if you work with those areas a lot.

If I can nail down more accuratly exactly when it happens ..I should of course bug report it.

Michael

mikkelen
08-22-2013, 03:25 AM
Would be greatly appreciated if someone LW3DG could shed some light on Bullet & LightWave...

bobakabob
08-22-2013, 04:24 PM
I do know that it often crashes when working with VPR and procedural texturing, mostly when switching between procedural layers and also changing fractal typeīs in some cases, and very very often with texture displacement layers.

If I can nail down more accuratly exactly when it happens ..I should of course bug report it.

Michael

Michael,
I've noticed this too on occasions, nothing drastic but there's something not quite 100% about changing procedurals and VPR. They may be DP textures, will try and replicate when I have time.

mikkelen
08-26-2013, 08:10 AM
I've begun the process of testing other tools. It's pretty obvious that the Bullet implementation in LW11.5 is quite narrow compared to other packages, why do Newtek add features with so little depth and flexibility?

Emmanuel
08-26-2013, 08:23 AM
Because :)
Speaking of the roadmap, its obvious, isnt it : 11 got Bullet with hard bodies, 11.5 got softbodies,12.0 gets hinges etc. Little by little. Works for me, as I love GoZ and LWtoAE and the new modeling tools and vpr updates. Might not work for you, such is life.

prometheus
08-26-2013, 08:50 AM
I've begun the process of testing other tools. It's pretty obvious that the Bullet implementation in LW11.5 is quite narrow compared to other packages, why do Newtek add features with so little depth and flexibility?

I think we need to have patience here, sure I felt a little like..what the heck..chronosculpt and nevron motion development focus!! but thatīs the way
they decided to work at for the moment, I can only hope that the chronosculpt and nevron motion modules wonīt interfere too much with further development of bullet and particles amongst others..but that
can only future updates tell for us when it arrives...unless some developer jumps in and talks about that roadmap, which is not to be expected as common.

You could for instance try houdini out and see how easy it is to setup compared to Lightwave, I actually feel itīs much much easier to get going with such stuff in Lightwave, though some limits.
and you could try modo and see if recoil works better.

Michael

JohnMarchant
08-26-2013, 12:08 PM
Houdini Apprentice for $99 per year is a steal if you want to learn Houdini.

mikkelen
08-27-2013, 05:57 AM
Houdini Apprentice for $99 per year is a steal if you want to learn Houdini.

Sounds good, but we deliver for commercials and short fiction.

JohnMarchant
08-27-2013, 07:07 AM
Well then Houdini is probably more than you need, depending on the commercials. However if you wanted to learn a powerful fully featured 3D Software then this would be my recommendation.

mikkelen
09-01-2013, 12:25 PM
Thank you, maybe we should test it!

JohnMarchant
09-01-2013, 12:46 PM
Prepare for s steep learning curve. Like Maya, if you have any programmers get them scripting, that's the way to really get the most out of this software. Fantastic Particle,Dynamic and cloud system