Questions about how to create models

Discussion area for user-created Allegiance models, textures, voicechats, music, and other ingame content.
Adaven
Posts: 1959
Joined: Sat Oct 25, 2003 7:00 am
Location: Greater Ozarks

Post by Adaven »

Your_Persona wrote:QUOTE (Your_Persona @ Mar 9 2012, 12:54 PM) I've started working on a blender export script to mdl. It's working for basic meshes as of last night.
Can you sort of explain what you are doing in more detail?

In particular:How are these text MDL's different than what we use in game or the meshes I would normally create in Blender? If text MDL's were part of regular Alleg, how did they fit into the old tool chain? Are you incorporating any of the MS tools into your processing steps or are you directly trying to create these files from scratch?
Adaven
Posts: 1959
Joined: Sat Oct 25, 2003 7:00 am
Location: Greater Ozarks

Post by Adaven »

Thanks MadP. I was about to post a bunch of questions, then I remembered the links you had posted. Looks like they'll answer most of the ones I had at the moment.
Compellor
Posts: 994
Joined: Fri Jul 06, 2007 12:56 am
Location: Columbus, OH

Post by Compellor »

Adaven wrote:QUOTE (Adaven @ Mar 8 2012, 08:07 PM) Sorry for getting the cart before the horse.

Looking back at the wiki, I found an article by Imago on using the MS tools to generate hitboxes. I'm hoping this is still fairly correct? Also, would this fall more under the Rigging/Meta-data phase or Optimization? I'm leaning towards the former.
That's mine actually. I was never 100% happy with that article, I was sure I was missing something because it seemed like the collision models didn't always end up working exactly as intended. I haven't done anything with models since mid-2010 so I don't know how well I could answer any questions now. Still, the current Rix Expansion model was done using that method, so it can't be all wrong.
Last edited by Compellor on Sat Mar 10, 2012 6:15 am, edited 1 time in total.
Any job worth doing with a laser is worth doing with many, many lasers. -Khrima
Beyond a shadow of a doubt if you don't watch them like a hawk they will stack their collective balls off - MrChaos on Alleg players
Imago
Posts: 1440
Joined: Tue Sep 23, 2003 7:00 am
Location: Minneapolis, MN
Contact:

Post by Imago »

Image

These bugs haven't been fixed yet because don't have any developers interested in fixing them up. --Tigereye
Imago's stupid-sensor is supersensitive. --RealPandemonium
The art is managing the flow of the drama to achieve the desired results. --Big_Beta_Tester
joeld wrote:But we’ve been amazed at the level to which some of the Allegiance fans have remained hard-core.
Your_Persona
Posts: 773
Joined: Sat Dec 04, 2004 8:00 am
Contact:

Post by Your_Persona »

Ok, here is a video of it working at the most basic level.

http://www.youtube.com/watch?v=yMlpj0wMqbQ

And here is the script if you want to play around with it.
Its very rough, and the save location is hardcoded at the bottom of the script.

http://pastebin.com/rruZbKyE
-->>Elitism<<--
I'm not Hamlet. I don't take part any more. My words have nothing to tell me anymore.
fuzzylunkin1

Post by fuzzylunkin1 »

Your_Persona wrote:QUOTE (Your_Persona @ Mar 10 2012, 03:23 AM) Ok, here is a video of it working at the most basic level.

http://www.youtube.com/watch?v=yMlpj0wMqbQ

And here is the script if you want to play around with it.
Its very rough, and the save location is hardcoded at the bottom of the script.

http://pastebin.com/rruZbKyE
Just FYI, you don't need the trailing \ at the end of lines (except for after "line = "). At least in my python version.

Code: Select all

>>> line = \
... ', '.join([
... ', '.join(map(str,[0,1]))
... , ', '.join(map(str,[2,3]))
... , ', '.join(map(str,[4,5]))
... ])
>>> line
'0, 1, 2, 3, 4, 5'
Although normally you put the commas at the end of the line :lol: . (And putting a comma after the last item (some do it for for consistency) doesn't hurt anything)
Last edited by fuzzylunkin1 on Sat Mar 10, 2012 9:16 am, edited 1 time in total.
Adaven
Posts: 1959
Joined: Sat Oct 25, 2003 7:00 am
Location: Greater Ozarks

Post by Adaven »

Compellor wrote:QUOTE (Compellor @ Mar 10 2012, 01:12 AM) That's mine actually. I was never 100% happy with that article, I was sure I was missing something because it seemed like the collision models didn't always end up working exactly as intended. I haven't done anything with models since mid-2010 so I don't know how well I could answer any questions now. Still, the current Rix Expansion model was done using that method, so it can't be all wrong.
Ah, pardon my mistake then. I must have seen Imago's name but misread the rest of the first section.
Adaven
Posts: 1959
Joined: Sat Oct 25, 2003 7:00 am
Location: Greater Ozarks

Post by Adaven »

madpeople wrote:QUOTE (madpeople @ Mar 4 2012, 09:24 AM) Unfortunately it's not as simple as just passing a .x file to mdlc / xmunge or the original MS .bat files in the source. Well, it is if you try to convert the .x files in the MS source, but they use an old format for .x files. Files from modern .x file exporters didn't work too well. I managed to get it working by comparing the layout of the MS .x files with the ones from my exporter and writing a program to convert from the new layout to the old (the old layout had a name but I don't remember what it was). After converting from new to old layout I think I abused the flattening functionality of xmunge(?) to make a .x file to feed into a heavily edited version of the MS .bat file for conversion.
According to the file headers, Alleg is using version 3.2 of the .x format. I've tried importing the MS models and then exporting them again from some different free tools. The few exporters I've looked at all use 3.3. Even then, there is some variability between exporters, even with the same supposed file format version.

The original models were made with Softimage 3D. Softimage is owned by Autodesk now, and they make a free non-commercial version, called Softimage Mod Tool. It appears to import the MS .x files really well. All the pieces are in the right place, animations work, joints present, and textures. Their exporter also uses the version 3.3 format, but just glancing at the output, it appears to be closer to the original than what Blender and Wings put out.

I'm not familiar with the MS tools very much, although I hope to start learning soon. I'll try to post some examples of the Softimage exported files next week, maybe someone can tell me if they are a better fit that what's been tried before?
Last edited by Adaven on Sun Mar 11, 2012 3:21 am, edited 1 time in total.
MrChaos
Posts: 8352
Joined: Tue Mar 21, 2006 8:00 am

Post by MrChaos »

KGJV wrote:QUOTE (KGJV @ Mar 6 2012, 04:30 AM) I wasn't aware of that and I'm not an expert in modeling and 3D file formats.

But we do have the source code of the MSR tools and we can change or fix them if they're not working correctly with modern modeling software.

This is a typical 'Chicken or the egg' situation added to lack of correct information broadcasting:

Most people are not aware of the tools and are told to use the Milkshape plugin.
The few who try the tools either find it too complex or are stuck by .X File compatibly issues. they stop using the tools but do not inform the dev team about the problem.

Now things we can do. I see 2 main tasks here:

I'm aware that the MSR tools are complex to use and that's is why I recommended having other people doing that 4th stage. People that are skilled at scripting and file manipulation.
Simplifying the tools so that a larger audience can use them is probably something worth doing. Documenting them also, Ada just started this.

If dealing with .X file format is an issue for 'art people' than we need to find a replacement for it or at least find an intermediate format to use between stage 3 & 4.
A format that is commonly accepted by everyone as a 'standard' (if such thing exist in 3D modeling? Collada may be? any experts in here?), a format that most modeling software can produce without installing addon or 3rd party module. A format that can support the 'meta' informations needed for Allegiance (directly or with a 'trick' like with .X format).
Once that format chosen we can write a conversion tool that will take this format a produce 'clean .X file(s)' for the MSR tools.
Other way of doing this would be to rewrite the MSR tools so they use directly that new format but I foresee too much work for the same result (at least for now).

The required skill set is more around scripting for these tasks but if you need C++/Allegiance engine expertise (MSR Tools share code with Allegiance engine) I'd be happy to help.

But like always with projects, small or big, we do need someone to carry it , a project manager that will lead the way. otherwise this will just be yet another forum discussion and the 'deja vu' feeling won't go away...

so 1st step is to ask MrC to find a volunteer for this I guess ? (or 2 , one for each tasks and they'll coordinate stuff together).
pkk wrote:QUOTE (pkk @ Mar 6 2012, 05:03 AM) Kage, I only seen 3-4 people, who could do this. All of them already posted in this topic...

Yourself
Madpeople
YP

Someone should step up, no need to ask MrC.
KGJV wrote:QUOTE (KGJV @ Mar 6 2012, 09:17 AM) yeah lots of 'deja vu'. There are some good infos in these topics that should be curated and put in the wiki.

AMT2 is dead and buried long ago. 90% of its complexity was to code a 3D editing system which already exists in all 3D modeling software...no point really.
It's very complex to do and was puzzling me till I found out about an article explaining 3D attractors and other stuff.
Once I understood how to do it, the immensity of the task to actually code it wasn't worth the result...
Also I'm now convinced that allowing to read MDL/CVH is a bad thing, it's like decompiling EXEs... MDL/CVH are final, compiled formats. you don't want to read them, only to write them.

So it's better to work at a higher level with mainstream modeling software and then send the digital assets into a processing pipeline (one-way pipeline).
That's how all games are built these days and there is a good reason for that.

Pkk, I don't think you understand what I'm talking about. The leader I'm talking about isn't the person who would actually DO the work. It's a team work.
That is why not much is done here these days because people assume that those who can DO will do. And then they just wait for them to do...
It's a bad habit we have in this community and I'm responsible in part for this. I used to do it all too but that only work when you have plenty of spare time to spend.
And it's not healthy for the project anyway.

So we need someone to LEAD 1st , that someone will find the people who would DO the actual work but before he will discuss and specify fully what to do.

I'll not edit any of the above and if I follow... let's see if I do first

The way we model things for Allegiance is a bit... out of date but works...ish. This causes unintended issues and the problem propogates across volunteers until someone is stubborn enough AND talented enough to bull through the issues to produce something useable. This creates a legacy and momentum to continue to do things tthe same way as was done in the past. As the years have ticked by this has gone from quaint, to a hassle, to a bonafide issue. The clued :roll: and it becomes another reason not to do stuff.

Someone, on their own accord, shows an interest in doing something for the game and runs smack into a roadblock. Never encountering it before they are naturally curious and even confused as the details of the issue. After some grumbling from the clued most just say $#@! it and move on. The clued discuss the way forward and well the whole topic is repeated ad infintium until it too becomes "the way we have always done things". The poor dude in charge is stuck with a big bag of legacy and no with the talent who ill solve the issue.


I think that is what is being said above converted to MrChaos Speak™


To be crystal clear, pkk is absolutely in charge of the development stuff and he has a big job on his hands. The idea of changing the way artwork is submitted, processed, and implemented in the game probably spreads across several zones... it all depends how you cut it in a nutshell. I AM the roadblock remover, pom pom waver, buck stops here dude tbs and after consulting with pkk, who is a fine fellow with a wicked sense of humor, he finds no issues we could get started.

Just we have limited resources in sheer numbers of people, and lots of concerns. One of the worst things that can happen now is for projects to die on the vine after I've whipped them into a frenzy via my pom pom waving and short skirts ;) We really cannot afford to waste the volunteers. I'm working on that issue too but that is the facts for the moment.

MrChaos
Ssssh
KGJV
Posts: 1474
Joined: Tue Jul 01, 2003 7:00 am
Location: Transilvania

Post by KGJV »

I had buried that topic since it's wasted with spam but since MrC kindly PMed about his reply, I'll answer to him.

I think a lot of people are confusing things here.

So 1st thing 1st. if someone NEW ask about how he can make models for Allegiance, you don't answer with some complex unclear stuff that will discourage him right away. nor do you display disagreement and spam behavior. You simply answer:

Mr X, thank you very much for your interested in modeling for Allegiance. Plz do with whatever modeling software you're used to work with. We will THEN take your work and process it so it can be used withing the game. Why ? because modeling for Allegiance and adding models to the game isn't the same thing , it's too very distinct things.
It's like writing a book and publishing it, 2 separate things requiring very different skills set.

So to simplify:

Q: How to create models for Allegiance ?
A: As you want, with whatever software you're used to work with (3Dmax, LW, blender, Google Sketchup, Maya, Wings, Milkshape,you name it).

Q: How do I publish models for Allegiance?
A: short answer: ask the Art ZL (whoever that is) to do it (or have it done by someone).
A: long answer: your model needs to be converted to the internal format used by the game. To do so, it must be adapted and processed by some specific tools, and eventually hand tuned. Then someone must DECIDE where your model with go in the game. Who decides this is not clear since it's both a core thing and an art thing. And will probably generates lot of discussions between people. Yeah you can't just add new models to the game like that.

now back to the topic which about PUBLISHING models (and not creating them):

QUOTE To be crystal clear, pkk is absolutely in charge of the development stuff and he has a big job on his hands. The idea of changing the way artwork is submitted, processed, and implemented in the game probably spreads across several zones... it all depends how you cut it in a nutshell. I AM the roadblock remover, pom pom waver, buck stops here dude tbs and after consulting with pkk, who is a fine fellow with a wicked sense of humor, he finds no issues we could get started.[/quote]

I don't get then why you PMed me back here then? nothing prevent anyone from getting started. certainly not me.
if you think blender is the solution, if think YP making scripts for blender is the solution. go ahead. I don't have any authority here.
If you want my opinion about this, I think I gave it already but that's just my opinion:
I don't think it's the way to do it because we learned that a specific software + plugin ( (Milkshape + AMT plugin) is a bad solution so why a new specific software (Blender) + plugin (new stuff) would be better ? this looks like history is repeating, in a bad way.
Image
Post Reply