CMP to SUR Conversion Tests
-
LS -
OK, never mnd anyone else, I like what I see, so gimme gimme gimme!!
Sushi-
The method used to get round the 18 group limit will not work with this, because it combines many or all of the original ship parts into one group.By this method, the ship will not break up into those separate groups when it is destroyed either. But this is not important for large ships because most are set to just disappear instead of âshatterâ.
Everyone-
The key is to separate your model into sensible component groups, up to 18. This has been enough for almost all of my models.And by doing it this way you should have enough real ship groups to ensure that you do not need to make any dummy parts, which I hate and detest!
So by separating ship booms and wings etcetera into one per group, LSâs sur generator will wrap each one in its own sur, and will leave the spaces between them empty, as they should be.
And when the ship is destroyed, those booms will break away and spiral off! Nice!
-
Sushi-
The method used to get round the 18 group limit will not work with this, because it combines many or all of the original ship parts into one group.Ya, it definitely does not do that. Thatâs why we use this method to hitbox our 7 piece vmeshdata spliced ISD. I think it would work
-
**Here are some test results so far.
Here is a pic of NX01 after I regroup it from 62 to 10 groups, including the sur overlay. Each part of the mesh was regrouped to fit into a specific section. When exported it was done as 10 parts. To do this in Milkshape simply select all of the groups you want to have together and regroup.
NX01 in HardCMPHere is a video of it being tested in-game. Notice how I can fly thru the central openingâŚ
[wmp=640,480]http://digitalbrilliance.org/Videos/SUR_Export04.wmv[/wmp]**
-
Amazing LS, simply amazing.
This work will make everyones life so much easier with modding. I cannot wait to see what else you can do with this (or is there anything else??)
-
Sushi, W0dk4 -
Sorry, must have confused that mehtod with the other âzomgbay thingâ that grouped all ship groups into the 1st group and used dummy groups!Iâll download it and have a look at yours now.
LS -
I notice that the saucer is shown in dull white whereas the other parts of the NX-01 are bright white. HardCMP also showed âgoodâ sur parts in bright white, and âsuspectâ and non-working sur parts in dull white, can you how did it do that? -
StarTrader wrote:
Sushi, W0dk4 -
Sorry, must have confused that mehtod with the other âzomgbay thingâ that grouped all ship groups into the 1st group and used dummy groups!Iâll download it and have a look at yours now.
LS -
I notice that the saucer is shown in dull white whereas the other parts of the NX-01 are bright white. HardCMP also showed âgoodâ sur parts in bright white, and âsuspectâ and non-working sur parts in dull white, can you how did it do that?Thatâs the multi-part SUR. The Freeworlds guys are so insane they have 100k+ models and needed a technique for CMPs, too
-
I had no trouble with it at all on the saucer as u can see in the video. It looks alot darker in the pic than it did on my machine, most likely jpg artifacting.
-
FriendlyFire wrote:
StarTrader wrote:
âŚ
LS -
I notice that the saucer is shown in dull white whereas the other parts of the NX-01 are bright white. HardCMP also showed âgoodâ sur parts in bright white, and âsuspectâ and non-working sur parts in dull white, can you how did it do that?Thatâs the multi-part SUR. The Freeworlds guys are so insane they have 100k+ models and needed a technique for CMPs, too
What do you mean FF, I donât understand either part of your reply old friend?
LSâs utility (this one) will make the sur from the cmp, no?
But my question is how HardCMP showed good sur parts in bright white and bad ones in dull white/grey.
It seemed consistent. It was a good clue when my surs were going to be bad even before I tested them. So I thought it was a deliberate method in HardCMP and maybe LS had done the same?
-
What I meant is that the technique you described with dummy models that put all parts in the root group is for large SURs, not CMPs. The Freeworlds people had to develop a technique to work around both SUR and CMP group limits, but the CMP workaround doesnât change how groups work (if there should be 36 groups, there will be 36 groups, which means that the models will work fine with LSâs utility).
-
FriendlyFire wrote:
What I meant is that the technique you described with dummy models that put all parts in the root group is for large SURs, not CMPs. The Freeworlds people had to develop a technique to work around both SUR and CMP group limits, but the CMP workaround doesnât change how groups work (if there should be 36 groups, there will be 36 groups, which means that the models will work fine with LSâs utility).Ah, right - but they also modified the cmp in the same way to add the dummy parts that matched the sur names, no?
Anyhow, yes the Freeworlds cmp method obviously is better and those cmpâs can hopefully then be used directly by LSâs 1024-group generator!
Wunderbar!
Well - come on then, LS, where is it?! Eh?!
-
**Hold your horses old man, it will be ready when itâs ready. BTW did you get that hedge trimmer I sent you?
Just for you, a model from Freeworlds, the X Wing (courtesy of Sushi), as you can see it isnât a perfect fit. If you want a form fitting sur, then you will need to build it by hand. This is a general purpose tool and isnât meant for perfectly fitting sur files unless the model is prepped for it.
Still having occasional sur corruptions, until I get them cleared out I wonât be releasing an update.
The testing is going good so far though, at the present all of the sur files have shown up properly in Hard CMP. Hopefully I will release a working version soonâŚ**
-
Hedge trimmer - nope!
But donât send a replacement yet, Iâll just blunt it!
Yep, thatâs pretty good.
For myself I would make the engines and weapons separate groups to get a close-fitting sur for it.
Tip - Could also remove the weapon groups and export a temporary cmp so the weapons would not generate sur parts, then restore the original cmp for the game. As long as you donât change any group names all should be well.
Ah - donât forget my optional shield bubble thingy, huh? (Donât worry about what anyone else wants, I donât care, theyâre all inferior!)
Just filling in timeâŚ
âInfamy! Infamy! (Theyâve all got it in fer me!)â - Kenneth Williams, âCarry on CleoâLike it? Oh. OK, hereâs another then:-
âArabs are very intense people - they do everything in tents!â - Charles Hawtrey, âCarry on, Follow That CamelâAnd one more for the recordâŚ
âNo offence meant, just feeling a gay person - Uh-oh⌠yep, he sure is offended!â - StarTrader, âAvinâ a Laff!â -
LancerSolurus wrote:
**Hold your horses old man, it will be ready when itâs ready. BTW did you get that hedge trimmer I sent you?Just for you, a model from Freeworlds, the X Wing (courtesy of Sushi), as you can see it isnât a perfect fit. If you want a form fitting sur, then you will need to build it by hand. This is a general purpose tool and isnât meant for perfectly fitting sur files unless the model is prepped for it.
Weâre going to be making a bunch of âserver sideâ hitboxes, to improve performance, and from what I see there, the hitboxes your program is making is perfect for that. I really cannot wait to use it
You are doing an amazing thing for the FL Community, LS, I cannot say that enough!!**
-
ST puts his white-bearded face round the doorâŚ
Still no sign of LSâs multi-part sur generator.
-
Sushi wrote:
Weâre going to be making a bunch of âserver sideâ hitboxes, to improve performance, and from what I see there, the hitboxes your program is making is perfect for that.Does that work? Server and client having different .surs?
-
If it does work then it opens yet more cheating.
Does it?
Iâm keen to try it out! lol
While Iâm waiting for LSâs generator!
-
Yup. However it will only affect the NPC ship hitboxes as well as how they perceive hitting a player ship. PvP interaction will be unchanged. The reason why weâre doing this is that weâve suspected more polygons in a hitbox ââ> more calculations â> increase CPU usage ----> drop in performance.
-
**Thanks for the encouragement, still doing alot of tests and trying different layouts for the sur file. Multiple meshes in a single cmpnd part is being the problem. This is a by-product of the way the model is exported. Thinking about making a cmp modifier so it can split the mesh properly into parts. This is due to that the models that are modular in design wrap tons better than an all in one mesh.
Note to model converters, regroup your models before exporting to make the overall shape similar to a sphere, cube or cylinder for each groupâŚ**
