DynDOLOD 3.00 Alpha 182

[sheson]

13 hours ago, MisterMorden said:

Should I be checking Force Dynamic or would that give me 2 models coexisting?

Edit - would the campfires have to be set as "is full LOD" and/or "persistent" for the previously mentioned mesh rule to work the way I'd like it to?  I haven't looked yet, but the mod author of Embers XD suggested I try setting "is full LOD" flag on the campfires so I'm assuming they didn't do that.

 Just set FarFull and maybe Replace to avoid brief overlap of 2 full models.

[z929669]

3 hours ago, MisterMorden said:

Should I be checking Force Dynamic or would that give me 2 models coexisting?

Edit - would the campfires have to be set as "is full LOD" and/or "persistent" for the previously mentioned mesh rule to work the way I'd like it to?  I haven't looked yet, but the mod author of Embers XD suggested I try setting "is full LOD" flag on the campfires so I'm assuming they didn't do that.

Please post back with the mesh rule mask you use for this, as I will add it to my own if it works.

[MisterMorden]

13 minutes ago, z929669 said:

Please post back with the mesh rule mask you use for this, as I will add it to my own if it works.

You got it.

[MisterMorden]

Really weird, but the best I can seem to get with the giant campfires is that sometimes the flames appear on the full model lod and sometimes it's just the glowing embers with no flames.  The same campfire will sometimes have flames and sometimes not when viewed at different times.  I tried using "replace" instead of "unchanged" however then the flames didn't ever seem to appear on the lod model nor would they erupt when I got close...just remained a glowing pile of logs.  Also really weird is that even using "unchanged,"  every once in a while the flames wouldn't appear once the cell attached.  It's like my lod aren't fully initializing nor unloading 100% of the time.

Clicking on the lod models using MIC did indicate they were from Dyndolod.

[sheson]

45 minutes ago, MisterMorden said:

Really weird, but the best I can seem to get with the giant campfires is that sometimes the flames appear on the full model lod and sometimes it's just the glowing embers with no flames.  The same campfire will sometimes have flames and sometimes not when viewed at different times.  I tried using "replace" instead of "unchanged" however then the flames didn't ever seem to appear on the lod model nor would they erupt when I got close...just remained a glowing pile of logs.  Also really weird is that even using "unchanged,"  every once in a while the flames wouldn't appear once the cell attached.  It's like my lod aren't fully initializing nor unloading 100% of the time.

Clicking on the lod models using MIC did indicate they were from Dyndolod.

... and also show that the full model is being used.

If there are no flames in that case, it could be because of some fade INI setting or in the NIF or because the base record in the DynDOLOD plugin is missing something. Compare to the original base record.

If these troubleshooting hints do not get you anywhere, set the reference drop down to "Original" and test what happens then.

[MisterMorden]

26 minutes ago, sheson said:

... and also show that the full model is being used.

If there are no flames in that case, it could be because of some fade INI setting or in the NIF or because the base record in the DynDOLOD plugin is missing something. Compare to the original base record.

If these troubleshooting hints do not get you anywhere, set the reference drop down to "Original" and test what happens then.

Thank you.  MIC showed the name of the model (giantcampfire01burning.nif) when I clicked on the lod.  I also did try to greatly increase LOD associated ini values but there was no change.  I did not, however, compare the dyndolod plugin's base record to the original...I will try to do so.  Any suggestions on what differences I should be looking for?

I will also try using "original" for the reference setting.  Thanks again for always trying to help with our issues!

[sheson]

56 minutes ago, MisterMorden said:

Thank you.  MIC showed the name of the model (giantcampfire01burning.nif) when I clicked on the lod.  I also did try to greatly increase LOD associated ini values but there was no change.  I did not, however, compare the dyndolod plugin's base record to the original...I will try to do so.  Any suggestions on what differences I should be looking for?

I will also try using "original" for the reference setting.  Thanks again for always trying to help with our issues!

"LOD associated ini" does tell us which INI setting you are talking about. I can only hope it doesn't mean INI settings for static object LOD under TerrainManager, since they are obviously irrelevant.

We have no idea what the original base record looks like - since no logs, no screenshots or form id was provided - anything that might be different but the form ID and editor ID.

[jiggajigga82]

Not entirely sure if this is the right place to ask, but is there a specific script that's used in DynDOLOD to detect mesh root blocks being NiNodes? I'd like to root them out in xEdit if they're a proven cause of CTDs.

[z929669]

I got the OpenGL error running TexGen. Logs attached.

No bugreport. I use AMD, so I will check my drivers next, but they have been same drivers for months now and no issues previously.

Logs.7z

---

Looks like I do have a driver issue maybe caused by Windows updates. I'm removing/reinstalling latest recommended AMD drivers now.

[sheson]

48 minutes ago, jiggajigga82 said:

Not entirely sure if this is the right place to ask, but is there a specific script that's used in DynDOLOD to detect mesh root blocks being NiNodes? I'd like to root them out in xEdit if they're a proven cause of CTDs.

DynDOLOD is all compiled code, so no pas scripts. Just use the list from the summary.

[jiggajigga82]

2 minutes ago, sheson said:

DynDOLOD is all compiled code, so no pas scripts. Just use the list from the summary.

Alright, because I've found some that didn't show up in the summary. Does that mean I shouldn't be concerned with them?

[sheson]

1 minute ago, jiggajigga82 said:

Alright, because I've found some that didn't show up in the summary. Does that mean I shouldn't be concerned with them?

Depends on the type of NIF and how it it used. You do not need to worry about any that are not reported. https://dyndolod.info/Messages/Root-Block

[z929669]

I can confirm that AMD Adrenaline 22.5.1 drivers resolved the issue.

For kicks, I first factory reset my display drivers and tried AMD Adrenaline 22.11.1, and this indeed caused the OpenGL issue, so I can confirm that as well

interestingly, TexGen processing of my same Step LO has increased approx. 4x from last version I used (alpha 98 and many previous versions). It used to reliably take 1:40-1:50, and now its upwards of 5 minutes processing. Generating billboards processing increased by a significant amount, but Creating textures portion at the end was the brunt of the increase.

FYI: logs from the successful run for direct compare to my last if interesting.

Logs.7z

[jiggajigga82]

7 minutes ago, sheson said:

Depends on the type of NIF and how it it used. You do not need to worry about any that are not reported. https://dyndolod.info/Messages/Root-Block

Perfect, thanks!

[MisterMorden]

2 hours ago, sheson said:

"LOD associated ini" does tell us which INI setting you are talking about. I can only hope it doesn't mean INI settings for static object LOD under TerrainManager, since they are obviously irrelevant.

What I meant was ini values with LOD in the names.  I didn't mess with anything under [TerrainManager], but I did try increasing fLightLODMaxStartFade, fLightLODStartFade, fLODFadeOutMultItems, fLODFadeOutMultObjects, fLODFadeOutMultSkyCell (although I didn't hold out much hope with these).  I wouldn't push my fBlockLevel0, 1 & Maximum past where I have them now (75000, 165000 & 335000 respectively).