PDA

View Full Version : New game development tools in 8 ?



Emmanuel
04-23-2004, 05:47 AM
Hi !

Any of the old NT promises kept this time ?
Any new game relevant tools/exporters/bridges ?

TriXal
04-23-2004, 10:04 AM
I would like to see smoothing groups in LW.
Most important game development feature in my opinion.

tudor
04-23-2004, 10:13 AM
Smoothing groups is like surfaces..
Imo smoothing groups are a pain.
Hard/soft edges like in Maya though! That would be great.

ghopper
04-23-2004, 11:57 AM
LW8 should really support games developers, but I don't think this is a high priority for NT at the moment, they have so many other things to improve.

For example, I couldn't find any info on whether Valve will release
Half Life 2 exporters for LW. Only found some info about exporters for Softimage, Maya and Max.

HalfLife 2 modding is gonna be massive and fortunately there is the XSI EXP version - smart move by Avid really ;)

But I still hope we will see a HL2 exporter for LW.

fortress
04-23-2004, 08:06 PM
value has no plans of releaseing a lw exporter

i have already written a static prop exporter the animation is a different story

it should be pretty simple for a better coder there file format is all text based so easy to figure out

accually it is almost the same as hl1

TriXal
04-24-2004, 03:09 AM
Originally posted by tudor
Smoothing groups is like surfaces..
Imo smoothing groups are a pain.
Hard/soft edges like in Maya though! That would be great.

I've used LW over 4 years and I can tell you that smoothing groups are nothing like surfaces.
Currently there's no fast and good way to do similar thing in LW.
You can always unweld vertices but who wants to do that?
I am a game developer so I know what i'm talking about.
At work I model and map my objects in LW and then I throw them to MAX for finishing.
I'd be very happy if I wouldn't have to go MAX at all.
Pivot tools is also something that I'd love to see in LW.
Oh, I almost forgot ability to manipulate hidden edges.

jin choung
04-24-2004, 03:51 AM
actually,

between max, maya and lw, maya definitely has the correct solution to the 'smoothing groups' issue.

just selecting certain edges and tagging them 'smooth' or 'hard' is much better than extrapolating edges from 'smoothing groups'. it's much more explicit and direct to deal with the edges in question.

and we ABSO-fing-LUTELY have to remove the 'smooth' from the surfaces! it is NOT rightly a material property! it should be externalized like in maya (perhaps making it an OBJECT PROPERTY) and let us deal directly with the edges.

lw can be a fing pain in the [email protected]# because it is is [email protected]#$ literal.

we have to WELD/UNWELD to edit UVs because that is what's really happening to the mesh. other apps ABSTRACT that away from the user for ease of use sake but lw doesn't.

same thing with hard/soft edges. in lw, they force us to UNWELD because in order to interrupt the gourad smooth shading, the object REALLY IS GETTING UNWELDED and creating more verts. but again, this is abstracted away from the user in other apps but we have to deal with it at the base level.

which is a pain in the ***** when you consider that one of the big clean up steps before exporting to a game engine (for me) is merging everything and see if you have any anomalous verts or junk geometry floating around.

[in a related note, the process of bone weight mapping is excessively and needlessly exposed to the user at a low level in lw. some things need to be re-examined and when it makes sense, abstracted away from the user.]

jin

p.s. (this is from a rare great article from GAME DEVELOPER MAGAZINE) if you have two materials on a single object, you are creating VERTEX SPLITS (unwelded verts) on the actual geometry. whenever you create discontinuous UVs you are creating VERT SPLITS. and when you do hard/soft edges, you are creating VERTEX SPLITS in the mesh in actuality to achieve that effect.... it would be nice if these vert splits are abstracted from user intervention in lw....

BUT it would exceedingly NICE to see a FINAL VERT COUNT in the 'w' statistics panel that gives you a final geometry count that shows you true, runtime totals.

jin choung
04-24-2004, 03:58 AM
oh,

and seriously guys... you gotta stop asking for newtek to write game exporters for lw!

that is well and beyond the purview of newtek's responsibility.

if the game developers themselves create exporters for your app, that's one thing but to cry foul when your 3d app company does not is something else.

i agree that it would be nice if they had a team of game dev guys who did nothing else but program in support for the more popular game engines (unreal, halflife, hl2, quake3/wolfenstein/callofduty, doom3, etc) but considering that newtek is the smallest of the 'professional' 3d apps, i don't think we're EVER gonna see that happen.

if it happens at all, it will be because of some gamer guy in the community with 'madskills (tm)' and a big heart or through another app:

SPEAKING OF WHICH

if anybody needs a translator to game engines, for heaven's sake, JUST GET MILKSHAPE3D from chumbalumsoft. $25 that's WELL WORTH IT if you use it for nothing else except a simple exporter app.

seriously. it's no use asking for game translators here. it will not happen.

ask for other necessary things like the hard/soft edges, real edge support and better uv handling and maybe dx9/ogl2 level features in RT redraws, but game specific exporters is just a lost cause.

jin

mkiii
04-24-2004, 03:59 AM
Me too. I've been using LW to create games content now for
around 6 years or so, and have always had to export the LW object, to 3ds at first, and now to Max, so I can export it using our in house plugins which are for Max only.

I am a one off indulgence. Everyone else here is using Max, and in a few cases, Maya.

Having a smoothing groups option would be nice, but it wouldn't make any difference to me, unless 3dExploration (my converter of choice) could read them, and pass the data over to Max. Same goes for the hidden edge manips.

Obviously, a 3ds file exported from LW would not contain this data.

The only solution IMO is to brainwash all Coders, and convince them that the LW SDK is far easier to write tools for than the Max one, and to convince our bosses that Max, Maya & XSI are not the only apps that can be used for game dev.

Brain rays at the ready. Over to you Newtek.

PS what jin said...... but without the expletives :)

rebelr6
04-24-2004, 04:04 AM
Originally posted by TriXal
I've used LW over 4 years and I can tell you that smoothing groups are nothing like surfaces.
Currently there's no fast and good way to do similar thing in LW.
You can always unweld vertices but who wants to do that?
I am a game developer so I know what i'm talking about.
At work I model and map my objects in LW and then I throw them to MAX for finishing.
I'd be very happy if I wouldn't have to go MAX at all.
Pivot tools is also something that I'd love to see in LW.
Oh, I almost forgot ability to manipulate hidden edges.

I'm also a game developer who has just purchased the 7.5 duo and patiently awaiting 8 (I am in the UK - Scotland).

I don't see the problem with unwelding vertices to finish off a model. Actually it's easier from a programming perspective if there isn't smoothing groups as you don't have to manually create unified arrays of vertices and detect which vertices are shared and which aren't - you just get Lightwave's array of vertices and output that.

In my view, the issue that concerns game developers most is not the exporting of models from Lightwave, but the lack of any real level building tools and game specific entity properties editing. I am converting my old plugin (from another 3d program) to Lightwave to allow me to define game entity properties, assign paths to entities, and save out a "level" file containing all this data. I may make this plugin open-source, but I'll have to think about whether it will be too specific to my game to do that.

There is (I believe) an easy way to do this in the API by using the "setTag/getTag" on items in Layout. You can associate a string tag with an item, so that is probably the route I'll go down when it comes to implementing these properties.

jin choung
04-24-2004, 04:07 AM
Originally posted by mkiii
I am a one off indulgence. Everyone else here is using Max, and in a few cases, Maya.

Having a smoothing groups option would be nice, but it wouldn't make any difference to me, unless 3dExploration (my converter of choice) could read them, and pass the data over to Max. Same goes for the hidden edge manips.

Obviously, a 3ds file exported from LW would not contain this data.


me too... i take my lw to model and texture, then export using 3d exploration to .obj (our exporters are from inside maya).

inside of maya, i have to tweak the hell out of all the material properties and re-assign textures and then do the hard/soft edges thing.

then export.

your second point about 3d exploration not supporting any changes in the lw format (at least initially) is a great point and i think it's one of the reasons that they may be treading lightly in changing the object format to include TRUE EDGE SUPPORT, UVs (as opposed to everything [even in the uv view] being verts), SMOOTHING being taken out of the surface properties and enabling us to tag edges as hard/soft (which is essentially a boolean 'EDGE MAP').

actually, i think that's giving them too much credit. honestly, i don't think that they've even thought these issues through.

but if I were developing it, this would be one of the big reasons that would give me pause in changing the object format: compatibility with 3rd apps and translators like 3de and zbrush and deep paint 3d....

alas.

jin

tudor
04-24-2004, 04:24 AM
Writing an exporter from LW to whichever program/game, is much easier in LW than any other program. Why? Well, wysiwyg. No history, no unessecary crap. It is not just me as an artist saying this. I have coders at work who are activly trying to get us to use LW as our main artist tool just because of this. We have a couple of guys at work who are making their own game engine which we might use in our coming games that relies on LW. It took them a couple of days to write an exporter that supports just about all that you need including bones, morphs, all surface attributes etc. LW's file format is just so easy and straight forward.

Hard soft edges.. Well.. The game engines treat them as unwelded points anyways, so we can already do them quite easily..

Emmanuel
04-24-2004, 06:11 AM
I was aiming at things that were told would come out of working closely with developers like Digital Extremes, probably the Croteam etc.
I mean, if You have a look at the "planned" exporter stuff for the NetImerse engine, Unreal etc that comes with 7.5c, You'd think that Nt planned to create really cool, universal exporters for engines.
If You really wait for the game developers, You better give up and change the app, because the easy bridge between Max and Maya and engines seems impossible for LW, despite the fact that people claim LW is EASIER to develop exporters for.
Tell that to Criterion (renderware), Crytek (CryEngine), Lithtech (Jupiter) etc.
The renderware engine is probably the most used one in the industry, only followed by unreal I guess.
NT *must* make it more attractive to use LW in a game development pipeline, and the only way I see is by giving developers some advance.
Let's face it, modeler is getting old by now, has no edges, no smoothing groups, no engine-usable tri preview, no advanced OpenGL display for normal maps, lightmaps, anything.
You can get Maya for just a few bucks more that is much more complete for gamestuff than LW, and supported by every game engine on the planet.
And YES, game engines do come with exporters for Max and maya, and I guess its because Alias was aggressively charming these engine developers, well enough to become a substitue for Max, which practically owns that place.NT was never able to get as far as Maya, and I guess the reason must that Maya is easier to integrate into pipelines, and NT got to ask themselves "why?".
The future is networked, so if NT really think they can go on with their
isolation from 3rd party render engines and game engines, then they will find themselves even more "outside" than they already are.
If NT aren't interested, fine, but they should stop teasing me with all the same promises release after release.

tudor
04-24-2004, 06:28 AM
It is quite sad, but thats the way it is.

There are no exporters for LW because almost noone uses LW to make games. Almost noone uses LW to make games because there are no exporters.

It is so sad, because Maya for example has a horrible api when it comes to exporting to game engines. At work we have been suffering for over three years because of NetImmerse and its poor exorting from Maya.

How much money is there really in games? Most game companies suffer right now. Overall games pull in a lot of money but 3/4 of all companies loose money instead of making a profit.
Hopefully I will get out of making games soon. Got an offer as Lead animator on a swedish CG film :)

TriXal
04-24-2004, 07:28 AM
Originally posted by rebelr6
I'm also a game developer who has just purchased the 7.5 duo and patiently awaiting 8 (I am in the UK - Scotland).

I don't see the problem with unwelding vertices to finish off a model. Actually it's easier from a programming perspective if there isn't smoothing groups as you don't have to manually create unified arrays of vertices and detect which vertices are shared and which aren't - you just get Lightwave's array of vertices and output that.


That might be but not with complex models.
I model cars which consists of thousands of polygons and many layers.
Usually when I finish editing or adding something I weld every vertices and then I use unify to check that there's no polygons I don't want. So if I have already made my "smoothing groups" they are gone after my welding operations and I have to do them again.
And I don't want to worry all the time that I accidentally screw up my "smoothing groups".

Castius
04-24-2004, 07:38 AM
"There are no exporters for LW because almost noone uses LW to make games."

IMO i don't think this is true.

Although i will say i think the reason we see max and maya exporters for engines is because of some of the stuff you guys have been talking about. Max and Maya make it easyer to create theses options and to the people making the engines it is easier working with Max and Maya to test there engines. Hense they all release there exporters with there engines.

Modelers style has alwasy been to make things open and exsposed. This is why we have to to clean up 2point polys and unweld to creat discontinuse UVs. I prefer this in most cases. I have resenty found out that LWO format supports smoothing groups though tags. Aparently it's just never been used in the interface yet.

I would like to see an a extra user layer control in LW that would make these things transparnet when there isn't a need to deal with them. Like if you switch to bigginer mode these items become easyer to deal with, edges in quads/ngons could be edited. UVs would not need to be unwelded, and smoothing could be controler easyer. switch back to expert mode and we can go back to workgin on base level eliments. What do you guys think?


We are in the prossec of writing our new exporter for nebula 2. personly the only thing i hope is easyer is exporting animation. Our exporter for Nebual 1 have been open to the public in lscipt for a long time and are very easy to modify. Nebula 2 exporter will be fallowing the same path.

tudor
04-24-2004, 07:47 AM
I got a neat little trick for you :)

The points that make up the sharp edge, ie. the point that will get welded if you merge points. Select those points and make a tiny tiny jitter.. 0.001mm or so. This keeps them from merging, but does not show up as any rendering artifact.


Originally posted by TriXal
That might be but not with complex models.
I model cars which consists of thousands of polygons and many layers.
Usually when I finish editing or adding something I weld every vertices and then I use unify to check that there's no polygons I don't want. So if I have already made my "smoothing groups" they are gone after my welding operations and I have to do them again.
And I don't want to worry all the time that I accidentally screw up my "smoothing groups".

TriXal
04-24-2004, 08:10 AM
Originally posted by tudor
I got a neat little trick for you :)

The points that make up the sharp edge, ie. the point that will get welded if you merge points. Select those points and make a tiny tiny jitter.. 0.001mm or so. This keeps them from merging, but does not show up as any rendering artifact.

Thanks for a good trick but I think I will stick with MAX when it comes to smoothing groups... for now :)

Castius
04-24-2004, 11:02 AM
tuder take that metod alittle futher. For alittle cleaner system I creat a morph called seams. then i just move points apart i don't wanted welded when i merge. the best part is this can be undone at any time.

pauhana
04-24-2004, 11:21 AM
Originally posted by tudor

How much money is there really in games? Most game companies suffer right now. Overall games pull in a lot of money but 3/4 of all companies loose money instead of making a profit.
Hopefully I will get out of making games soon. Got an offer as Lead animator on a swedish CG film :)


I respectfully disagree with this statement. There are a lot of companies making money in this industry. Most of the companies that fold do so because they put out crappy games or they don't have follow up/overlapping projects to carry on. If Newtek were in demand for game development they could sell at least 20-40 licenses per company (Independent game developers) and much more for large developers. Also, with solid, functional export tools for engines like Netimmerse and Renderware (to name a couple) they could create really strong partnerships. This may not be Newtek's goal, but it is feasible.

Peace...

jin choung
04-24-2004, 08:54 PM
actually,

don't forget the major reason why there are so many exporters for max and maya-

the developers USE these apps to make their games. they're not doing anyone a favor in developing these exporters... it's part of their work.

and so, i say again, that it is unreasonable to ask anyone to make exporters if it's not part of their bread and butter interest.

jin

pauhana
04-24-2004, 10:07 PM
Originally posted by jin choung
actually,

don't forget the major reason why there are so many exporters for max and maya-

the developers USE these apps to make their games. they're not doing anyone a favor in developing these exporters... it's part of their work.

and so, i say again, that it is unreasonable to ask anyone to make exporters if it's not part of their bread and butter interest.

jin

That doesn't make sense. The reason developers USE those apps is because there is game support for exporters. Do you have experience in the game industry? I don't think you fully understand the pipeline.

pauhana
04-24-2004, 10:36 PM
jin choung,
Scratch that last part of my reply...that was rude of me. Doesn't really matter if I would like game support for Lightwave it's up to Newtek to foster those relationships. I just wanted to be able to use Lightwave for more than "personal projects".

Peace...

Adrian Lopez
04-24-2004, 10:57 PM
Originally posted by pauhana
That doesn't make sense. The reason developers USE those apps is because there is game support for exporters. Do you have experience in the game industry? I don't think you fully understand the pipeline. Sorry, but that doesn't make any sense. Compared to all the work that goes into programming a modern game, writing an exporter for a game engine is a trivial task. Of all the technical and artistic reasons why developers might choose a particular modeling and animation package, built-in exporters are surely at the very bottom of the list (right along with other "it would be nice if we had this" items). Frankly, it would be utter madness to choose a modeling/animation package according to such a peripheral feature as a ready-made exporter for your chosen engine.

pauhana
04-25-2004, 12:04 AM
Yes, but with all that work going into modifying the engine to suit your games needs there needs to be initial support for the geometry, textures, and animation. The engine we're using has stronger (animation) support for another package than the one we use. I've been having to trouble shoot a lot of things they hadn't encountered with exporting from the other package, especially since we are using more complex rigs than other studios have used with this exporter. We will ultimately have to rewrite the exporter to suit our needs, but we couldn't hit our milestones if nothing was there.

jin choung
04-25-2004, 01:07 AM
Originally posted by pauhana
jin choung,
Scratch that last part of my reply...that was rude of me.


yes it was.

gotta say that i'm happy for both of us. if you track my behavior on the boards, i am very careful not to start trouble of a personal nature.

but boy, do i have a very unchristian knack of refusing to turn the other cheek and going for no holds barred smackdown when insulted.

but good. there will be peace then.

------------------------------------------------------------------------------------

FIRST:

darn the luck... i just happen to WORK PROFESSIONALLY IN GAMING.

SECOND:

i guarantee you that half life 2 developer VALVE was not waiting for xsi to make exporters for their game. same thing for doom3, same thing for unreal tournament.

and same thing for my company.

WHATEVER EXPORT SUPPORT there exists for these games was coded by the GAME DEVELOPER THEMSELVES!

and so i say again,

THERE ARE AS MANY EXPORTERS FOR MAX AND MAYA AS THERE ARE BECAUSE THEY ARE DEVELOPED BY THE GAME DEVELOPERS.

------------------------------------------------------------------------------------

if we're talking about third party game engines like NETIMMERSE, GAMEBRYO, RENDERWARE, all the exporters are created by the maker of these engines.

if support for lw does not exist in these engines, it is a testament to the demands of the development community.
------------------------------------------------------------------------------------

ESPECIALLY if we're talking about PROFESSIONAL (paid, funded, money making) game development, the burden of creating exporters from a 3d app to the game engine lies SQUARELY on the shoulders of the game developer.

if your programmers cannot develop an acceptable exporter, that is THEIR problem and that is their inability. consider firing.

the serious sam folks were able to do it so it's not a problem with the sdk - so again, it's not rightly newtek's problem.

it would be NICE. but you can't fault them.

jin

pauhana
04-25-2004, 01:38 AM
"if we're talking about third party game engines like NETIMMERSE, GAMEBRYO, RENDERWARE, all the exporters are created by the maker of these engines.

if support for lw does not exist in these engines, it is a testament to the demands of the development community."


That's what I was refering to. We develop for the XBox, PS2, Game Cube and having "basic" export support early on in development is pretty important when choosing a game engine. Publisher is dropping millions of dollars, so we'd better hit our milestones.

pauhana
04-25-2004, 01:51 AM
Jin,
All that crap aside. I noticed that you posted elsewhere about Maya's IK strengths and how it would be nice to have some of those features in Lightwave. I'm guessing you use it at work for game development. I've done a lot of character work with it and feel the same way. Why do you think they choose to make the animation tools so different from Maya & Max? I feel that gimmicky character tools can never replace solid, simple workflow.

jin choung
04-25-2004, 02:23 AM
yup,

i use lw (my own personal copy that i take into work) for modeling and uv mapping but all of our games exporters were programmed for maya.

so all the game assets have to be finalized in maya and i do all animations from inside of maya.

as for why newtek does not simply drop pretention and JUST COPY maya's ik system...

man, you got me....

it is in my mind, beyond insanity, to 'upgrade the ik' with ikb when so far, it's lookin' like ikb can't even be a good substitute for the current lw ik.

it seems unlikely but perhaps maya's ik is proprietary gold that is so super secret and inaccessible that it is completely contrary to what everyone else knows about ik in the industry.

???

either they don't have the know-how, they don't have the money, both or maybe they really like it the way that it is... some unsettling opinions (imo anyway) seem to indicate that that might indeed be the case.

jin

jin choung
04-25-2004, 02:32 AM
oh,

and as for the situation at your company....

well, ya might want to consider using ANOTHER 3D APP for game development then!

too bad for newtek....

like i said, it's not rightly their responsibility to create exporters and such.

but if they don't, they gotta live with the realities of the market and come to grips with the fact that they might lose out to other apps that have popular support in things like 3rd party game engines.

so yup, they may be losing money by not being more on the ball but then again, they really may not have money enough to deal with the issue properly.

them's the breaks.

jin

pauhana
04-25-2004, 02:42 AM
I'm glad you got that off my chest:D Heh heh heh. I don't think Lightwave sucks...it's just strange that the other big 3 packages address the character animation tools more. I've been waiting patiently since version 5.0...

Emmanuel
04-25-2004, 03:54 AM
Hi guys,

well (irony on), 8 *is* about improving the character tools, or is it ?
(irony off)
I have to second that: before introducing stuff like bone dynamics, IKB etc, they should have updated the character tools to a level that was at least on Maya's.
To me, it still all looks like the separation between modeler and Layout is the biggest obstacle for integrating ANYTHING in the same way that other apps have it.
Integrating the ortho tools sure was a major step, but I fear we'll end up with half-baked stuff until the core architecture gets overhauled.
Apart from that, Jin, I would say that the situation for NT is a bit different: if they want to be able to compete with Maya and Max in the games development business, they simply have to offer some kind of
advance/value added stuff to make people interested.
The integration of LW would have to be smooth and flawless, so they probably need to create exporters or something similiar, because there is nothing else in LW that would justify a game develoepr to add it to the pipeline, let's face it.It has not the necessary display options for DirectX games, not the robust character tools, not the accessible API, not the support for things like Havoc, no smoothing groups, not realtime game content statistics (tris, sommthing groups, number of vertices after texture layer addition etc).
Unless they really work in that direction, I don't buy any commitment to the games industry at all.

ghopper
04-25-2004, 04:17 AM
All I can say is that NT should learn from Avid.

Look what smart move Avid made, they've worked together with Valve to help them make Softimage better and easier to use for their game development.

Just check out Softimage's website, you can clearly see how heavily they are going to support HL2 community. I mean HL1 modding is big and for sure HL2 modding is gonna be even bigger. So this means even more publicity and hype for Softimage.

But then again, I guess the games industry is probably not NT's target audience.

Anyway, what can NT do, to make LW a game developer's best friend ? Or is it too late already for NT to break into the games industry, I mean now that you can get Softimage Foundation for ~$2000. I wonder how many users are actually gonna switch over.

jin choung
04-25-2004, 04:38 AM
Originally posted by Emmanuel

Apart from that, Jin, I would say that the situation for NT is a bit different: if they want to be able to compete with Maya and Max in the games development business, they simply have to offer some kind of
advance/value added stuff to make people interested.
The integration of LW would have to be smooth and flawless, so they probably need to create exporters or something similiar, because there is nothing else in LW that would justify a game develoepr to add it to the pipeline, let's face it.It has not the necessary display options for DirectX games, not the robust character tools, not the accessible API, not the support for things like Havoc, no smoothing groups, not realtime game content statistics (tris, sommthing groups, number of vertices after texture layer addition etc).
Unless they really work in that direction, I don't buy any commitment to the games industry at all.

howdy emmanuel,

yah, this opinion has been voiced many times and it's one that i wholeheartedly agree with.

problem is this:

THEY HAVE TO, THEY MUST, ABSOLUTELY, DO THESE THINGS!

BUT

they simply don't have the cash.

sure, it's certainly a wise saying, 'investing in the future.' but that only works if you have the money to invest in the first place.

i know in this day and age of easy credit for individuals, it is difficult to imagine being in a scenario where you absolutely MUST MUST MUST get something done (broken toilet, timmy's braces, broken transmission) but you have 0 money to work with.

but i do believe that this is newtek's current position. they've been in a game of poker and they're just getting raised out of the game.

sure, it's possible for businesses to take on loans and such but what if they're already stretched out to the limits?

yep, there's much to be done but it remains to be seen what they can actually afford to pull together. i'm looking forward to seeing how the 8.x series works out to be.

jin

jin choung
04-25-2004, 04:50 AM
oh,

and ghopper brings up a good point about switchovers.

it used to be that lw cost much much less than the 'big boys'. certainly, this was part of the reason that the big boys created 'jr' versions that were about as cheap.

true, if you count the cost of upgrades (both frequency and price), lw is still MUCH cheaper.

but this presents newtek with a 'ROCK AND A HARD PLACE' dilemma.

they may need money to radically improve. but they can't get this kinda money unless they jack up the price (perhaps radically). but if they jack up the price, there's an immediate temptation for many users to jump ship to other apps that will be similar in price and probably far more established, stable and refined than any new incarnation of lw can be initially.

i can't speak for others but price is a tremendously large reason why i currently own lw and not maya. it's certainly not the best (honestly, it ain't) but it really does a hell of a lot for the money which i believe is lw's PRIME SELLING POINT right now.

so if they jack up the price, they risk losing the likes of me and the question remains whether they will pull away 'BIG TICKET CLIENTS' away from the likes of xsi and maya.

imo, they're stuck.

they're a niche product and the best that they can hope for is to keep the price/performance ratio high enough to keep their core customers.

and pray like hell that the new batch of programmers are a bunch of latent geniuses that can work miracles on a shoestring budget.

jin

Emmanuel
04-25-2004, 05:28 AM
Hi,

well, I agree with the money theory a bit, yet, wouldn't the lack of ressources make it imperative to make the core and basic features as robust as possible ?
I ain't no coder, but I imagine that bone dynamics are harder to implement than a three click IK setup like in Maya.
And these would be features that really would benefit workflow and character setup (which NT was their focus with 8, if I am not mistaken).
Easy IK setup, easy character workflow, pole vectors, better deformers, faster feedback, more/better undos, easier GUI customization etc.
How hard can it be to have a close look at competing products and just follow the leader ? One reason would be: because it saves re-inventing
the wheel, another would be that users could find it more appealing to work with LW because they can use their knowledge.
LW is a follower today, not a leader, so everybody would benefit of it beeing similiar to other apps in setup and workflow behavior.
Again, without integration of modeler and Layout, this will be very difficult.
Yes, there are a lot of people who like LW for what it is and for how the GUI workds, but: its about offering the choice and do the same as the competition without repeating their mistakes and by ironing out their bugs.
Its about beeing able to say, yes, LW can do it, too, and I can get the result easier and faster.
Having to click 12 times in LW vs 3 times in Maya (I am just making up these numbers, these are just examples) doesn't speak for LW, especially for tasks that are repetitive and need to be done on an hourly basis.
I mean, basic stuff like rendering from any viewport, undoing everything to a certain point, using all the three mousebuttons, maybe integrating something like Maya's stroke GUI, or finally adding a basic walk generator, these are things that LW would benefit from much more than bone dynamics, IMO, because how many times do You animate a parachuting zombie where You need the wind to affect the bones vs daily production work that requires basic features which just aren't there or to tedious ?
Especially with a low budget, You need to make priorities, and IMO fany gimmicks aren't the way to approach this, especially if companies such as Pixologic, Maxon etc, which are also as small as NewTek, are on the fast lane :/
I have been advocating this direction since LW6, and at the time asked why we don't have undo in Layout but superfancy Hypervoxels.
Nothing has changed, and I am about to loose my faith and loyalty.

Emmanuel
04-25-2004, 05:36 AM
Oh, and BTW, I talked yesterday to a MAXer who said the same happened with the first integration of Havoc into Max.
It had flaws and problems, and people were compalining.
What I don't like about this industry is how easily users can be used as beta-testers and work-in-progress-customers.
Why is it so hard to focus on one features and refine it to make it as usable as possible ?
Could it be that all the betatesters just *loved* 8 ?
Or would You rather think that NT thought, well, we need to release something, and we will add the necessary functionality and bug fixes in the next revision ?
Features that "can be used but are not as fully implemented as You'd wish" are much too often in such products like Maya, max, LW and even XSI.
This makes some people say " I will rather wait for the next revision so tht all the bugs and edges are ironed out", which again shows some kind of resignation towards the software companies, I would say.
I mean, if You are accustomed to the fact that new features only work after two, three revisions, something si wrong in this business :/
Its like selling car by saying "its got wheels, but they don't turn very well yet, maybe with the next revision".

jin choung
04-25-2004, 05:42 AM
well,

in the case of lw8, most of the new stuff is well and truly external plugins that they acquired.

but to be fair, this is a reflection of the situation at newtek... they had a whole new programming staff so most of the new stuff HAD to be 3rd party plugins really....

so we have developments that were perhaps not the very best next steps that they could have taken but... it is what it is.

shrug.

hopefully, the development staff ramps up to the point where they can start addressing core issues.

haha,

seriously... they should just use the latest release of maya as their R&D! pride goes before a fall... dudes, just copy!

seriously, i've said this before but pride at this point is simply not an option for them. they MUST have a copy of MAX, MAYA, XSI and C4D (houdini would be overkill) and be as comfortable with these apps' featuresets as they are with lw.

it's difficult to see what you're doing well or badly if you have no point of reference.

i've never ever advocated copying another app JUST to be like another app.

but the essential question should always be IS IT BETTER? and if it is, copy them.

jin

private
04-25-2004, 06:11 AM
Copy or rewrite from the ground up. I think some other companies thought the same thing. I would disagree that Lightwave's strength is only price. It's also modeler, workflow (modeler), upgrade pricing and educational software pricing program. Throw Proton and generous plugin writers too.

I agree Maya's IK solution is much better system. I just don't know how much patch work can and will be done to the 8.x code. I hope a lot.

jin choung
04-25-2004, 06:21 AM
didn't say that its ONLY draw was price.

but in my estimation, lw's 'core competency' is delivering 'enough for the price' or the ol' price/performance.

i would agree that a total teardown rewrite is in order but talk about pricey... the best we can hope for i think is that they just keep abreast of what everyone else is doing and keep up in ways that don't break the bank.

jin

Emmanuel
04-25-2004, 09:12 AM
The way I see it most of the time: do a rewrite or loose.
NT can hardly rely on just keeping the hardcore loyalists, they need to expand, like any good business.
I wouldn't even use the low eductaional price, haven't I seen somewhere that XSI goes for 295$ to interested students ?
Talk about a low price !
The thing is: both Maya and Z-Brush were able to get a foot into new areas, Maya achieved something LW never could: become as attractive as Max for games development, from scratch !
Z-brush seems to be partially used by WETA, ILM and growing, and it doesn't even do animation and is probably a 1-man-company or so !!
LW desperatly needs a killer feature that pushes the whole industry forwards, they did it with metanurbs, radiosity, hypervoxels, surface baker, HDR lighting.
Okay, due to internal trouble, they lost that pioneering position, but there are still the same NT supervisors at the helm (Tim J., Andrew C.,
Deuce etc...), I hope they will get back on the horse and give us a "Seabuiscuit"-performance-surprise :)
FPrime is truly the only thing right now that makes LW shine.
Expand that into a new render engine with full LW compatibility, and there is a chance....

Castius
04-25-2004, 11:49 PM
I'd like to say that I hope to god they don't copy Maya IK. Trust me I know all to well that LW IK has some issues. If you want easy IK setup just use autoIK or fast IK scripts. There are plenty of issues with Maya's own IK no need to copy it.

In terms of what I think they need to do to get LW back on track. IMO it's very closly related to game development.

IF the new team doesn't have it already, then establish a set of statements that embody certain elements that make LW great.
Make a list of what needs to be add/changed/fixed or taken away.
Priorities the list.
Decide how deep to the root of the app needs to be changed to support the future changes.
Establish how long that will take, bring in outside input someone that has not been involved and can give a fresh outlook.
Reevaluate to make sure you can get it done in a reasonable time and you’re market will be there and will be interested in investing in the changes made.
Then get cracking you now have milestones to meet.

We all have our own opinions on how newtek should do all that. Maybe it would be fun to create a contest for the forum. It would consist of a set of ? like the outline I just through together. The best set of answer wins a coke. 8)

1) list the 10 best things that you think make LW great
2) list 1-10 things that might improve these ideas
3) list 1-10 things that should be add/change/taken away in each category
a. GUI
b. Rendering
c. Animation
d. Rigging
e. Modeling
4) Priorities your list
5) Guess how long it would take to complete

jin choung
04-25-2004, 11:56 PM
having a public todo list would be great.

so far, newtek reaction seems to be, 'it ain't gonna happen'.

actually, what problem do you have with maya's ik?! i would really like to know.

cuz i used to think setting up a character was laborious in lw but in maya, i got very predictable results from rigging to ik anims....

so yah, i'd really genuinely like to know what you feel are maya ik weaknesses... and if there's a better example you can think of.

jin

jamesl
04-26-2004, 01:20 AM
Originally posted by jin choung
so yah, i'd really genuinely like to know what you feel are maya ik weaknesses... and if there's a better example you can think of.

jin

Well, I've been using Maya since 1.0, and I can think of plenty of weaknesses. First off, if you use all the bells and whistles in Maya to pull off a multi-character shot with any complexity, you end up with a shot so heavy and lethargic as to test the even most seasoned animator. Think of teams of animators animating scenes with dozens of characters in a single shot. TD's can set up hunderdes of blendshape/set driven key relationships, add individual expressions between surface clusters and joints, relate lattices to joints, etc. But the higher level of abstraction makes things a bit sluggish, and prevents the ability to swap out models for quicker response (something LW users take for granted). But I'll say this... the ability to create your own IK solver through the Maya API is a real good idea.

j

jin choung
04-26-2004, 01:31 AM
well actually,

my question was specifically regarding castius' critique of maya's ik system. so my question was - what about maya's ik didn't he like and what better has he encountered.

as for the swapping out with proxy meshes... errrr... if they couldn't do that... errr... well... uhhhh... see... ummm... most movies are made with maya... ummm... and well... ummmmm... they don't always feature single characters... so ummmm... yeah...

i would say that the assertion that it is 'DOABLE' is a euphemistic understatement....

jin

tudor
04-26-2004, 02:17 AM
Ok.. Mayas IK.
Or, well.. It's contraints.

I have a habit of constraining the last joint in a chain to the IK handle. Just one IK, and rock solid. No jitter..
But why can't I have any other FK keys on at the same time?! I must first clear my FK keys on the joint, and then constrain, and thereby losing all previous animation on that joint. You can't say the same for match goal orientation.
No biggie you might say. Well, If you have the main character in the game with 160 or so moves, and they want another move that require me to add IK, it is a problem.
Now I solve it by first copying the keys to a locator, do the ik/constrain work, bake that, then copy the FK keys back.

Also, when using more than 3 joints in a chain, Mayas IK solver leaves a lot to desire. LW's IK, where it is eaier to use multiple goals is a lot better when it comes to this.

LW's IK has a great feature that I like to use aswell. The non fulltime IK. Great for posing, but gives FK tweening. In Maya that means many clicks more than in LW. If we could only keyframe the full time IK button I would be happy..

jin choung
04-26-2004, 02:52 AM
well,

i never ever use more than two bones on an ik... i've never had to animate crazy tails or tentacles but i seriously doubt i would expect a multibone ik to pull through... maybe spline ik.

as for fk/ik mixing, all the tutorials on the net for maya seem to use a really ingenious solution....

instead of having one right arm... you have THREE!

only one of the arms is actually bound to any geometry. another arm is controlled by ik, the last arm is fk.

and then, you set up a CONSTRAINT SLIDER so that your 'real arm' is either following IKARM or keyframe a transition to FKARM.

for games work, as long as your original mesh has no IKARM, FKARM, after you bake the animations, you should be able to simply delete the IKARM, FKARMS before exporting your animations and you don't even incur extra overhead - although since no geometry is bound to these bones, it really shouldn't make that much of a difference if they were in there.

jin

tudor
04-26-2004, 03:23 AM
And LW'er complain about workarounds ;)

jamesl
04-26-2004, 05:57 AM
Originally posted by jin choung
[B]
as for the swapping out with proxy meshes... errrr... if they couldn't do that... errr... well... uhhhh... see... ummm... most movies are made with maya... ummm... and well... ummmmm... they don't always feature single characters... so ummmm... yeah...

That's about the state of it. Lightwave isn't just good because it's cheap. I think they actually have something because they have a lower level architecture... ala Houdini (if you squint hard enough, it's there) than Maya, while providing artist level tools for real animation. Anyways, that's the only reason I use it.

j

Castius
04-26-2004, 08:38 AM
jin in terms of Maya IK it has come a long way since Maya 1 but I still harbor some ill will from my first contact with it. I'll admit I don’t know the latest Maya as well as I should to be giving a full evaluation of it but I will try.
Over all I think it's harder in Maya to make sure IK hide that the look of IK. Ik snapping is harsh, and the simplistic nature of the IK is some times to simple. What I still see it only spline IK for multigoal IK cause alias tried multigoal IK and FAILED badly. Because of this you need pole vectors. As much as I'd like the ability to set up easier pole vectors in LW I’d like to set up IK like LW in maya. A strong case of grass is greener on the other side. So it's not necessarily that I don't like Maya IK it's just that saying someone wanting to copy Mayas IK is to simple a statement for a complex issue. You know what I mean.

In many cases LW IK can be as easy and as fast as Maya’s IK. BUT the motion system in LW makes constraints very poor ATM. So what I would rather see is the heart of the motion system inside lw fixes and I think we would see LW IK improve along with it.

Tesselator
04-26-2004, 02:48 PM
Ya Cast, the IK in M6 is nearly as good as L8.

...nearly!

Gwot
04-26-2004, 06:30 PM
My biggest complaint these days (post LW 8 release). Has more to do with performance in layout. I think LW is a pretty decent package overall, but I've been spoiled by the real time performance of Motion Builder's viewport rendering.

I was just bitching to fortress on icq about having to work with 5 characters at once in both LW and MB today. 5 characters for a total of approximately 30000 polygons, 100 bones per character (plus 4 complete ik rigs in MB only) and Layout was choking so bad it was unbearable.

Being able to transform bones/rigs, manipulate the camera, scrub the timeline and get instant feedback, or even full smooth playback in the viewport is a luxury that only MB seems to be able to provide. Even in bounding box display mode, MB was smoking LW.

I think this has become my number one wishlist item for future LW updates. I can live with most other issues I have with the app. They are minor compared to this, and every other app out there has it's equal share of minor issues too.

mechis
04-26-2004, 09:25 PM
Hi Gwot,
I'm guessing you've told your complaints to Newtek yourself, but here's some other info-:

I e-mailed Newtek's feature request, asking that they work on increased feedback when animating (like Motionbuilder) and Deuce e-mailed me back to say they were working on it. When we will see the fruits of their work?... I don't know, but it's good to know that they see it as a priority.
~Mechis

jin choung
04-26-2004, 09:49 PM
jamesl,

my point was that swapping proxies in maya is done everyday. as for level of interactivity, the largest, cgi scenes ever committed to film have been done with maya (though probably not from a single maya scene!). i would seriously doubt that under similar workloads, especially considering lw's opengl redraw rate, lw would come out on top in multicharacter scenes. but that's not a slam. how much less do we pay for lw? nature of the beast and that's fine by me.

as for lower level architecture, although lw is more primitive (not a slam, just a statement of fact), i think that maya is actually much more lower level and without a doubt, much better designed in terms of elegance, consistency and software architecture.

if you think about maya's interface as a GUI riding on top of mel, this becomes immediately evident. EVERY SINGLE MAYA KEYSTROKE is actually a mel command. so in terms of getting very low level access to data, it is much easier to do this in maya than lw. and very easy to learn mel by starting out by just copying a bunch of keystrokes to create batch melscripts....

lw's gui does not 'ride on top of lscript' like windows 3.1 to dos. although lscript commander brings it close but alas, that's only for layout and i don't think it's exactly as low level as maya. i actually think there's much more processing going on there simply because the gui does NOT ride on top of lscript... everything's being re-interpreted....

[software architecture note: windows 3.1 was REVILED for simply riding on top of dos... and everyone would agree that the 95 and NT technologies are far better. however, because of the current speed of computers, having a 3d app's gui ride on top of a command line system does allow for some really cool 'low level access' with not a lot of loss of speed.

in an example that exemplifies the opposite end but the same principle:

i used to think that a gui was a complete waste of cycles... it uses up so much resources... little did i imagine that computers would come to a place where such resource usage becomes negligible.]

castius,

that's cool.... but i'd like to hear what you think about it with the current versions of maya. and as for the ik snapping being harsh, i guess that's a matter of taste... for me, i would describe it as rock solid and tight and most importantly, predictable... it is up to me as the animator to get rid of any 'ik feel'... lw's seems a bit 'mushy' to me.... but especially with the thanks of another thread here talking about ik, i do think of lw's ik now as being definitely usable.

but i WOULD like the ik setup procedure imitate maya's if nothing else. seriously, maya's is 3 clicks to setup the simplest setup... lw's is a trip through a bunch of drop down menus.... the fact that you have to create a null for lw is in my mind a dumb idea... i mean, if the very notion of IK REQUIRES a goal for it to make sense, it should be PART of the ik system that's automatically created/removed....

jin

KiDCoDEa
04-26-2004, 11:29 PM
Originally posted by TriXal
That might be but not with complex models.
I model cars which consists of thousands of polygons and many layers.
Usually when I finish editing or adding something I weld every vertices and then I use unify to check that there's no polygons I don't want. So if I have already made my "smoothing groups" they are gone after my welding operations and I have to do them again.
And I don't want to worry all the time that I accidentally screw up my "smoothing groups".


exactly my prob. i miss smoothing groups also, and feel surpassed by 3ds noobs...

just coz as usual, lw doesnt support this BASIC game dev feature...

Jake
04-27-2004, 08:38 AM
jin--


but i WOULD like the ik setup procedure imitate maya's if nothing else. seriously, maya's is 3 clicks to setup the simplest setup... lw's is a trip through a bunch of drop down menus.... the fact that you have to create a null for lw is in my mind a dumb idea... i mean, if the very notion of IK REQUIRES a goal for it to make sense, it should be PART of the ik system that's automatically created/removed....

Isn't that the whole point of the new options for importing pre-existing rigs? Is setup that much of an issue if you are able to reuse rigs and easily adjust them to fit other characters?

tudor
04-27-2004, 08:41 AM
Well.. In Maya you might have a standard FK rig, and by just a few clicks you have IK.. Do the stuff needed with IK, Bake and delete IK. Later you may need IK on another part. Easy to add a new IK chain.
Atleast I don't just use IK in the main rig setup, but add it as I need it. Just as any tool there is.

04-27-2004, 08:45 AM
Jin: Proxies are fine and I'm totally down with that. My complaint was more over the fact that even if my display was set to bounding box in LW, MB still blows it away at it's highest display settings. There's just no comparison at all between modern hardware accelerated viewport rendering and outdated stuff like LW's.

I don't need to render out previews in MB to playback my animation and I don't even need to use proxies. In fact, as an example of how much better modern realtime 3d rendering can be, I get better playback in our game engine with our characters fully textured and lit by the engine than I do in LW. So I either animate things in MB or export to the engine to see how it looks. I can even manipulate my viewport while I'm playing back in MB and it updates just fine, so I can examine my characters from any angle while they are in motion.

I agree with you on the IK stuff and pricing too. LW is dirt cheap compared to everything else, especially for what you get. But that advantage is dwindling as Maya and XSI have drastically lowered their pricepoints.

IKB is a step in the right direction towards simpler setup and use. But it's not finished yet, and not as well integrated as it should be. I'll take it as is though over not having added it at all. I'm patient enough to wait for Newtek to develop it further.

Mechis: Yep, I sent my list to Deuce as well. Most of it had to do with IK/bone tools. ;)

Edit: sorry, [email protected] = gwot :)

scott_krehbiel
04-27-2004, 09:27 AM
Originally posted by pauhana
.... If Newtek were in demand for game development they could sell at least 20-40 licenses per company (Independent game developers) and much more for large developers. .....

I fear that if they were lucky, they'd sell maybe one license per company for the smaller shops. A previous game-developing employer of mine didn't buy a single license of the several installations of Max and Photoshop that they were using. (OH! Did I say that out loud?)

I kept raising a stink, and finally got them to purchase the copy of Max for me to use. After I continued, they finally purchased Photoshop for me also. Turns out I was labeled a "troublemaker" and was almost fired. :cool:

In case anyone thinks that money was the issue, they kept buying out other companies, but wouldn't buy the software that their own people used. :rolleyes:

I hope the rest of the industry isn't as criminal as this previous nameless employer.

Peace,
the troublemaker

BTW: Turns out that they're basically folding (or imploding, depending on how you look at it). I knew that it'd all bite them in the --- eventually. :p

cgolchert
04-27-2004, 10:45 AM
Originally posted by jin choung it would be NICE. but you can't fault them.
jin

Except that game support has been promissed twice over the years.

If exporters existed, a company might allow artists to use Lightwave and fit it into their pipeline as opposed to having to crack off a programmer to write one.

fortress
04-28-2004, 05:59 AM
speaking of game tools any word on the lw8 sdk

jin choung
04-28-2004, 06:16 AM
actually,

that's an excellent question; if development of the SDK was touted as a feature of the upgrade, it would be tremendously nice to know what exactly has been changed or revamped.

jin