3D Programs, MSTS (.s) & OpenBVE/BVE

Hello all,

What has irritated me a fair bit, is how you cannot create objects and 3D stuff for OpenBVE/BVE using programs like Sketchup or Blender. So I've decided to set about trying to make a tool to make it easier. Now, this won't go to the devolopment section until im fairly confident I can do such a thing.

________________________________________

Well this all started when I was attempting to turn somthing into a .b3d from Blender. All the Blender --> .b3d exporters seemed either to be out of date and unsupported or just unfinished. Having finally found one that worked, it didn't work. Later, I found a tool that'd convert .x to .b3d, so I exported it to a .x file in blender and then played about with the converter to make it into a .b3d. However all the .b3d types that the converter listed didn't seem to work with OpenBVE and BVE at all. Abit saddened by the failure, I set about looking at all the diffrent formats that blender could export, and soon discovered the .x format wasn't too far off the .csv file we use. Then a lightbulb appeared over my head and I had an idea. How's about a .x to .csv converter? That way I could build an object in sketchup and send it to blender then export it as a .x and then run this converter to make a .csv!

So the first step I did was to create a cube in blender. Having exported it as a .x, I opened notepad++ and stole the values from the .x and made a cube with a .csv!


Here is the first ever box...in wireframe mode.

The reason the x format doesn't export correctly is that openBVE and BVE originally only ever implemented a small subset of the format. Because all x format objects where converted via structure viewer from the other object formats.

If I remember correctly then I think Milkshape3d has a bve compatible addin although this is shareware and has to be paid for not sure what the price is.

The thing to remember with blender is that it uses mathematical axis ie Z is up/down, verses bve Y is up/down, so you'll need to take this into account.

Hello again,

Thanks Derryck. One of the aims of this project is perhaps to make it easier for newbies to devolp things. I really hope that it encourges more people to devolop content, and may even speed up the devolpments of current devolpers.

LabRatAndy, I wasn't even aware the .x file worked directly...until today when Derryck pointed out some working .x files. I have also noticed myself the x and y axis issues, thanks for confirming that! I heard of Milkshape3D during my reasurch, however you have to pay sadly. I gather its £20 (€25).


Derryck also pointed out in chat earlier that we should be careful, cause using 3D programs may cause people to over-detail, killing the fps too.
____

Talking of fps killing objects...



A .csv ball in objectveiwer. 8064 vertexs with 2048 faces !

For a start on routeveiwer, the CPU usage easily craws up to 50% if I place this ball every 100m 4 times, compared to around 30% on the bits with no scenery and just track and grounds. In-game I only noticed dips of 5-8fps however. If anybody thinks they can measure preformance better...here is the ball's .csv file: http://dl.dropbox.com/u/21670777/BVE/ball.csv

I'll try it on my test route. My PC is weaker and therefore I expect more significant differences. Expect feedback in the evening.

CPU and FPS are only very loosely correlated. It's a GPU thing. The better a sim is programmed, the worse a CPU can be without having a negative effect on the FPS. (The better a sim is programmed, the more it does with the GPU)

Yep, my GPU is also not very good, so we'll see.:-D

Leaving aside Chars-Gisors, which sadly is of very poor quality when it comes to operational aspects (signals wrongly plated all along, wrong aspects, etc.), I think I never had below 20FPS, and often I even have >30FPS even on modern routes - and I use an Acer Extensa with Intel Onboard graphics. You don't get far lower than that

Well, I'll be more than happy if you tell me your fps on my route (after it is updated tomorrow).

Not needed anymore...

Scrap all the above, I've found a new method to get the .s to a .x file! I'll explain soon...

Could cause an addon boom... :-D

Yeah, but also will require us to check up addons of newbies coming to this board with something great and finished. That's a problem that arose with EEP also; since there is a possibility to easily convert other files to EEP, there are, on a regular basis, people coming and showing off their great own work which then appears to be plagiarized; the somewhat lash reactions of the EEP community engraved their esteems in the eyes of other railsim communities. That's no argument against making such tools, not at all; it's only a reminder that we will collectively have some form of responsibility. Simply get curious if somebody emerges out of nothing, without having any visible history in any railsim, and shows a ready-to-use addon; ask questions like "wow, how did you solve that and that problem" or "that's great work, what was your approach, do you have some work images? You know, as inspiration?" and maybe check some sites with addons for other sims. We could even prepare some sort of check list; this way we could simply post it when a suspicion of an addon being plagiarized arises and those of us having the time to check one or two points, check and answer.
Also, maybe it would be possible to program the converter so that it leaves some discreet trace? You know, some small syntax error which is ignored by OpenBVE; that's something only people knowing the csv-format can correct. Or even discreeter, nonsense lines like two or three vertices not used?

Don't take me wrong, I find this concept great and I hope it'll work! Maybe we'll even be able to lure some M$-TS-dev here this way...? The inability to migrate ones former work is always a strong counter-argument. Just remember how many people round you used OpenOffice.org back when it only was able to process open formats, and how many now, when it is fully compatible with M$-Office?

And I'd even have a suggestion Another file format that could be of interest for conversion is the file format of LokSim3D. It's xml, but the data stored is very similarly structured as in OpenBVE files. As LS3D has problems with modern hardware, there's regularly people finding, they can't use it anymore after buying a new computer (just like myself, several years ago); maybe we could lure some of them here?

And with all the time wasted on trying to convert s**t to treacle and then back again you could have made a dozen or more objects easily with notepad and paint.net in the happy knowledge that it works.

lol Yeah... But depends on what objects. If somebody has a highly detailed model of some loco, which as we know are weeks of work, how high is the probability, s/he will be interested in rewriting it for BVE?

Quork you seem to speak sense here. The thought of people stealing others working has arisen in the back of my mind before, and really I can't think of much of a way to stop it. A syntax error might be a way, however it would be abit irritating. Perhaps I could connect the tool with a online database to log inputs and outputs? I do agree that there should be a way to track and report and a system to prevent people stealing others work. To be honest this is the only thing that keeps me off the idea of finishing this whole project.

Yes, perhaps and hopefully though it will lure MSTS devolpers into the world of OpenBVE and that'd be a good side. I can remember the day I migrated to OpenOffice when MSWorld become compatable...!

Suggestions! - Love them! LokSim3D...I have heard of that german trainsim. Never managed to move the train and I don't understand german, however I shall take a look soon. No promises yet...

If somebody has a highly detailed model of some loco, which as we know are weeks of work, how high is the probability, s/he will be interested in rewriting it for BVE?

I wouldn't really want to guess the answer. I only use BVE. A friend had msts and I was never impressed with what I saw, so I don't imagine there can be many msts things at all which are better than some of what we have. And I am prepared to bet that my routes are better than any Irish content for msts, but I'm never going to write Ballyfeckin for Microsoft, even if I might get some money out of it.

Anyway, it isn't so much that more trains are needed. That is less important than having more good quality routes. Too much dependance is put on a small number of people who are doing the most demanding job of all. Without new route mileage being added you can certainly book the funeral for the sim, indeed that applies to any sim, msts survives not because of the quality of drive experience or visuals but by sheer numbers of users and available routes to play with.

graymac wrote:
Anyway, it isn't so much that more trains are needed. That is less important than having more good quality routes. Too much dependance is put on a small number of people who are doing the most demanding job of all. Without new route mileage being added you can certainly book the funeral for the sim, indeed that applies to any sim, msts survives not because of the quality of drive experience or visuals but by sheer numbers of users and available routes to play with.

You have a point, just what could be done about it...

Well, I have tried to add something today... now that I am satisfied with all objects, some mileage can be added.:-)

You have a point, just what could be done about it...

To be honest Bruce, it's entirely up to the people out there. I've been involved with BVE for two years now, which makes me a newbie in terms of how long BVE has been around and how many years many of our friends here have spent using it and creating for it.

I do all I can to encourage anyone who wants to try and make stuff to get involved. It isn't particularly difficult to learn the simple mechanics of writing a coded mesh and attaching a piccie on it. See the links to help guides in the knowledge base, none of this stuff's top secret and plenty have gone before and done it.

Plenty of folks have wished for something to replace the old routebuilder program which is now well obsolete. That might churn out more material more quickly, but the quality and character of the work would be seriously compromised. Taking aside the "rat in a drainpipe" railways, which work to different rules, if you asked the membership their fave routes I guarantee NWM is going to come top. The greatest number of members registerd here are resident in the UK, Derryck tell me if I'm wrong. Brits like familiar stuff so not many will often bother with "foreign" railways. So NWM leads the field because of its appeal and quality. That quality has nothing to do with a number-crunching route factory program. It is the result of talent and sheer bloody-hard work.

So the dilemma is that we ALL want to see more quality stuff like NWM. Chance are it was exactly that which gave the sim the WOW! factor and made you decide to play it. Because I'm not going to be ashamed to admit that it was the reason I got interested in the sim. And I found other excellent quality work too. That inspired me to WANT to build a route.

But here's the crunch. You gotta really WANT to do it. It will take a LOT of your time. It will also give back, there's great satisfaction in finishing a good work, like anything else creative. Like when you get that bitch of a guitar lick sorted, and can play it faster and better.

So I admit making it sound too easy in the past. Sorry, I was only trying to encourage a few more to get working and help save our sim (SOS????)
But nobody's gonna do it, it's too hard. Gotta spend time watchin' eastenders and (removed) 'bout ain't got me no routes or trains, aint it a shame.

That, Bruce, in a nutshell, is what's going to be done about it. So let's wait see who's got the balls and who's got the talent.

Well, I cannot check myself for talent, that is up to someone else to decide...

No kidding, we're lacking good routes, but good routes come up with good objects. It goes hand in hand. So anyone should do what suits him/her the best - the best way to achieve quality result is a good cooperation, knowledge, and ENOUGH RESOURCES.

Yeah... I know it from LS3D, too; there's people making great objects but not having no feel at all for routes, and vice versa.

Now to the routebuilder; there are some great routes done with routebuilder, this doesn't automatically mean bad routes. Look on the screenshots of the LS3D boards; the routes over there are great too, although there never has been anything handcoded over there, the editor is quite comparable to routebuilder.

but good routes come up with good objects.

How very true. Which is why it is necessary to learn to make your own objecs, as far as possible, to guarantee originality, quality and character. Some "bought in" objects help to put simple projects together a bit quicker, and give the learners some practice. You won't get the most satisfaction or the best result that way, though.

@Quork There are some good routebuilder examples. But as a proportion, I've seen some awful ones too. And out of the worst work I've ever seen (no names, they should know who they are) the very worst was definitely a routebuilder one. Routebuilder made routes but it didn't necessarily make routebuilders

Yeah sure, but you don't need to, e.g., build each tree, each signal etc. anew. Also, for Germany, I once created a set of points, using the switchgen-program (dunno anymore where I got it); in Germany mostly standardized switches are used. My set is designed to have the look and feel of the switches and on the same time sensibly fit into the 25m-raster of OpenBVE. I don't see why any route dev for Germany should do this work again? The switches look all the same whether in one-horse Jerkwater Town North or in Cologne Central Station, whether in Flensburg/Weiche or in Oberstdorf/Allgaeu. Why shouldn't they then ingame?

I think dennis lances route builder was good, i got me a route out of it But then again you need ALOT of time to do great routes to a good standard, Gray for instance, you only been here for 2 years, if OpenBVE had a builder like railsimulator or trainz you would have tonnes of great routes, but you have a handful of GREAT routes made for OpenBVE, its a shame the process is not faster, but then again OpenBVE always has that "come back and better yourself" feeling to it

im a train builder and im always going back to the 350, not because its not finished, but because i always want to better it