Jump to content

Fallout 4 Optimization and Performance Systems Explained


Recommended Posts

Hello can someone tell me if moveable statics, in this case an effect, can break the precombine, PreVis cluster aswell? Its about XPRI reference here:

GSMist03Dim is part of the PreVis Reference Index for whatever reason ???

unknown.png

Link to comment
Share on other sites

Not so much break as cause undesirable effects. Objects in XPRI are rendered by the occlusion system based on information in the previs (uvd) file. If you move an object in XPRI it will be rendered based on where it originally was (so will appear/disappear incorrectly).

Link to comment
Share on other sites

A fun thing to do is add an object to XPRI. As it has no info in the uvd it won't be rendered, but it is still placed in the world and has collision - just invisible.

  • Haha 1
Link to comment
Share on other sites

Alright another thing i dont understand. If any object which is a precombined object was refernces and set "Persistent" by the CK that object

  1. looses its "[Placed Object]" state in FO4Edit
  2. will from now on displayed "Placed Object" instead

Any ideas?

 

unknown.png

It might not be crystal clear so some more notes on this:

- lets say i would like to have one object linked to a static which is precombined

- the CK will flag the object reference "persistent"

- by adding the persistent flag the object is no longer marked [Placed Object] and at least in FO4Edit shows "Placed Object" instead. Im pretty sure it will still break the precombine and maybe is a display bug in FO4Edit ?

Link to comment
Share on other sites

This is all explained in my deep dive article - but anyway, it depends on whether the reference is in an interior or exterior cell.

Precombines in Interior cells do NOT care about the persistence flag. Many persistent references are precombined (which is ridiculous but - Bethesda). Changing that flag does not effect precombines itself.

Precombines in exterior cells do not care if a precombined reference is moved to another cell (by overriding it's coordinates OR setting it to persistent). Such a reference is however just ignored.

This means that reference will now be rendered as if it is NOT precombined, HOWEVER the precombines it was in will still be rendered (moving it does not break precombines) so it will also still appear in it's original location as well.

If it was just changed to persistent then that may cause z-fighting. It really depends on how the occlusion system decides to render the precombine it was originally in.

All very messy 

And yes, I think there are display bugs in the way xEdit handles the [] and I brought it up over a year ago and was told "it is working as designed".

I am not 100% sure what that design is, but it seems to be "the ref is in the XCRI of the cell it is a child of".

So making it persistent moves it to the persistent cell so it is not in [] in that, however it is still in the original cell's XCRI so "should" still have []. If you move a reference to another cell by it's coordinates it keeps it's [] in the original cell. So it is inconsistent (to us) but consistent with the design...

So I feel I cannot rely on the [] meaning 'it is in some precombine somewhere' - which is what I really want to know.

Link to comment
Share on other sites

6 hours ago, PJMail said:

"it is working as designed".

It is working as designed but the design does not work all time ...

  • Haha 1
Link to comment
Share on other sites

Don't say that too loudly - the sky may fall.

Though I am extremely grateful for xEdit. The CK is so good in some respects but so crude in others. Horses for courses as they say. I mostly work in xEdit so it is obvious where my loyalties lie...

Link to comment
Share on other sites

Seriously FO4Edit, xEdit is working 99.5 % of the time. And the CK is great as well.

I discovered like 2 bugs for Fo4Edit till today in the last 4 years its absolutely brilliant.

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...