Yakuza

From the wiki: No?? This is not my only source, but the wikipedia is one of the most trustworthy resources. https://en.wikipedia.org/wiki/Mipmap In 3D computer graphics, mipmaps (also MIP maps)[1][2][3] are pre-calculated, optimized sequences of textures, each of which is a progressively lower resolution representation of the same image.[/b] The height and width of each image, or level, in the mipmap is a power of two smaller than the previous level. The wiki just states it's the texture quality. Misworded maybe? No. Reread what the wiki says; they have no idea what mipmapping is. https://en.wikipedia.org/wiki/Mipmap#How_it_works Tl;dr It has nothing to do with texture quality. It is a Level Of Detail modification and strictly an LOD. The STEP wiki fails to even reference mipmap size... If the wiki was correct, why would these two popular mods exist? https://www.nexusmods.com/skyrim/mods/12801/? https://www.nexusmods.com/skyrim/mods/5755/? The second is DDSOpt. A mod recommended by STEP. How much more of the STEP wiki is incorrect? EDIT: The "iTexMipMapMinimum" images were removed, lol. Pretty sure the text was modified as well (before this was posted) but I can't prove that. The images were 100% removed, however. Side note: You will never see mipmappings' effects indoors, at least not very much, and MipMapSkip will never change the base texture yet they included this in the wiki. There is an unbelievable amount of wrong in the wiki.

This forum isn't open. I thought it was, but info actually regulated by admins very heavily. There's bias towards what admins say and they're reluctant to take feedback seriously, even mildly. Example: I criticize the wiki on any level, not even at the level where I have proof of you, STEP admins and the like, being totally wrong, and I - a contributor for free with no affiliation - get shut down. That isn't how wikis are meant to work. I didn't make money on this. I am trying to help when it's obvious you know very little about 3D rendering and simply tweak settings to see what's different when you do. I took a lot of time to research things before testing, let alone posting here. Overall, it feels like you're pushing STEP and STEP alone, rather than stable Skyrim without a ton of tweaking or performance modification. It can be done in Vanilla, and you totally missed this reegardless of how obvious I made it. Granted, I never showed you any of my information, but you treate me as if I had and I was very wrong about it. I'm pretty sure I'm done at STEP as a contributor. If everything I said thus far was dismisse, that's okay. But you don't seem to take criticism very well. Look up Shadow bias, how it's done in games and 3D rendering. Best of luck to you if you ignore the fundamentals. I'll make one final topic showcasing how little you know, and it has nothing to do with shadows.

Noted! I saw this and wondered what was up. I do like the effect it has on some water though.. Here is a very good comparison of various AA types. Wish I found this sooner. https://arstechnica.com/civis/viewtopic.php?f=6&t=1271159 SMAA is supposedly like FXAA, but has no blurring. When it was first released, people crapped themselves at how good it is. Sadly my card doesn't support it, but yeah. I would recommend SMAA over all in every environment I think (unless you can stomach performance losses to extremes).

Increasing it beyond a certain point creates indoor crashes at specific resolutions. The lowering/changing of it at all was to prevent people from creating crashes or get unnecessary performance hits when using one of the provided settings. Besides, I'm not really sure how much of the wiki is worth holding close for shadow info at this point.

Sorry to any users that felt this was helpful at all. It seems many STEP-related individuals act in a typical way for modmakers, and I'm not going to help with that. https://forum.step-project.com/topic/9291-yakuza-shadows-optimized-shadow-presets-for-any-pc/?p=146282 Read the above post from this topic if you're truly interested as for why. Or read the whole topic. Amazing.

NVidia-controlled 2x2 SSAA + 2/4X MSAA controlled + FXAA looks best for performance, but you're running FXAA over a postprocessing AA as well as 4x supersampling (meaning you are rendering 4* your game's resolution per frame). NVidia Inspector (get it!) is what you need to manipulate AA settings to the max for Skyrim with NVidia cards, but most of the settings blow compared to in-game FXAA, surprisingly. Stuff to know before messing with AA much: -- EQAA is the AMD "equivalent" to CSAA. They do slightly different things; I would say EQAA is better for picture quality overall/is closer to MSAA, but CSAA gives far better edges. -- CSAA/EQAA/MSAA/FXAA/SMAA/TrSSAA/TXAA are all post-processes (I think, not sure on last two). They alter the image after a certain point. This means that multisampling/filter/post-processing options in drivers should apply to them. -- FXAA should ideally not be ran with any other filter/post-process, possibly not even supersampling/downsampling processes. -- It's either/or for Transparency Multisampling and Supersampling. Don't run both. -- SSAA processes each frame's resolution by a multiplier given by its AA value and then shrinks it back down to your resolution. This is downsampling, and gives the best quality image overall, usually. But it's a first-off process and it's a ton of work so it's quite demanding. #x# SSAA means WxH or "Horizontal resolution by Vertical"; 2x1 SSAA is rendering twice your resolution width (1920 if 1920x1080 monitor, for example), while 1x2 processes double your vertical resolution (2560x1440 = 2560x2880 frame rescaled to 1440p). Which you choose is dependant on the game and what you see more of, aliasing along horizontal lines (bridges, rails, roadways in racing games, etc) or vertical ones (guard rails, fences, trees, etc). Order of AA options from NVidia in order of roughly best to worst from my testing for Skyrim, taking into account my personal performance hits vs. opinion on the image quality (though both will vary depending on your system specs and monitor/resolution/maybe mods to a degree). WORST ====== NVidia FXAA toggle in driver (This used to be good; they removed functionality to it to push launch ready drivers) No AA 1x MSAA Skyrim 2x MSAA Skyrim 2x Quincunx MSAA 1x2 SSAA 4x MSAA Skyrim 4x MSAA 8x MSAA Skyrim 8x MSAA 4xS SS+MSAA (1x2 SS + 2x MS) 8x CSAA (I rate this as 2nd best for performance, next to 16xQ or Skyrim FXAA if you can't handle 16xQ) 8xQ MSAA 2x2 SSAA 8xS SS+MSAA (1x2 SS + 4x MS) 3x3 SSAA 8xSQ SS+MSAA (2x2 SS + 2x MS) 16xS SS+MSAA (2x2 SS + 4x MS) (Too performance costly for a slightly better image than 2x2) 16x CSAA (Slightly better image, worse performance than its brother) 12xS SS+MSAA (2x2 SS + 4x Gaussian MSAA) 4x4 SSAA (that is 16x supersampling, 4 times your resolution times 4 downsampled - garbage for the performance) 16xQ CSAA (This one is very good for the performance) 32xS SS+MSAA (2x2 SS + 8x MS) (Didn't run it; too heavy for my pc. I assume it's great) 32x CSAA - Hands down the best I used. Sadly it does not entirely kill shimmering. It is so damn good though. ====== BEST Once I hit 2x2, the difference in quality wasn't as easy to see. I paid attention more closely to edges and kept Transparency Anti-Aliasing options off for comparisons (on when testing), and that helped me judge which is best for my eyes. Now, what I found in the end was that Skyrim's default FXAA - all by itself, as its meant to be run - blows them all out of the water. It has a little blur to it but the performance is retarded compared to 16x/8x anything that can even begin to touch on its overall quality I think. Every single AA option and combination I tried did not match up well to it alone. Except 1, which isn't supposed to be done, didn't run great and is supposed to be pretty screwy with frame stability and the likes... 2x or 4x MSAA Skyrim + 2x1 or 2x2 NVidia "Enhance application setting" + Skyrim FXAA, like I said at the top. This produced the best quality picture of all, at least if it was 2x2 SS + 4x MS. It had a little bit of FXAA blur left, but just enough to remove pretty much every shimmering pixel in Skyrim.. Performance wasn't great and I'm pretty sure it's very unstable but that is by far the best setup. I just can't use it myself since I already cut corners to run 16x CSAA. The weird part is that NVidia's driver 16xS could not match it despite me leaving Skyrim FXAA on.

Yeah I'm 99.99995% sure I know how to control shadows now. I have no problems running high distance shadows (8-12k+)

So like, what else is needed for vanilla shadows..? I'm pretty sure I just solved the case, Jones, and I need to know if you guys think you understand shadows yet. Didn't know about this topic beforehand, I just looked into it myself.

I'm compiling a shadow configuration list that will do just that! Hopefully people will check it out/find it useful. May be a good idea to see something similar in the STEP config wiki, as you said. I only found out shadows through tweaking for days, and I'd hate to be the guy that settles for less than what they can get out of the game because they can't be bothered. Adding extra info for various systems like that is the way to go imo.

Hi. So I have a different setting used than what is recommended, based on what the STEP wiki provides for information on the two Sun values. And I wanted a couple questions answered since it's confusing. Am I wrong in thinking this causes the game to do more work in a burst, and/or less work over time? Is the game still checking for sun/time changes every 3 seconds or whatever for 1 minute to pass in-game? Are any of the 3 below statements true? - More GPU work in a burst because it has to move all shadows at once for a longer period of time. - Less GPU work over time because it's not frequently changing the shadow position. - Less CPU work over time because it's not constantly checking the sun time (i.e. fSunUpdateThreshold = gametick checking?) Maybe I'm misunderstanding what these settings do/expecting more out of them. I actually enjoy 3/3 settings because shadows are more static (which I prefer but it's probably not as good as constantly moving shadows), but they also sort of shimmer like they would in real life with birch trees and such when the wind blows. And best of all, they don't drastically change in position. I was mostly hoping it helped with game performance at high shadow distances, though.

It gets even stupider when you realize that eventually the difference between 1 frame and the next is potentially only a 30% change in your monitor's displayed pixels or something, hardly noticeable in most scenes. After a certain refresh rate/fps, you are completely limited to what the game is offering you for very different images. Obviously if you're spinning in circles in Skyrim with ultra-high mouse sensitivity, this totally doesn't matter because you will notice that sweet, buttery-smoothness of high FPS and high refresh.

You guys do realize FPS and monitor refresh rate are directly connected. Motion blur is a technology that masks tearing and such, which may be why you percieve ENB as being "Good", or the experience with it better than it actually is. But,.. I don't know if most people use ENB and get some form of blurring. The difference between the two, assuming ENB is blurring, is that motion blur can compliment higher fps if you're into it, yet it is 100% unnecessary because the frames are fluidly moving in front of your eyes and no tearing should be happening. If you have a high refresh rate and stable, high fps, you will know this is true. If you have adaptive refresh on a monitor, you will know this 24-30~fps crap is the dumbest thing ever reverberated within gaming culture. Exactly 30fps on an older CRT monitor will look better than 60hz LCD due to CRT refresh and how it displays pixels/images. Exactly 60fps on 60hz will probably look better than on 120hz depending on driver settings unless you have adaptive framerate control like Freesync/G-sync. 61-120fps will only be displayed at 61-120hz+ because the monitor is only refreshing that many times per second; it cannot physically display more than its refresh rate. Unstable 120fps on 120hz Adaptive (Freesync/G-sync) looks better than unstable 120fps on 120hz anything else, and so on. The cut-off is 144hz/144fps because 240hz+ monitors use technology to flash twice or more, displaying the same exact image twice per GPU frame time, which only cures part of tearing and actually does nothing in terms of displaying more new information to you. It's not showing you more than 120 unique images per second (FPS) even if you have 240 fps locked. And, seeing as everything is in a theoretical locked, stable 30/60/whatever FPS unless it's able to utilize adaptive refresh, motion blur and low fps seems really good. It's not.

Should have an option on that R9 380 in Catalyst to limit FPS to 55-60. Do that. Not above 60, sometimes preferably below 60 if you have trouble maintaining 60 to begin with or use questionable or new drivers. This is supposed to be for useless power consumption, at least marketed by AMD, but limiting FPS is more than just saving money. It's very logical to do in games to stabilize things. https://www.amd.com/en-us/innovations/software-technologies/technologies-gaming/frtc Disable V-sync unless it provides functionality to ENB (dunno why it ever would but I don't know ENB), or you'll get extra input lag from old frames being thrown in late when they miss a v-sync cycle. The only time this matters less is when you have a high refresh monitor and stable fps near/above your refresh rate without FPS limiters. That's just a recommendation but it may provide a lot more stability. Despite what the STEP wiki says, I am pretty sure these values are very wrong. The STEP wiki states that each shadow setting is used for a specific purpose individually, where in my testing and the way these settings normally work in 3D rendering they work in tandem and cannot/should not be heavily warped. Ideally they shouldn't be weird values, and these are at weird values. I'll be making a separate post on this soon. If these are config settings that ENB depends on for functionality, you can ignore it for now. I highly recommend you tweak them for testing's sake though. Try leaving BiasScale at 0.15 and continue with other settings. Try reducing your iBlurDeferredShadows setting in [Display] to 1 instead of 3 as you have it currently. But you should be able to tweak Blur without it affecting any of your other shadow settings in a negative way. I find its intensity is dependant on resolution though so maybe it has a significant impact on performance when you begin rendering further shadows. I believe this is why you are lagging/having stuttering issues. Reduce this to realistic/recommended levels, in the .8 to 1.2 range. The Ultra setting for this is 1.5 and you do not have an Ultra level CPU for Skyrim (read my other post). At the very least, lower it to 1.2-1.5 and increase incrementally if you notice better performance. Basically, you are rendering what is labeled as "distant objects/terrain" by your config with that setting. But this increases CPU load I think, as it's producing full meshes and textures for objects further and further away from you as you increase your Split multiplier. In short, you'd have to have a huge, high resolution screen with the eyes of a hawk for this to be necessary, and even then you probably wouldn't have a PC capable of utilizing it properly in Skyrim. And reducing it in my game is actually a boon in performance, just like my testing with shadows. But in my case I am CPU-bound (similar to you but worse); my graphics card has far more room to fill than my CPU and I hit my CPU's limits much earlier than my GPU can in Skyrim. It is not an entirely GPU-bound setting as far as I've been able to tell. Read my first post a bit up from here for more info on how Skyrim handles CPU power. Besides the above, with the rest of your config settings this particular line should not need to be tweaked to an extreme value of 4.000. That is an insane number for any PC in existence. I actually only made an account to report these 2 findings, fSplitDistanceMulti= tweaking and the relation between the 3 primary shadow config settings (Bias, Distance and Resolution). Kind of a funny coincidence, but I still haven't had others test it to prove that I'm right. If I'm wrong and this doesn't help, I apologize in advance. But I am pretty sure it will. Like 99.997% disinfectant-sure.

Skyrim is heavily CPU bound/is single-threaded. AMD processors aren't great per-core and Skyrim does not utilize multiple threads very well. Instead, you want very powerful cores along with, ideally on any system, a good (6GB+, faster the better) amount of system memory. The higher tiers of Intel have higher clock speeds but I am positive you can still play Skyrim with a Pentium G3220 dual-core and reap benefits over an AMD quad or six-core that has roughly an extra 500mhz and 4/6 total cores. This is because Intel's latest architecture, apples-to-apples as you put it, versus AMD's Bulldozer+ architectures is roughly 35-55% faster. On a 4/6-core AMD FX-4300/6300, you have 2/3 physical cores to hose information through and 2/3 logical ones. But, since each are 35-55% slower than Intel's, you've basically got a $60 Intel-equivalent AMD processor. Now that's not to say that AMD is bad. This means that for CPU-bound games and processes, Intel's bottom line is closer to AMD's bottom/middle, and the top-end of AMD is totally thrashed in terms of value (6/8 cores with heavy clock speeds that don't matter). But, it also means that Intel's i3 processors are only barely faster, if at all, than an AMD processor that's 15-40% cheaper. They are simply different tools for different jobs in the computing world that are akin to Swiss Army knives in terms of how broadly they can actually be used.