Jump to content

A little suggestion to esim for their commercial SB Pro PE


[]_--__[]KITT

Recommended Posts

  • Replies 67
  • Created
  • Last Reply

Top Posters In This Topic

  • Members

We shall see if artwork is going to become a bottleneck. Up until recently is was programming time, so considerations about the artwork were somewhat moot.

In all honesty I have to say though that our experiences with the work of amateurs in the past hasn't impressed me. There are people who can create models, and do a fairly good job at it. But at some point you need to go to the unwrapping of the models and efficient use of limited texture space, and that part seems to be where most hobbyists realize that working on these models takes more time than they thought it would. This is where they may be losing interest, or where they fail to adhere to deadlines (even if they are self-imposed).

A professional artist may not necessarily deliver better quality with models. But he usually knows well how much time he needs for a certain task, and can do the work efficiently enough to make a living from the money that he gets paid for a model. Reliability is worth a lot since a lot depends on the adherence to deadlines. I mean, I try to be as liberal and laid-back as I possibly can be about deadlines, but in the end I need to plan the development roadmap, and I need to plan with something, and there are dependencies (e.g. sub-project X cannot be started or finished without artwork Y), and if the deadline for artwork Y is busted, the entire sub-project grinds to a halt. That way we can lose a lot more than just a certain piece of artwork, and the fact that we don't have to pay for the work until it is finished, and that we have to pay less for a model than it would cost from the hands of a professional artists cannot possibly make up for the havoc that busted deadlines create, even if they are very, very liberal.

I have offered this numerous times, and I do it again: If there is a talented modder out here who wants to make a specific model, or already has some models that he wants to see included in SB Pro, we're willing to look at the stuff and if it is good, we'll buy it. Contact us, and we'll talk. Initially we'd prefer to get one completely textured model and an honest assessment how much time it took to complete it. Once that we have that, we can talk about model modifications, texture modifications, and additional models.

But I need reliability. A slow pace can be acceptable, just don't let me down.

Link to comment
Share on other sites

There are people who can create models, and do a fairly good job at it. But at some point you need to go to the unwrapping of the models and efficient use of limited texture space, and that part seems to be where most hobbyists realize that working on these models takes more time than they thought it would. This is where they may be losing interest, or where they fail to adhere to deadlines (even if they are self-imposed).

A professional artist may not necessarily deliver better quality with models. But he usually knows well how much time he needs for a certain task, and can do the work efficiently enough to make a living from the money that he gets paid for a model. Reliability is worth a lot since a lot depends on the adherence to deadlines.

True, true.

This afternoon, I talked with a friend wich help me to make a mod for ArmA2, and basically we looked at each key steps of the development.

-Programming stuff took us few months.

-Modelling part took me 3 months to have a decent looking model (roughly twice the polycount of the Chally in SB).

-And UV mapping the model is still occupying my time (after 10 month on it and a huge number of failed attempts).

Unfortunately the work done do not equals by far professionnal work, especially when you got two maps (2048*2048) for a tank and additional maps (1024*1024) for fancy stuff like roof MGs.

My advise for the non-profesionnal modellers, who want to put their job in SB:

Use the time that eSim guys are developing tools, to get a lots of experience for optimised models.

You have everything to get from that. ;)

Link to comment
Share on other sites

  • Moderators

...and this is where I say, there are practical limits on polygon counts. If you say model X is twice the polygons of the Challenger 2 in SB then, well, that is simply not applicable to SB. The models MUST be efficient first and foremost as there are (potentially) four to six such LOD1 tanks in a scene, and of course many other vehicles possible at various ranges in the distance. That of course is the difference between ArmA and SB, the fact that SB must be able to efficiently represent up to battalion and even brigade level (Red Leopard featured a Soviet tank brigade after all).

Next, making a model goes well beyond making just a visual model: the things you never see are well beyond the detail and complexity of the model itself in many cases. Also, there must be three separate models (three LODs), and it must have a woodland, desert and (ideally) a winter texture. It must also have a TIS texture as well, arguably the least favorite part of the whole process. And finally and most difficult, it must have an armor model, which is something that people never see, but is probably the most important part of the the entire vehicle. Then there is the requirement of a multitude of blank frames all over the model that must have specific names - the frames determining (basically) reference points for various things related to the simulation.

The point is, it is probably not what you might think. Any idea you might have from making ArmA2 mods doesn't really apply to SB, because SB's vehicles are a great deal more involved. What we don't need is someone who wants to add their favorite vehicle X to the sim - as a sort of flash pan interest (not saying that is you). The complexity alone would ensure that a flash pan interest approach would be doomed to fail. And, generally speaking, there is probably a greater likelihood of rejection than acceptance due to the whole complexity of the entire process - but hey, I am just explaining the darker reality of it.

Link to comment
Share on other sites

...and this is where I say, there are practical limits on polygon counts. If you say model X is twice the polygons of the Challenger 2 in SB then, well, that is simply not applicable to SB.

I agree, but between those two simulations the specifications are different and the viewer has different viewing liberty.

The fact that I know roughly the polycount of the challenger 2 is a proof that I pay attention to that part of the equation.

The models MUST be efficient first and foremost as there are (potentially) four to six such LOD1 tanks in a scene, and of course many other vehicles possible at various ranges in the distance. That of course is the difference between ArmA and SB, the fact that SB must be able to efficiently represent up to battalion and even brigade level (Red Leopard featured a Soviet tank brigade after all).

You're talking about resolution LODs wich are displayed depending of the ranges, and the "calculating load" of the computer. Is that right?

Next, making a model goes well beyond making just a visual model: the things you never see are well beyond the detail and complexity of the model itself in many cases. Also, there must be three separate models (three LODs), and it must have a woodland, desert and (ideally) a winter texture. It must also have a TIS texture as well, arguably the least favorite part of the whole process. And finally and most difficult, it must have an armor model, which is something that people never see, but is probably the most important part of the the entire vehicle. Then there is the requirement of a multitude of blank frames all over the model that must have specific names - the frames determining (basically) reference points for various things related to the simulation).

Absolutely, visual don't give any kind of physical data, articulations, or view occlusion. I don't know every LOD for sure, but I don't underestimate their workload.

The point is, it is probably not what you might think. Any idea you might have from making ArmA2 mods doesn't really apply to SB, because SB's vehicles are a great deal more involved. What we don't need is someone who wants to add their favorite vehicle X to the sim - as a sort of flash pan interest (not saying that is you). The complexity alone would ensure that a flash pan interest approach would be doomed to fail. And, generally speaking, there is probably a greater likelihood of rejection than acceptance due to the whole complexity of the entire process - but hey, I am just explaining the darker reality of it.

Thank you for your honesty!

Still using A2 as a "reference", the strong point of that sim is it's weakpoint as well:

When people touch the content, it can become a real mess ("hey my soldier can kill a tank just shooting on it's motor! OA power!")

For that I trust you to keep it clean.

My interest is far beyound adding a vehicule to the sim, in fact it's just a good way to learn and maybe cooperate with other member to achieve a common goal.

Sorry, if I appear to you like a "Battlefield 2" guy, in fact I'm really a fan with all the bad points wich come with it.

Link to comment
Share on other sites

Darkangel,

I hate to say this but his base unskinned model looks great. I challenge you to do better.

Volcano,

Myself being a advid simmer, playing ArmA2, Freefalcon/Open Falcon, and SBpro. I can say that ArmA2 does a GREAT job of efficiently modeling and representing battalion sized engagements for infantry, and light vehicles, even some mild helo close air support. Granted the helo modeling is the weakest point in the three I mentioned.

I have been involved in online games with over 400 AI combatants and it not even slow down my computer enough for me to make note. But I also kept the engagement limited to light wheeled vehicles, APCs, and infantry only.

SBPro has some marked advantages when compared to ArmA2, BUT ArmA2 destroys SBPro in the infantry area and the graphics. And in some cases sound too.

But SBPro is a dedicated TANK SIM. Not a modern infantry combat game/sim, like ArmA2.

Hell even Free Falcon/Open Falcon/Falcon4.0 can run battalion sized engagements without much of a issue.

Link to comment
Share on other sites

I have been involved in online games with over 400 AI combatants and it not even slow down my computer enough for me to make note. But I also kept the engagement limited to light wheeled vehicles, APCs, and infantry only.

Sure but that’s an order of magnitude lower than what I often use SB Pro / SB Pro PE for - namely a BlueFor “heavy” Battle group complete (“complete” = both fighting and admin echelons depicted) against a RedFor “heavy” BDE (i.e. 3 x BG+ with the same detail).

SBPro has some marked advantages when compared to ArmA2, BUT ArmA2 destroys SBPro in the infantry area and the graphics. And in some cases sound too.

And I believe the Infantry “shortfall” as you perceive it, will be addressed very soon.

To be clear though I don’t have a pro SB or anti VBS2 / ArmA2 stance but overall I’m more than happy with SB Pro (and the various compromises it has made - esp. given that some of these are “about“ to be rebalanced.).

Link to comment
Share on other sites

To be clear though I don’t have a pro SB or anti VBS2 / ArmA2 stance but overall I’m more than happy with SB Pro (and the various compromises it has made - esp. given that some of these are “about“ to be rebalanced.).

Well, I'm biased...I admit to it!!

But anything that uses health bars for tanks is far out in my books ;-)

If SB is able to handle civilians/multi-party missions VBS2 is off the desk!!

Link to comment
Share on other sites

  • Moderators
Darkangel,

Myself being a advid simmer, playing ArmA2, Freefalcon/Open Falcon, and SBpro. I can say that ArmA2 does a GREAT job of efficiently modeling and representing battalion sized engagements for infantry, and light vehicles, even some mild helo close air support. Granted the helo modeling is the weakest point in the three I mentioned.

Well, firstly, I am not here to challenge ARMA, so no need to defend it like that; all I am saying is that SB has a conscious limit placed upon models to allow it to have large level / high intensity engagements at high frames per second. You say you can have hundreds of combatants, but whether or not you can have over a hundred vehicles, vehicles in platoon formations in close proximity of each other (and not lone vehicles driving here and there), with as many infantry, all in a given area with 5km lines of sight, and at high frame rates, well, if you can do that in ARMA then hats off to you and your computer.

Regardless, my point on that specific comment is that just because you can make a model for ARMA, it doesn't necessarily relate to SB directly, especially if said model is twice the number of polygons as one of the most advanced in SB. It is not as if Dejawolf doesn't know how to add thousands of additional polygons to a model; what makes a good modeler is someone who can make a vehicle look as good as he does, while at the same time maintaining an efficient polygon count. This approach is what allows SB to do what it currently does. It is no disrespect to ARMA or to DarkLabor.

Link to comment
Share on other sites

  • Members

The polycount, while not unimportant, is usually not the make or break point, at least not for the LOD1. Modern graphics cards can handle high polycount scenes pretty well. Being frugal and efficient with texture space is a challenge, and above all the work with the unwrapping.

If you can unwrap a model efficiently you can fix the rest of your model under some guidance (like naming conventions, texture variations, and other simulation-related tasks). The big hurdle is to finish a modeling job in a time frame where you still see enough progress in your own work to motivate you to keep going. The goal must be to unwrap a model in under ten weeks (in fact, Raino cranks out about two or three external view models per month (!), with all the texture variations etc.).

Link to comment
Share on other sites

Sorry but that looks like ass.

an old modeler saying goes like this:

a great texture can make a bad model look good,

a bad texture can make a great model look bad.

i've looked at darklabors models, and they're not bad, just a bit on the heavy end, polycount wise,

and some less than optimal polygon distributions

the textures needs a ton of work though. dirt, wear and tear, all that good stuff.

these are the specs for Sb:

LOD1 ~12000 polygons

LOD2 ~2000 polygons

LOD3 ~1000 polygons

1024 diffuse texture

128 track texture(if its a tank)

Edited by dejawolf
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...