LF111 Posted February 17, 2025 Posted February 17, 2025 I don't think Vortex can do what MO2 does there, but I am able to just check the files. I downloaded a texture viewer (WTV) to be able to see the .dds files, and Data\textures\dyndolod\lod\dyndolod_tamriel.dds is blank. It's telling me that the DX10 format is unknown. I tried in GIMP as well and got a similar warning about an unsupported DXGI format. I can check the files in the actual folder, I just need to know what is meant by timestamp. If it's referring to date modified, they don't match (although it might be close enough, based on what you mean by "about the same"). The .bto has a date modified of 2/17/25 at 1:15 AM, while the texture has a date modified of 2/17/25 at 1:09 AM (though the "date created" is listed as 2/17/25 at 1:17 AM... after it was modified?) Sorry if this is not at all what you mean by timestamp. In case it helps, I've uploaded the texture .dds file to the folder.
sheson Posted February 17, 2025 Author Posted February 17, 2025 5 minutes ago, PRieST said: Started DynDOLOD second time without the error. 030090B4 is not overwritten. I know DynDOLOD is based on the most recent xEdit version - with that version I get the warnings regarding the FishingModsMerged.esp which wasn't the case with earlier xEdit/DynDOLOD versions. Do you know how harmful these are? I think they were from Simple Fishing Overhaul before I merged it with some other mods. Same goes for the mentioned navmesh error which is new - all mentioned FormID from the log are not overwritten, so is this a false positive? I could provide the new logs regarding these questions (from the successful run) if needed. I am not sure what warnings in FishingModsMerged.esp you are referring to. There aren't any in the uploaded logs I can find. The Navmesh error is unrelated and should be of no consequence to LOD generation. I believe the missing triangle #382 is added by the unofficial patch, you may have a plugin undoing that change.
PRieST Posted February 17, 2025 Posted February 17, 2025 17 minutes ago, sheson said: I believe the missing triangle #382 is added by the unofficial patch, you may have a plugin undoing that change. Haven't seen any conflicts in my whole loadorder. And es said before, wasn't mentioned with earliere versions of xEdit/DynDOLOD. 24 minutes ago, sheson said: I am not sure what warnings in FishingModsMerged.esp you are referring to. There aren't any in the uploaded logs I can find. Oh, thaught it was mentioned in that one, too. I'll take a look myself and if I need further guidance I'll ask again.
sheson Posted February 17, 2025 Author Posted February 17, 2025 1 hour ago, LF111 said: I don't think Vortex can do what MO2 does there, but I am able to just check the files. I downloaded a texture viewer (WTV) to be able to see the .dds files, and Data\textures\dyndolod\lod\dyndolod_tamriel.dds is blank. It's telling me that the DX10 format is unknown. I tried in GIMP as well and got a similar warning about an unsupported DXGI format. I can check the files in the actual folder, I just need to know what is meant by timestamp. If it's referring to date modified, they don't match (although it might be close enough, based on what you mean by "about the same"). The .bto has a date modified of 2/17/25 at 1:15 AM, while the texture has a date modified of 2/17/25 at 1:09 AM (though the "date created" is listed as 2/17/25 at 1:17 AM... after it was modified?) Sorry if this is not at all what you mean by timestamp. In case it helps, I've uploaded the texture .dds file to the folder. You need to use programs or plugins that support BC7 texture compression format. IrfanView has support for example. However, the problem seems to be missing textures, e.g. a file not found. Filemanagers like Windows Explorer typically show a filetime, which usually the last time a file has been modified. The creation time is when DynDOLOD saved it. The modified time is when Texconv converted it to BC7. According to the provided log, anything within 8 minutes is probably from the same generation: [01:31] [Tamriel] Executing "C:\Skyrim DynDOLOD\DynDOLOD\Edit Scripts\LODGenx64Win.exe" --inputfile "C:\Skyrim DynDOLOD\DynDOLOD\Edit Scripts\Export\LODGen_SSE_Export_Tamriel.txt" --logfile "C:\Skyrim DynDOLOD\DynDOLOD\Logs\LODGen_SSE_Tamriel_log.txt" [01:42] Creating texture atlas C:\Skyrim DynDOLOD Output\textures\DynDOLOD\LOD\DynDOLOD_Tamriel.dds .. [07:17] Waiting for LODGenx64Win.exe generating object LOD for Tamriel. [08:25] [Tamriel] Generating Pre-Computed Occlusion TVDT Having all that explained, missing textures are not the issue it seems. The uploaded diffuse texture atlas has normal map textures included (the *_n.dds textures), which are usually blue and can look like missing texture color in the game. They should only be on the DynDOLOD_Tamriel_n.dds. Check the data folder for the textures\lod\wrbuildingslod01.dds. It should be from the TexGen output. Verify that the file in the data folder has the same filesize and modfied time as the file in the TexGen output mod, or better check their CRC32 to be equal (for example with the right 7zip CRC sha option if you have that installed for example. It should not be blue. Upload both textures\lod\wrbuildingslod01.dds and wrbuildingslod01_n.dds if you want me to check.
sheson Posted February 17, 2025 Author Posted February 17, 2025 10 minutes ago, PRieST said: Haven't seen any conflicts in my whole loadorder. And es said before, wasn't mentioned with earliere versions of xEdit/DynDOLOD. Oh, thaught it was mentioned in that one, too. I'll take a look myself and if I need further guidance I'll ask again. I suggest to ask about those on the xEdit discord.
PRieST Posted February 17, 2025 Posted February 17, 2025 31 minutes ago, sheson said: I suggest to ask about those on the xEdit discord. Thaught the same as you confirmed it will not cause any LOD related issues. Thanks.
CarlWG2 Posted February 17, 2025 Posted February 17, 2025 8 hours ago, sheson said: The crash log indicates a problem with an NPC (and their hair physics?) probably related SKEE64.DLL (from Racemenu). The NPCs happen to be in a cell which DynDOLOD.esp and Occlusion.esp are the last plugins to overwrite. Troubleshoot and fix the CTDs that happen without DynDOLOD in the load order. No crash log without DynDOLOD in the load order was provided for comparison. Test without SKEE64.dll / Racemenu. Test with a new game. If it only happens while load existing saves. it might be related to racemenu data in the save and/or SKSE co-save. You can clean (most of) the data of mods from the save and co save by temporarily disabling the mod and making a new save. Make sure to check comments/forum related to Racemenu and hair physics mods for similar reports, updates etc. Check NPC 0x00013B9B in xEdit and the involved plugins overwriting/patching it. Skyrim.esm -> unofficial skyrim special edition patch.esp -> cutting room floor.esp -> AI Overhaul.esp -> PAN_NPCs.esp -> AI Overhaul - Cutting Room Floor Patch.esp -> Olfina replacer.esp -> Bashed Patch, 0.esp -> Synthesis.esp Once the CTD without DynDOLOD are fixed, generate LOD for that load order. In case there are CTDs with only DynDOLOD output activated, upload new logs and crash logs. Thank you for the quick response. If desired, I can provide additional crash reports from this same save. I load the save, attack a guard in Whiterun, die, and then the game CTD on autoload. Each time the crashlog has different information. The one I posted happened to have Olfina, but the two tests I just did had different references to Stormcloak Soldiers and no reference to Olfina (all produced from same save...load, attack same Whiterun guard, die). As noted, if I uncheck "DynDOLOD is active" in the MCM, the crashing does not happen on autoload. I'll do the following next: I will try disabling Racemenu and the physics hair distribution mod I am using and then attempt to recreate the issue with a new character (as well as with the existing save used for the report I posted, cleaned as needed). As requested, I'll also take a look in SSEEdit at Olfina to see if anything is unusual. I will also attempt to get a crashlog from when DynDOLOD is disabled. That is harder to do as the game doesn't crash often when DynDOLOD is disabled/not installed. I am leaving soon for a trip so my response will be delayed a few days. Thank you!
sheson Posted February 17, 2025 Author Posted February 17, 2025 34 minutes ago, CarlWG2 said: Thank you for the quick response. If desired, I can provide additional crash reports from this same save. I load the save, attack a guard in Whiterun, die, and then the game CTD on autoload. Each time the crashlog has different information. The one I posted happened to have Olfina, but the two tests I just did had different references to Stormcloak Soldiers and no reference to Olfina (all produced from same save...load, attack same Whiterun guard, die). As noted, if I uncheck "DynDOLOD is active" in the MCM, the crashing does not happen on autoload. I'll do the following next: I will try disabling Racemenu and the physics hair distribution mod I am using and then attempt to recreate the issue with a new character (as well as with the existing save used for the report I posted, cleaned as needed). As requested, I'll also take a look in SSEEdit at Olfina to see if anything is unusual. I will also attempt to get a crashlog from when DynDOLOD is disabled. That is harder to do as the game doesn't crash often when DynDOLOD is disabled/not installed. I am leaving soon for a trip so my response will be delayed a few days. Thank you! Upload the other crash logs, also the one without DynDOLOD being active. That would be the crash to fix first to rule out any interaction that would otherwise not happen. Remember, troubleshooting is to remove everything else that is not required for the issue to occur until only the required parts to remain. When investigating crashes while DynDOLOD output is active, make sure it was generated for that exact load order. If you want to troubleshoot crashes with DynDOLOD active further, test what happen if you just disable the DynDOLOD.DLL. Ignore any error messages in the game related to that and do whatever you do to trigger the crash to see if it has any affect. Based on that result we can try narrowing it down further. Reply whenever you are ready, no probem.
rustywd Posted February 17, 2025 Posted February 17, 2025 Recently added The Gray Cowl of Nocturnal - 10th Anniversary to my load order and since it adds a new world needed to re-run dyndolod 3, it had been a while since I last ran it so an update was required (DynDOLOD 3.0 Alpha-187). After updating it ran fine but there is something seriously wrong with Atlas maps (I think that is what you call them). In several areas the _n map is swapped with the color one. I am seeing this in several of them but not all, DynDOLOD_DLC2SolstheimWorld, DynDOLOD_Falskaar, DynDOLOD_MarkarthWorld, DynDOLOD_PalePass, DynDOLOD_Sovngarde, DynDOLOD_WyrmstoothWorld all have the issue, ironically the new ones that were added (not listed) do not have the issue. Here is the maps for DynDOLOD_DLC2SolstheimWorld, where you can clearly see the color and purple versions for a few objects are swapped and placed on the wrong map. Since none of those areas had any changes the only thing that changed for those was using the new dyndolod version. In game this results in the map showing purple like missing textures and distant lods doing the same. https://imgur.com/a/OZ4EtdE Been using Dyndoload for years and never had an issue previously, this is my first time posting so if I did it wrong, my apologies. Thanks!
sheson Posted February 17, 2025 Author Posted February 17, 2025 29 minutes ago, rustywd said: Recently added The Gray Cowl of Nocturnal - 10th Anniversary to my load order and since it adds a new world needed to re-run dyndolod 3, it had been a while since I last ran it so an update was required (DynDOLOD 3.0 Alpha-187). After updating it ran fine but there is something seriously wrong with Atlas maps (I think that is what you call them). In several areas the _n map is swapped with the color one. I am seeing this in several of them but not all, DynDOLOD_DLC2SolstheimWorld, DynDOLOD_Falskaar, DynDOLOD_MarkarthWorld, DynDOLOD_PalePass, DynDOLOD_Sovngarde, DynDOLOD_WyrmstoothWorld all have the issue, ironically the new ones that were added (not listed) do not have the issue. Here is the maps for DynDOLOD_DLC2SolstheimWorld, where you can clearly see the color and purple versions for a few objects are swapped and placed on the wrong map. Since none of those areas had any changes the only thing that changed for those was using the new dyndolod version. In game this results in the map showing purple like missing textures and distant lods doing the same. https://imgur.com/a/OZ4EtdE Been using Dyndoload for years and never had an issue previously, this is my first time posting so if I did it wrong, my apologies. Thanks! I moved you post to a thread where we already seem to be dealing with the same issue. See https://dyndolod.info/Official-DynDOLOD-Support-Forum#Post-Logs which TexGen and DynDOLOD log and debug to upload. Then, if possible try to identify at least one of the affected textures and check if the source textures which should be in the TexGen output are correct or already wrong. Assuming none of the LOD textures are from other mods.
Zorcan Posted February 17, 2025 Posted February 17, 2025 Does it matter that in the most recent update to DyndoLOD, Alpha-187, that TexGen is still reporting as Alpha-186. The DynDOLOD executeable is Alpha-187 but the accompanying TexGen executeable is Alpha-186. Will that pose any issues?
sheson Posted February 17, 2025 Author Posted February 17, 2025 17 hours ago, Zorcan said: Does it matter that in the most recent update to DyndoLOD, Alpha-187, that TexGen is still reporting as Alpha-186. The DynDOLOD executeable is Alpha-187 but the accompanying TexGen executeable is Alpha-186. Will that pose any issues? No it won't cause any issue. The last update was for DynDOLOD only and I just forgot to update the TexGen version number. It will be the same version number again next alpha. 1
rustywd Posted February 17, 2025 Posted February 17, 2025 Here are the log files https://ufile.io/f/d7qjm Went through the TexGen outputs and did not see any instance of them being incorrect there, normal maps have _n and the diffuse maps do not, so that part seems fine. I have the outputs overriding everything else in MO2.
rustywd Posted February 17, 2025 Posted February 17, 2025 (edited) Just another observation, I tested running it twice without making any changes and get different results as shown in the First Pass and Second Pass images here, it looks like it not mapping them correctly as well as not having the right image type. https://imgur.com/a/OZ4EtdE Edited February 17, 2025 by rustywd
sheson Posted February 17, 2025 Author Posted February 17, 2025 22 minutes ago, rustywd said: Here are the log files https://ufile.io/f/d7qjm Went through the TexGen outputs and did not see any instance of them being incorrect there, normal maps have _n and the diffuse maps do not, so that part seems fine. I have the outputs overriding everything else in MO2. Check if this test version of DynDOLOD fixes it https://mega.nz/file/9RBwhB7A#Phv0FkY7MqncuumOQ_LHcitoM-rp-we9rU_lcOSIpLg Report results. If issue persists upload new DynDOLOD logs.
Recommended Posts
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 accountSign in
Already have an account? Sign in here.
Sign In Now