z929669

I tested with basically with minimal STEP, so VRAM was not an issue. Go look at the screens just after the benchmark results of the DDSopt guide (6-14 FPS using those examples, and they are direct compares): Screenshots (CrossfireX): Vanilla baseVanilla + HRDLCDDSopt'edVanilla + HRDLC (optimized as above)Vano89-Hybrid(1024+2048)Vano89-Hybrid+ Vanilla normal mapsVano89-Full 1024Screenshots (Single GPU): Vanilla baseVanilla + HRDLCDDSopt'edVanilla + HRDLC (optimized as above)Vano89-Hybrid(1024+2048)Vano89-Hybrid+ Vanilla normal mapsVano89-Full 1024

You can get to the guides from the main page

Lots of great info in the guide. It will be popular, and you will find yourself answering many questions ;) I found it useful as well for my own purposes, as I was not sure what the top-of-the-line mobo/processor combos were for sure. Note on Crossfire using 1 Gb cards: I use the ATI 6850 (Sapphire, 1Gb) in Xfire, and see a significant performance improvement in FPS (10-20 FPS consistently), so while my VRAM limit is still capped identically, I think that the GPU performance gain is of significant benefit. The ATI Xfire drivers are well suited to Skyrim, and there is even an Xfire app config for TESV. Bang for the buck is definitely not worth buying two 1 Gb cards, but if you already have them, it would be highly advisable to put both to good use ;)

@work I use both Chrome and IE. Former for personal, latter for work-related stuff. Love Chrome because of Chrome sync feature (saweet!)

I am personally not a fan of WIS. Perhaps a fully randomized, vastly toned down version would be nice. Just seems strange and unrealistic to come upon a cohort of spriggans battling a cohort of elemental wizards in the middle of the wilderness :P

TC's or no SRO are much better IMO; however, I cannot tell the diff between the latter largely because of the slight change in camera angle between screens :P

Discussion thread: Skyrim HD - 2K Textures by NebuLa Wiki Link Re-accepted for v2.2.9.1

Thanks for the heads up. I will peruse it.

It occurs to me that this could all be a mistake... I don't see anything about Riverwood being included with this pack, but what if the author is also including vanilla textures in the pack? I have oft seen mod authors pack all kinds of extra baggage into their mods, and that second screen looks like vanilla to me ;) Take a look at the Nexus screens . They look much better than what we are seeing here ...

It is a sweet little program. However, I am still out of the game and hoping for some stability soon before I get back in for testing... need stability for benchmarking :@

Yeah, I have forgotten to check subs in the past as well, but this is only a minor issue if you are aware of the differences between complex and simple. Great for organization, but adds a layer of complexity. Glad that it worked ... That mod is a great example of a complex BAIN too. I have retained all of the alternate textures in the complex BAIN (see attached). This allows me to swap out very quickly (uncheck one sub, check another, anneal), and look at differences in game. I LOVE options.

First, click the 00* Sub-Package in the middle right pane of BAIN. Did this fix the issue? If so, you need to always check the subs of interest when the root folders in your BAIN pack are named anything other than that of a standard Skyrim directory. See Figure 6 in the WB Guide. If this is not the issue: Please, post an explorer image like I have attached with folder structure AND files (this is a copy of mine, which works just fine)

Sure. It's skipping the: armor\glass\f\curiass.dds armor\glass\m\boots.dds armor\glass\m\curiass.dds armor\glass\m\gauntlet.dds armor\glass\m\helmet.dds I assume it just doesn't recognize the structure..? The structure is like that for ALL the armor variations... The BAIN is fine. WB will recognize ANY standard file type under ANY standard directory including: Data/ Data/textures/ 00 Core/textures/ It doesn't matter what the structure under textures is, only the number of directories above textures matters (there cannot be more than one above before root). When you select that BAIN, what do the package details reveal? Are the files of interest listed there (upper right corner under [General] tab)?

+1 on the gen consensus being reported here. I prefer Skyrim HD ...

Discussion thread: Skyrim Realistic Overhaul by Starac Wiki Link Since Starac pulled his mod from the Nexus and all share sites, he has requested that we not host his files from STEP. We are reluctantly honoring his wishes, and this means that our users must do the same. If abyone wants to host Staracs work privately or from some other site, that is fine, but don't let those links connect to this site, or you risk a severe penalty. STEP fully supports the concept of "open modding", but others do not ... the Nexus does not, and so STEP reluctantly adopts a position of "no position" with regard to the Nexus and their community's view. While we fully support the "cathedral" view within the STEP community, we prefer to stay in compliance with the greater modding community where our paths intersect. This means that with respect to mod authors on the Nexus, we adhere to their wishes with regard to their work. Torrents are different, and anyone on STEP can point to the SRO torrents, since Starac himself has pointed users to these. STEP and its users simply cannot host the files ... that is our agreement with Starac. We know that this seems contradictory, but we do this more for the sake of honoring our agreement and a mod author's wishes as opposed to doing what would seem most logical or convenient. WARNING: This mod link points to a torrent site, which is one of the riskiest places on the internet in terms of computing security. If you choose to download the torrent, please use the link at top of page titled "Download Torrent", and do so at your own risk. That said, thousands of people have safely downloaded this mod using this torrent. STEP 2.1 + old SRO ONLY-landscape (1024): https://imageshack.us/photo/my-images/825/tesv2012033100250537.jpg/ STEP 2.1 + FULL-SRO v1.1 (1024): https://imageshack.us/photo/my-images/521/tesv2012033114523540.jpg/ You draw your conclusions about the SRO 1.1 update and what to take from it Tell me what you think ! TC Dropped in 2.2.9.1. See 2.2.9.1 changelog.

@GraPhix Can you post a screen of the Explorer view of the 'BAIN' structure (as I do in the guide)? It will be easier to spot the error then.

Yes, I do; however, you are probably going to benefit most using the optimized DLC pack by Vano. These are already DDSopt'ed. I still have testing to do with DDSopt on vanilla and Vano texture combinations

You are correct with respect to benchmarking procedure... whatever is more tedious and boring is probably the best practice RE DDSopt'ing my textures, I always operate on my BAIN installers. Unextract > DDSopt > repack ... I am doing this with EVERY mod with textures in step, so I will have 2x BAINs of many packs in my installer list. I will test each independently, and then cumulatively to determine impact. Then I will likely run with the optimized textures and keep the pre-optimized versions around just in case I notice any errors in game as I play normally. Then I will troubleshoot. It will take awhile, but I plan on beginning as soon as I finish transferring the WB and the updated DDSopt guides to the wiki. I will then create a table on the wiki to house the DDSopt results. Wiki will also be a good place to keep all of the best tips/tricks/tweaks once we work them out in these forums.

Correct, will fix as I move the guide over into the wiki. about 25% complete right now.

Ahh you found TESCosi. Good site and good author(s). Not that well maintained anymore, but it looks like he(they) hasn't left it completely. That installation order applies to Skyrim in a general sense, but there are idiosyncratic details for Oblivion that do not apply. STEP order is best, as it is specific to ... well ... STEP ;)

agree with Monty's assessment. You are probably looking at CPU limitations more than anything. You may need to tweak your INI to get the best results on that front, and mind heavy script-based mods, overall resolution, shadows, and various other known FPS hitters from the CPU perspective. Also, it is worth benchmarking under a fairly anal protocol. Wait at least 30-60 seconds after load to begin benchmarking, and look at the average FPS over a 10-20 s period in the exact same places and under exact same conditions. PITA, but the data are at least relatively certain ...

good points... all I can say is that I work in the spirit of Unix/Linux, so prefer forward slashes. Windows understands either, but Python interprets in a much more discriminating fashion, so I will change all '/' to '\'

I understand exactly what you are saying ;) I went through the rather exhausting process of explaining complex package restructure because ... well ... I am pretty anal :P If anyone else out there is like me, they can follow along, and those that are not so persnickety can simply say 'screw it!' and toss the garbage and go simple My personal general rule is that if two variants modify the same thing and were either packed by or created by the same mod author(s), I complex BAIN them for record-keeping purposes. No need to worry about looking at conflicts, since I KNOW the nature of the conflict by the very fact that they are adjacent sub-packages within the same BAIN.

sounds like you are missing meshes. Did you check your installers?

Short answer is 'Yes' Using the ideal DDSopt settings, there should be benefit to processing 4x textures and higher as well, as compression will be optimized, redundancies will be deleted, and mip levels will be optimized. The key is not to reduce by constraining resolution (always leave at 8192x8192)... Ideal DDSopt settings, confirmed by Ethatron