Tutorial about the LOD system in OFP

[Matthijs 40]

I created a tutorial, which was inspired by the fact that many add-on/model makers do not seem to give much attention to the resolution LODs for their Operation Flashpoint models. The reasoning behind it is probably something like: "Why spend hours working on models you don't need anyway?"

In a certain way, they are right. You don't need to add LODs to your work, to create an add-on that seems to work ok. But the truth is: if you do so, the OFP engine has to work twice as hard, and use lower levels of detail for all other models to compensate for the shortcomings of your work.

This tutorial has two objectives:

1) Introduce you to how resolution LODs help improving your gaming experience, and convince you that the LOD system is an indespensable feature.

2) Help you in setting up resolution LODs in your work.

[ READ THE TUTORIAL ]

[orson 0]

Great tutorial

Thank you for taking the time to make it

[simba 0]

thanks that's really helping

[SelectThis 0]

Very nice, should be required reading for all budding modellers

LOD's are very important.

I like the analysis of the BIS model versus your custom one. Wonder how hard it would be to generate a equation to help with LOD number...

SelectThis

[Lester 0]

Nice Turorial, but one thing is definetly wrong :

Quote[/b] ]# If anything is behind another object or hill, don’t render it.

- behind hills OFP render as no hill will be there.

- it seems that only behind closed objects (this one you can't get in) other objects will not be rendered

[Matthijs 40]

Nice Turorial, but one thing is definetly wrong :

Quote[/b] ]#  If anything is behind another object or hill, don’t render it.

- behind hills OFP render as no hill will be there.

- it seems that only behind closed objects (this one you can't get in) other objects will not be rendered

OK, thanks for the feedback.

But: "behind hills OFP render as no hill will be there."

I really don't understand what you are tying to say. Can you explain yourself a bit more?

[raedor 8]

he means that ofp does not check if there is a hill or not. ofp renders it anyway.

"OFP renders behind hills as if there is no hill"

good tut btw.

[shinRaiden 0]

Quote[/b] ]- it seems that only behind closed objects (this one you can't get in) other objects will not be rendered

I'm not so certain about this. There seems to be visual occulsion on certain things that should be transparent, ie 'invisible' stuff not seen through windows. I think that this is CPU processed, just not displayed on the screen because it's not sent properly to the video card. Don't know for sure, just a hunch though.

[Matthijs 40]

Quote[/b] ]- it seems that only behind closed objects (this one you can't get in) other objects will not be rendered

I'm not so certain about this. There seems to be visual occulsion on certain things that should be transparent, ie 'invisible' stuff not seen through windows. I think that this is CPU processed, just not displayed on the screen because it's not sent properly to the video card. Don't know for sure, just a hunch though.

This is an issue with PAA files, not a performance improving feature.

[KaRRiLLioN 0]

Is there a way to make O2 automatically adjust the # vertices for a particular RLOD? If not, how does one best manually adust this? Do you just copy and paste your entire 0.000 RLOD and paste it into say a 1.000 RLOD and then just nip and tuck by deleting vertices? Also, do the main named selections have to stay intact for each RLOD, i.e. OtocVez, etc. or can you do like in Geo LOD where you can simplify things down to "blocks".

I like the gist of the tutorial explaining how to calculate, but I'd like to see a specific example or tutorial explaining how you actually go about adjusting each RLOD in O2.

At least it makes more sense to me now why I sometimes see vehicles with no textures when they're 500 meters away in game, most likely, I suppose because they forgot to add textures to their respective RLOD.

[orson 0]

I would answer yes to all your questions .

  Try to imagine that what you make in , lets say the 3.00 LOD will be what you would expect to see at 300metres on an object .

  Pasting the previous LOD into the next highest and then face culling ensures that names are carried through all LOD's correctly , just make sure you get the first one right

    eventually the last LOD will be a very simplfied version required to be under 300 faces , this is for the shadow .

This LOD also requires   LodNoShadow   with a value of 1 in the Named Properties .

http://ofp.info/brsseb/

Simple tutorials explaining the basics .

[KaRRiLLioN 0]

Thanks, I've actually already thoroughly scoured the tutorials in that link you gave me, but he doesn't really go into RLOD's.

In O2, I've noticed there is a menu "Structure" and under that "Face Cull (from pin)" and also "Multiple Face Cull".

Is using either of those adviseable, or should I go though and delete a vertex at a time?

[orson 0]

I really havnt played with those options yet .

May have to try them and see

[Master_Chief 0]

-If anything is behind another object or hill, don’t render it.

-ofp does not check if there is a hill or not. ofp renders it anyway.

That's kind of like 'does the fridge light go out when I close the door?' There isn't really any way to know for sure.

[colonel_klink 0]

Thanks, I've actually already thoroughly scoured the tutorials in that link you gave me, but he doesn't really go into RLOD's.

In O2, I've noticed there is a menu "Structure" and under that "Face Cull (from pin)" and also "Multiple Face Cull".

Is using either of those adviseable, or should I go though and delete a vertex at a time?

The face culling you mention are internal for O2. If you load and object and click in the 3d wireframe view then press the F key (for face culling) you will see what I mean. Face culling is an aid to modelling in that view, not culling (deleting faces or points) for lower LODS. I may be wrong but this is all it has ever done for me.

What face cull from pin or multiple face cull does I have no idea. Because of this I do the lower lods manually. It takes time but as mentioned below at least all of the parts wilol be correctly named from one lod to another.

I have tried auto lod methods in programs like trueSpace and Max but they have never been particularly successful.

[csj 0]

Quote[/b] ]What face cull from pin or multiple face cull does I have no idea.

Face cull from pin highlights those faces that are visible from the pins posi.

ie Editing View Pilot Lod:

position pin in pilots seat then select face cull. It will show you which faces are visible from the pilot seat.

So you should be safe to delete the rest.