<span style="text-decoration: underline">The SUR model</span>
Ah yes the SUR, the release of the exporter was a cause of great excitement amongst us all, until your shields dropped, that's when the b*tching started and most of you classed the exporter as useless, those of you that complained please take a look in the mirror, that jackass that's looking back at you is what's useless. You jumped in with both feet even though you didn't have bloody clue what you were doing, is it any wonder they didn't work, I knew from my own work on SURs that the Exporters SURs were done correctly, I just couldn't figure out why they didn't work. Multiple component SURs have given me my best results, but the Root was always a problem as it just wouldn't register, then about Nov/06 I realised the problem, concave polygons in the Root of the SUR stopped it from working. Are you starting to see why your SURs never worked, yep by exporting the CMP & SUR as single components (Roots) you rendered them useless yourselves, hundreds of concave polygons in each SUR Root made sure of it, I've been saying this for a while now but most of the custom CMPs that we have are useless in their current form, so model makers get your fingers out of you lazy fat arses and redo your models, OR I’LL DO IT AND TAKE THE CREDIT FOR YOUR WORK
The SUR model needs to be a low poly version of your CMP model, yes that does mean two models folks but that's just for starters as you could have a dozen ms3d's for the one ship (I have 11 for the ship in the above multi coloured pictures, 3 for the lods, 1 SUR, 5 damage meshes, an icon and a cockpit).
I'm using Ship_A again here, if I was to export this as a SUR I would get one of two results, a CTD on launching into space or a Root that doesn't register, why? because the guns and plates create concave angles in the Root, they can either corrupt the SUR completely (hence the CTD) or just stop the Root from working (almost as bad).
Open SUR_A.ms3d instead and you'll notice that there are only 5 groups and no materials, I still have my 3 components but I now only have one group being used for component 1 (Root), I deleted the groups that would give me a broken SUR.
Unfortunately as this model is nothing but primitives it’s of little use in learning how to create a close fitting SUR for your custom models, but don’t discount primitives as they can be used for a SUR if you don’t want to do the extra work to get a close fitting SUR. Take the NCC-1701 ‘Enterprise’ (TOS), a geo-sphere scaled to saucer shape, 3 single stack cylinders for the nacelles and engineering hull and 3 rectangles for the pylons and neck, this will give you a decent multiple component SUR just not a close fitting one.
First thing is load in the model you used for the CMP, now “save as” whatever, so you now have two ms3ds that are identical except for their names, now delete the Materials as you don’t need them with the SUR model. Take a look at the model in the 3D window (flat shaded and wire frame overlay), about two thirds (probably more) of those polygons will be getting deleted, what you will be doing is making a low poly version of the model using some of the vertices from the original, this way you keep the shape of the original but lose most of the polygons.
You’ll have a CMP model that goes from something like this-
To a SUR model similar to this-
The “Groups:” part for the SUR is important as you really only want one group for each component in the SUR model, you can get away with a couple but you can sometimes get errant polygons created where the groups meet.
In the above picture you see the Root from my CMP model, there are several groups and a lot of polygons in this, as it just so happens the main body is the only group I need to keep, the rest are built over for the SUR. The RED dots are vertices in the CMP model that are the main edge vertices, these are the ones that I want to use in the SUR.
Now it’s a case of deleting polygons so you can replace them with one or two big polygons, as I have done in this next picture.
As you can see the areas that had dozens of polygons now only have a couple, the Root had several groups but now only has one, and the poly count has dropped from a couple hundred to a couple of dozen. The RED dots are still the same vertices I identified earlier, they just have fewer polygons joined to them now. You will notice that the back of the models in both pictures are different, the first one has a lot of concave polygons that will destroy the SUR, and so you need to “skim over” these areas. I do this in the same way I did the front, I identify the edge vertices at the back of the model and join them up to the other edge vertices at the front with big polygons.
While concave polygons don’t affect the other components, you do need to lower their poly count by about two thirds.
The above picture shows one of my engine components, the top one is from the CMP while the bottom one is from the SUR, as you can see by the RED dots I’ve done the same thing here as I did earlier and selected the main edge vertices in this component. I’ve done exactly the same with this component as I did with the Root, areas between my desired vertices with more than two polys get deleted and replaced with a couple of big polygons. Even though concave polygons only affect the Root, you’ll notice that the engine intakes have been “skimmed over”, this saves polygons especially when you’re trying to keep it to between 300 and 400 polys with 500 being the absolute MAX for a FIGHTER/FREIGHTER similar in size to a vanilla ship.
Don’t expect to do this in an hour as it can take me the best part of a day to get to this stage. It takes me between 2 & 3 hours just to convert the model for the CMP, and a good 3 to 5 hours to do the SUR, and to date only my station SURs have caused me any problems, all of my ships work beautifully using this method.
<span style="text-decoration: underline">**Edit One**</span>
While doing the SUR for the APS Pegasus
I took a number of screen shots, I hope they help you to understand my methods.
The CMP model has 14 groups that are combined into 6 components (parts), seven of these groups make up component one (the Root) here
they are, I only want to keep some of those polygons, so after a look through each group I find that I only need to keep two
of the groups, the other five can be deleted. Now it’s time to decide which polygons stay and which get deleted from the two groups, after a bit of deleting this
is what I’m left with, regroup the two into one and your ready to start making faces, after 15 minutes work this
is what I have.
Next it’s component two the Engine
, it has two groups and after a little poly deleting I have this
, the small ring of polygons to the right are all that’s left of group 2, and their reprieve is only temporary as I only need them for the outer vertices, and after a spot of face making I have this
, you’ll notice the ring is gone and the end has been closed with a single vertex.
Component three has two groups
, the cage and the glass, the glass is the only group I’ve kept
, it also doesn’t need any faces added to it.
The fins are three separate components, each having only the one group, the bottom fin
like the other two has a lot of polygons, most of which we don’t want in a SUR, so after a little deleting I have this
didn’t leave much did I, eight faces later and the fin is done
, I did the exact same thing to the other two fins.
The end result is this
, a SUR model that has 6 groups that export as the 6 components (parts).
The CMP model has a poly count of 1820, while the SUR model has 304, but the most important thing is the names of the components in both CMP & SUR.
The CMP model has the following names:
are the component names that are added to the CMP.
The SUR model has these names:
Yes they are the exact same names as the components from the CMP, if they’re not then Freelancer wouldn’t be able to assign the appropriate parts of the SUR to the CMP.
Like I said I hope this helps.
Edited by - Bejaymac on 10/25/2007 8:08:49 AM