Jump to content

DynDOLOD 3.00 Alpha 182


sheson

Recommended Posts

9 hours ago, sheson said:

Do not post screenshots of text. Copy and paste the text instead.

The large reference bugs means that the engine does not unload the static object LOD for all large references in a cell when it is supposed to. So the large reference full models and their LOD models show at the same. This can manifest as very obvious texture/lighting flicker between the two models. Any ESP overwriting a large reference triggers this bug for the entire cell. Hence the log messages as seen in the screenshot to warn about plugins that are not fully visually compatible with Skyrim Special Edition.

If a plugin (ESM in this case, since it is not supported by ESP) overwrites a large reference, it should also list the large reference in its worldspace RNAM. That is what the game files do. If an overwrite/plugin is removed then whatever the overwrite record  or plugin changes simply does not happen.

Thank you! Did not fully understand though why any of DLCs to the game overwrite already existed RNAM records. In my understanding a RNAM record is a place in a world where the large reference LOD should be placed, but if skyrim.esm beeing overwrited by dragonbord DLC it should mean that all these objects are replace in dragonborn DLC, but they do not change already existed objects in tamriel. So, what are the changes they make in RHAM for?

Link to comment
Share on other sites

13 minutes ago, jared_maxwell said:

Thank you! Did not fully understand though why any of DLCs to the game overwrite already existed RNAM records. In my understanding a RNAM record is a place in a world where the large reference LOD should be placed, but if skyrim.esm beeing overwrited by dragonbord DLC it should mean that all these objects are replace in dragonborn DLC, but they do not change already existed objects in tamriel. So, what are the changes they make in RHAM for?

A reference is placement of on object with position coordinates and rotation for example. A large reference means, the form id of such a reference is listed in the worldspace record RNAM list. 

Large reference are not LOD. A large reference simply continues to show the full model outside the uGridsToLoad cells but only to the distance of the uLargeRefLODGridSize setting.

When records are overwritten by other plugins they typically keep most of the original data and only change a detail. Like if a rock is being moved a bit all but the position information of the record stays the same. Check the CK wiki and also start using xEdit to see how records in plugins work.

In case the rock is a large reference, the RNAM data in the worldspace record of the overwriting plugin simply lists the reference again.

Link to comment
Share on other sites

6 minutes ago, sheson said:

A reference is placement of on object with position coordinates and rotation for example. A large reference means, the form id of such a reference is listed in the worldspace record RNAM list. 

Large reference are not LOD. A large reference simply continues to show the full model outside the uGridsToLoad cells but only to the distance of the uLargeRefLODGridSize setting.

When records are overwritten by other plugins they typically keep most of the original data and only change a detail. Like if a rock is being moved a bit all but the position information of the record stays the same. Check the CK wiki and also start using xEdit to see how records in plugins work.

In case the rock is a large reference, the RNAM data in the worldspace record of the overwriting plugin simply lists the reference again.

'It is not just moving, right? The ref rows are different. REFR0010 beeing replaced with REFR0200 in Dawnguard.

 

xedit.png

Link to comment
Share on other sites

2 minutes ago, jared_maxwell said:

'It is not just moving, right? The ref rows are different. REFR0010 beeing replaced with REFR0200 in Dawnguard.

 

xedit.png

The RNAM lists are one a the few types of data that is merged at run time and do not adhere to the rule of one. Hence xEdit does not order the lists.

Link to comment
Share on other sites

2 minutes ago, sheson said:

The RNAM lists are one a the few types of data that is merged at run time and do not adhere to the rule of one. Hence xEdit does not order the lists.

TYVM!

I am wondering about

<Warning: Overwritten large reference in Lux.esp [REFR:000CE0ED] (places FXMistLow01 [MSTT:00077772] in GRUP Cell Temporary Children of BlackreachWeatherTest2 [CELL:000237D2] (in Blackreach "Blackreach" [WRLD:0001EE62] at 3,3))>

Lux is overwritting some stuff in blackreach at cell 3,3, so according to your explanation the whole cell should be flickering because of the engine bug. Any way how i can teleport myself into that exact cell?

Link to comment
Share on other sites

10 minutes ago, jared_maxwell said:

TYVM!

I am wondering about


<Warning: Overwritten large reference in Lux.esp [REFR:000CE0ED] (places FXMistLow01 [MSTT:00077772] in GRUP Cell Temporary Children of BlackreachWeatherTest2 [CELL:000237D2] (in Blackreach "Blackreach" [WRLD:0001EE62] at 3,3))>

Lux is overwritting some stuff in blackreach at cell 3,3, so according to your explanation the whole cell should be flickering because of the engine bug. Any way how i can teleport myself into that exact cell?

What happens is that object LOD for the large references in this cell won't be unloaded outside the uGridsToLoad and inside the uLargeRefLODGridSize. If there is actual texture flicker to be seen, depends if any of the large references actually have object LOD and if that is the case if the LOD and full model occupy the same 3D space for triangles to z-fight.

Use the cow console command from the main menu, like:
cow blackreach 3 3

Link to comment
Share on other sites

12 minutes ago, sheson said:

What happens is that object LOD for the large references in this cell won't be unloaded outside the uGridsToLoad and inside the uLargeRefLODGridSize. If there is actual texture flicker to be seen, depends if any of the large references actually have object LOD and if that is the case if the LOD and full model occupy the same 3D space for triangles to z-fight.

Use the cow console command from the main menu, like:
cow blackreach 3 3

Well, there is some weird stuff going on with mushrooms out there but it happens either i have enabled dyndolod ouput and lux or not, may be the distance just not enough.

Link to comment
Share on other sites

8 minutes ago, jared_maxwell said:

Well, there is some weird stuff going on with mushrooms out there but it happens either i have enabled dyndolod ouput and lux or not, may be the distance just not enough.

To see large references, turn off LOD with tll in console. Whatever full models remain that area are the large references and IsFullLOD references (dynamic LOD from DynDOLOD for example). That's the area to check.

Link to comment
Share on other sites

2 hours ago, sheson said:

You sure you got the right cell? 6, -1 is inside Whiterun about where Heimskr is preaching. The last plugin Smash Patch.esp overwrites the record already and "unsets" the flags.

Beg your pardon but I didn't specifically show the bugged cell on my imgur pic, I'm just showing that there are some cells that are like that on the smash patch and I intend to fix them all. So how do I fix records like that? Thank you.

Link to comment
Share on other sites

Just now, gmahadhika91 said:

Beg your pardon but I didn't specifically show the bugged cell on my imgur pic, I'm just showing that there are some cells that are like that on the smash patch and I intend to fix them all. So how do I fix records like that? Thank you.

This cell is already "fixed" by the smash patch.esp, which unsets the flags. Otherwise copy a record as overwrite into a new plugin with right click in the left tree view and then unset the flags.

However, Open Cities setting the flags is correct and the smash patch reverting it is wrong. There is no need for terrain in that particular cell, as walled cities typically have references placing normal 3D models acting as the ground.

Link to comment
Share on other sites

12 minutes ago, sheson said:

This cell is already "fixed" by the smash patch.esp, which unsets the flags. Otherwise copy a record as overwrite into a new plugin with right click in the left tree view and then unset the flags.

However, Open Cities setting the flags is correct and the smash patch reverting it is wrong. There is no need for terrain in that particular cell, as walled cities typically have references placing normal 3D models acting as the ground.

So I tried searching the cells like you said, and this came up. Am I searching it right? When searching on xEdit I typed the coordinates with minus (-) just exactly how it is on DynDOLOD's My Location. Is this the problem? If so, how do I fix it?

https://imgur.com/a/FYOTRcP

Link to comment
Share on other sites

9 minutes ago, gmahadhika91 said:

So I tried searching the cells like you said, and this came up. Am I searching it right? When searching on xEdit I typed the coordinates with minus (-) just exactly how it is on DynDOLOD's My Location. Is this the problem? If so, how do I fix it?

https://imgur.com/a/FYOTRcP

The xEdit screenshot shows the landscape record for the cell 0,0 obviously.

If you search for -16,-2 for Tamriel with the Cell Browser it will find the cell for -16-,2 with the form id 000099F4

The last plugins to the right are the winning overwrites. Whatever they do typically counts.

To test if "this" is a problem remove the last overwriting plugin that does "this". Whatever "this" is...

Link to comment
Share on other sites

40 minutes ago, sheson said:

The xEdit screenshot shows the landscape record for the cell 0,0 obviously.

If you search for -16,-2 for Tamriel with the Cell Browser it will find the cell for -16-,2 with the form id 000099F4

The last plugins to the right are the winning overwrites. Whatever they do typically counts.

To test if "this" is a problem remove the last overwriting plugin that does "this". Whatever "this" is...

I found it! Smash patch set it to force hide with the prior record being Skyrim.esm. I don't know where smash patch gets that "idea". Thank yu so much sheson it is now fixed!

Link to comment
Share on other sites

Hi, i'm having a problem with dyndolod 3 alpha, running No Grass in Objects mod aside it all perfectly setup (to the best of my knowledge) and it's all in-game, however the grass is a set distance in my game, changing the ini file for distance does nothing at all, same with skyrim's ini's, the distance is identical, also attempting to change dyndolod's settings in the mcm only affects far away distance, the grass close to me will not reduce with the distance change in the ini,

Disabling dyndolod fixes this, i'm wondering if the grass distance in the ini before running dyndolod is a set distance and you can't change it once dyndolod is ran?

Could anyone clear this up, not sure if this is normal behavior with dyndolod alpha 3's grass lod generation with this mod...

How i did it :

- Install the mod into .NetFramework plugins folder, edit the ini and change UseGrassCache to true, ExtendGrassDistance to true, OnlyLoadFromCache to true, overwrite grass distance to my specified value & DynDOLODGrassMode = 2 (also tried 1)

- Generate grass cache using mo2, full generation finished, no crashes, create mod with cache, make sure it's read by the mod and dosen't regenerate it in the overwrite folder.

- Edit Dyndolod config and set Set GrassLargeReference to 1, change grassdensity to 50 and enable Grass=1

- Run TexGen with the grass option, then proceed to run Dyndolod.

End result is the grass is there, and far away, however the distance setting does nothing.

Very confused...

 

modlist.txt.2021_05_03_14_17_19

Edited by darknightbacca
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...

Important Information

By using this site, you agree to our Guidelines, Privacy Policy, and Terms of Use.