This is NiranV's Typepad Profile.
Join Typepad and start following NiranV's activity
Join Now!
Already a member? Sign In
Recent Activity
@Candy68 What has this build being built with mesh to do with the fact that there are lots of complaining about bad mesh practises and content optimization? Mesh != bad.
This. THIS is exactly what i'm talking about. I went there to see it myself and i was getting 40 FPS minimum everywhere i went, deferred, shadows, ssao, volumetric lighting, ssr, everything on and with an almost 9 year old AMD FX now. That is pretty respectable. The place also looks very nice and even with quite a bunch of animesh around i've never seen my framerate hitch at all. I've made some quick pictures there too and checked wireframe as well. The place doesn't look like any of the latest triple A titles but it really doesn't have to. It looks quite nice and unlike most of SL doesn't look like 2007 garbage cobbled together. The lighing is decent, lots of material usage, the texture usage was decently low and it was perfectly used to mask all the super low poly models around. Wireframe: I'm quite happy with what i saw and i wish this would be more common. SL could be so much faster and look nicer too.
I can only support what Penny says. I'm still using an 8 year old CPU, an old AMD FX 6200. I've been building my Viewer practically since i got that CPU and i've been nothing but taxing my CPU to the max every day and its still running. I can run SL with 60+ FPS if the place is good and the mentioned 30 FPS with nice graphics are easily optainable if the place shows optimization. The only thing newer on my PC is my GPU, now a 1060 was previously a 670 (also ~8 years old almost) sadly got fried when we had a power takeout and my PSU broke, the PSU took my GPU with it. My videos more than show that what Penny says is absolutely realistic. The only thing really important for SL is the CPU, all you need is a CPU with around 3ghz and 4 cores (intel would be better). The GPU can be basically anything that can run Deferred, even a GTX 460 (i used it while i waited for my 1060) can easily run SL with all the fancies on and still get 30 FPS easily. Prior to my AMD FX i had an AMD Phenom X4 for years with exactly the same performance as i get with my FX today. SL really isn't that demanding if people just stopped wasting our resources. See the videos here:
@amandamagick I'm sorry but it won't ever be available for Mac unless someone does it for me. @Igor Romanov I highly doubt that Alchemy looks like that. Infact i know Drake and from what i heard none of the graphical changes he did so far have been in an official release meaning Alchemy still looks like the plain old LL Viewer. But then again most of the pictures aren't particulary telling to what makes Black Dragon great. Which is the reason i usually stay with my own pictures because i know i don't edit them (at all) and i try to use as many unique features as possible (and often not only just subtle but actually visible) to truly show what Black Dragon can do. I agreed with Wagner however that he should show other people's images (specifically human ones) so people don't get the impression that the Viewer is for furries only or only makes furries look good. While it can make everything look it certainly does make furries look much better in comparison to humans which see much less improvement as they usually use less of the unique features of Black Dragon, they would require a whole lighting change to look fundamentally different. I'm not saying that these human pictures are not good but sadly i often don't see much difference in them myself compared to other pictures taken with Firestorm because again they often don't make much use of all the fancy new graphical features which leaves them with only the subtle improvements in SSAO, Shadows, Depth of Field which don't stick out as obvious as say Screen Space Reflections or Volumetric Lighting, the first of which is more often used on furry avatars (dragons, lizards, kobolds or anything with scales profit a lot from this). I've always planned to make a comparison compendium at some point showing ALL the actual differences between other Viewers and Black Dragon but since things are constantly changing and stuff is getting tweaked all the time, i've held off on doing so, not to mention that doing actual real 1:1 comparison shots is extremely hard, it would require me to add in an option that sets the Viewer back into the original LL Viewer state to have an easy on/off comparison between them. It's something for... someday, until then people simply have to trust me when i say that almost everything has been touched and improved wherever i could improve something. @JK What would that work be? 7 years of work into the Viewer, most of which went into the UI, not the graphics, the graphics are one of the smallest parts of my work. @Blaise Glendevon I'm not resistant to suggestions i just basically don't get any and those few i get are rarely ever good. Typical suggestions are: "The UI is shit." and "Everything is so different." and of course the worst "Make it look like Firestorm." None of these are suggestions and none of these tell me anything at all. None except Firestorm, that one does tell me something but not much. I know how Firestorm looks like and i absolutely despise it because its a broken, half-assed, uncreative, patched-together UI that looks worse than the Official UI. Firestorm offers no consistency, no UI rules, no straight design (no design at all), the skins are all lifeless recolors and variations from each other. Yes they have lots of UI and stuff needs to get done in a timely manner but the quality takes a huge hit. It took me something like 5 years or so to just come up with a skin that offers a clean look and can be used everywhere, it took me 7 years counting to find a UI layout that i'm somewhat happy with, one that offers consistency, cleanliness and most important works technically everywhere. UI design isnt as easy as "just make it look like Firestorm", that's not a solution, Firestorm's UI is bad and i'm avoiding copying any other Viewer's UI, i want to find my own unique style, one that works because it works not because people were already used to it for almost a decade. Then again i don't expect to ever get some serious suggestions because people don't know what i want and what rules i design my UI in mind with. The vast majority of SL doesn't know UI's at all, they are in SL only and never really move outside of their tiny little comfort bubble. If they did they would have much less issue finding their way around my UI, or any UI for that matter. Every app is different, has a different UI, why does it work everywhere else just not in SL. It's the users, not the Viewer. I can switch to any Viewer any time without having issues finding my way around because to me they all are the same, they don't just look similar, they are. To me they don't differ much from my own Viewer but i don't start up a Viewer and are confused because the button has a different texture now. The people in SL simply need to learn to work their way through the UI in a plain and straight way one that doesn't immediately causes confusion when a button has moved from left to right and triggers the immediate "OH NO EVERYTHING IS DIFFERENT" thinking that will hinder you and possibly make you fail to understand completely. I can't help people but tell them that they have to be less ... "strict" with how they approach the UI, if they bail out the moment a button has moved from the first to the second row rather than simply zooming out a bit and looking at it from a bit further away... at the bigger picture... or more generally then i can't help them. By biggest issue with the UI design is that i don't want to break or dumb down the UI, i want to keep the UI intact, on a high quality level, accurate to the point and most importantly technically correct. I don't like using dumbed down names for features like "enable-this-to-do-exactly-this-and-that-feature". An example of this is the well known "Advanced Lighting Model" (previously known as Lighting & Shadows). Advanced Lighting Model is basically saying nothing, advanced lighting model, what does that even mean? what about the lighting is advanced, what does it do? Well it makes the lighting model advanced i suppose. It's correct technical name is Deferred Rendering. Deferred Rendering doesn't tell the normal user more than Advanced Lighting Model did but it offers the unique ability to google it because Deferred Rendering is something used in other 3D applications. Those who truly seek knowledge can actually find it, if they are willing to look for it that is. Similarly i spent a lot of time improving the little helper features, one such feature are tooltips, sadly something completely overlooked, many people either disable them or don't look at them but i spent days writing extensive tooltips for all features that clearly explain what these features do and even make notes about things you should keep in mind such as possible performance impact or problems they might cause or extra features that they might require. Once again if people just take the time to look at things, they would have much less issues figuring things out. Whenever i change a window i try to reduce space wasted and fill it with purpose, i try to make the UI as obvious and clear as possible given certain restrictions, sometimes i have to cut somewhere, icons rather than text due to space limitations for instance. The UI would look a lot more chaotic if i didn't spend as much time on the UI as i do, see Nirans Viewer's UI, it was all over the place, constantly changed to be even more all over the place and generally there was little thought put into the UI, all that mattered was cool looks and trial-and-error experiments. It's quite sad to see that so many people prefer the old chaotic UI of Nirans Viewer over the much cleaner and organized UI of Black Dragon. @Op You are a perfect example of the above. If you spent less time "having a hard time navigating around" and instead tried to find a solution you would have found out (either yourself or asked someone) that you can easily change all the controls to your liking with just a few clicks in Preferences - Keybindings. @Priscilla Sadly that is just how it is, Op basically just proved that people simply don't look (and not even hard, its right in the preferences window, its a whole extra tab just for it). They want everything to be perfectly set up for their personal use from the get go. It's sad to see that even though the solution to the problem at hand is basically right in their face most people fail at finding it and then fail again because they decline my help offers that i have basically scattered everywhere in the Viewer. From my name appearing in the friendlist, not to just look at it but to IM me if they need help (even tho most people who IM me think that they don't deserve my time. If i don't want to talk to people i wouldn't put my name into the friendlist at all times now wouldn't i?) to tooltips, to the viewer group, the discord chat, emails or simply finding me inworld. If people would just ask... @The Giving Tree Black Dragon is much more than just a photo tool. It's an every day Viewer like any other. I've been all over content creation of all kinds for years. The Viewer works no different than any other Viewer in that regards, the build tools are the same, the inventory is the same, the script window is the same, none of them differs to what LL offers in any meaningful way. It just doesn't offer special features other Viewers have, like copy-paste entire object parameters because... well those are their specialties just like other Viewers don't offer Black Dragon's special features. If you can't "work/create" with Black Dragon you can't work with any other Viewer other than Firestorm because of your own dependency on these extra QoL features.
The reasons for this running at only 30 FPS (worse even according to the choppiness) could be many, starting from the very beginning with the actual recording itself, unless this is recorded with NVidia Share (Shadowplay) i doubt its the original untouched framerate. Further we don't know the settings here, touching shadow resolution can quickly decimate your framerate for apparently no reason. Then there's the fact reshade was used here for the color correction and lens flare (color correction could have been done with the Viewer itself), reshade alone lowers the framerate just by being there. We don't know the system specs either and we don't know whether the Viewer might have been still downloading/decoing something. Sometime later today i'll go visit that place and have a look myself, with 8 year old hardware i'd expect 30 FPS there as well but we'll see, i've gotten surprisingly much FPS in several places i expected to be worse before.
@Candy68 We don't have a choice. SL is SL, it's completely unique. @David We don't get ruffled because A) no one ever sees them and B) no one dares to tell them in their face in SL... or do you walk up to a Linden and tell him his stuff is shit? I would. Not exactly like that but i'd tell him why its shit, if it is shit but then again i don't fear the Lindens, they are humans just like anyone else and if you talk to them they will either ignore you (because they are busy) or they will talk with you just like any other person.
@Galathir We tried the carrots for years. They didn't work and they will never work unless people are paid for it (which is the ultimate golden carrot). LL is not going to do that, the users are paying them for their content, since users don't care and buy everything they think look good, they are just "confirming" these bad content creators in what they do. This whole "don't buy what you don't like" doesn't work like that. A few hundred people not buying something doesn't change anything. That's exactly why this kind of topic is coming up over and over. People need to be reminded, people need to be educated and people need to be slapped in their face sometimes. We are getting nowhere derendering these people either, it's a workaround for an issue that needs fixing, something we have been doing for far too long. (LL is VERY guilty of this for many years) Also, we do have a way to limit memory limits for scripts, i use it for my own scripted AO (which is just one script and ~40kb). You should check the LSL wiki. I've been using this for years now so its nothing new either. I'd rather have less but better content than more but shittier content. I can only speak from a furry point of view but with meshes, furry avatars have not had any quality drop, infact the quality increased many times while at the same time reducing optimization across the board, this didn't hurt sales nor amount of content available. Infact, this whole "less content" thing is nonsense. Optimization doesn't mean less content, actually, it would mean MORE content, creators would be spending less time on their meshes as meshes would be simpler, less triangles means simpler rigging, less potential bogus vertices, faster upload, less upload cost and its only a one time thing. Seeing that most human mesh content has been created once and never actually updated (the CATWA head in particular hasn't been changed since it became Bento) further invalidating that "less content" statement. People simply don't understand that optimization doesn't mean "less" which immeddiately seems to ring all warning bells. Optimization simply means making more of the available resources and SL really doesn't have much resources available at all but way more than enough to make good looking content without completely murdering everyones framerate around for no reason. I've been complaining about this for years now because i've been using, making and visiting very optimized content many times already and there is simply no comparison to the usual junk we see. They look the same if not better and you can smack ten times more of them everywhere before you notice anything bad happening. This isn't wishful thinking this is simply a fact and a goal we should be working towards but with people showing simply no interest in doing so nothing will change, and the same people will simply continue complaining that LL should fix all their shit and make SL faster... or replace the engine or make SL2 or whatever. @Stuart I think you are just talking a whole lot of trash. First off i know from several people with high end machines that you aren't going to get 100FPS especially not with people around unless you are running a good Intel CPU, a decent GPU and have most if not almost everyone derendered and your settings lowered. And i can tell you i can't even get more than 30 FPS with people around and thats with 95% of them being jellydolled. There is absolutely no way you are getting 100 FPS with SL looking like a triple A title. You don't even know what "looking like triple A" entails. It's incredibly hard to get SL to look "triple A" outside of still pictures because the content in SL is simply not made for it. What stinks most however is that you think i "self professed" that "Graphics Expert" title. No i didn't. Infact i never ever even said i'm a graphics expert. That's just what everyone thinks because i'm the most vocal about graphics most of the time. I'm not a "Graphics Expert", just a "Graphics Fanatic". You don't need to be an "expert" to see that the vast majority of SL content is an unoptimized mess. Even worse is your believe that me saying "my stuff is so much more optimized than yours" is an ego thing. It is not. It's simply a matter of fact. When you educate yourself a bit more on how SL works and what's good for SL and what isn't (which really isn't that hard to do since most of these things should actually be quite obvious) and you spend some time applying this knowledge to reduce any impact you may or may not have on everyone around you, then you can say "you are more optimized than them" because that's simply a fact, the probability that whoever hears it didn't spend a single minute on even THINKING about optimizing their avatar is 99%. So no, it's not my ego, i'm just saying how it is, i've spent a couple hours exporting and combining textures, removing invisible body parts, making sure no body part uses two different textures which look identical (due to uploading multiple versions of it), making sure my scripts are reduced to an absolute minimum (like not wearing a whole arsenal of weapons when i'm not on a combat SIM or something), stuff like that. It's really not hard and most of it should be so obvious that you should know how it works. Heck, i even detailed and wrote down the process how i optimized my avatar in a blogpost (sadly i had to make it private because it includes ripping off textures from your avatar to combine/reduce them, which could be considered copybotting). ---- Optimization is no magic or some status or something to boast with, optimization is a process, in SL a very simple process, to reduce the resource usage to an absolute minimum without visual impact in best case and with minimal visual impact in worst case. You should be striving for it not call out people who do it as snobs.
@Galathir @Nara You just don't get it, don't you? It doesn't matter if SL is not a game, or a game experience. It doesn't matter what engine SL uses. This whole things about about the ignorance towards the absence of optimization in SL. Regardless of whether SL is a game or not, whether it uses Unreal or CryEngine or its own, doesn't matter. SL, like any other platform needs to employ some rules, some guidelines for optimization and this has to be burned into both the user and the creator's head. SL could run so much better for everyone, it could be so much less laggy both client and server side if any given content was more optimized. SIMs lagging like crazy because a bunch of avatars are on them? Guess what it's because they are heavily scripted with a whole bunch of unoptimized scripts wasting unnecessary CPU cycles of the server which could be used for more important things. The SIM almost dying every time someone teleports in? Guess what, you are wearing way too many heavily scripted attachments. Your Avatar taking 5 minutes to load? Guess what, it's because your avatar consists of pieces each of which are probably 5-20mb worth of data to download. If that's not bad enough, multiply it by the amount of avatars around. SL could be so much more by simply adding optimization to the equation. Optimization in content would free up resources, it would mean people have to download less, waste less time downloading, sacrifice less FPS, bloat memory less, lag the SIM less and all this while looking visually no different. Freeing up these resources would mean you gain them as available resources to use for MORE. If one tree kills your framerate like a hundred avatars, you're only going to put 1-2 of them down before you'll be unable to put any further down because your Viewer is dying, now replace this tree with one that looks identical but is optimized and takes 100 times less impact on your framerate, guess what you can now put a hundred or two more of them around before the same happens. This allows you to be more open and creative with your content in SL. Also your entire point of "LL offers us tools to limit what we want to render" is pretty much complete bogus. While LL offers us tools, these tools are useless. LL doesn't offer us derender and LL's complexity calculatios are a fucking joke, they mean nothing, you can't limit how much polygons you want to see, how much VRAM you allow per avatar, nothing of these, all you get is a slider for an arbitrary number that means nothing because it is inflated by harmless object features that are basically free to render, giving people the impression that their unoptimized crap is fine in comparison to people who put thought and time into their content. Besides, LL is NOT interested in offering you effective tools for performance tweaks and neither are they planning to punish bad content creation aside from their stupid LI which is basically another useless complexity calculation for when you rez things on the ground, making it unusable for objects attached to avatars.
@Anonymous Nice try but nothing of what you say holds any value. You throw in user error cases, explain them in the most generic and uninformative way and sell them as my fault. First and foremost, if you happen to have 8 hours of pain in your hands from simply gripping the mouse and hitting keys on your keyboard, the very thing they were made for then maybe you should see a doctor, this is hardly a Viewer issue. Secondly and more importantly you make it sound like any of this is my issue, i don't know about you but across all Viewers i can use the mouse and keyboard both to look around and to move around the same way and if for some reason i was not able to do so i could still open the camera controls and move/look with the onscreen controls, again just like in any Viewer. This is 100% Viewer independent, is nothing new and has been like this since basically forever. Thirdly, user friendliness. I think i waste enough time into the UI to make it user friendly, starting from the very design to make it usable, to all kinds of quality of life features and little bits and bobs such as writing extensive tooltips for each and every option in preferences. If what you said was even remotely true you wouldn't have any of this, there would be no tips, no icons, no help, everything would be labeled non-informative, the movement controls wouldn't exist at all (since you can bind them yourself), no camera presets except one basic one would exist (because once again you can create them yourself), i wouldn't even think that keybinding was a thing since that is userfriendly too, you know, allowing each and every user to customize the controls the way they need them (for different keyboards as example). The only reason my updates take longer than i'd like to is because i take the time to consider the users. This very NWN post here shows a feature that i'm currently working on and struggling to get out because of the user friendliness. Creating an interface that allows both super advanced controls over an extremely complex and deep system is one thing, making it easy to understand and simple to use is a whole different world. Just because you are too ignorant and stupid to be open enough for new things and not act like a complete moron when you see something slightly different doesn't make an entire application not user friendly. Further, you come here and spew your random bullshit about your problems but it is obvious that you have made no attempts to find the underlying issue. There are thousands of users using my Viewer who are happy and can see mesh fine, it could be YOU who is the problem, maybe you are running an outdated version, maybe you're running a bad GPU driver, maybe your GPU or CPU is simply trash, maybe another application is hammering your hard drive, maybe you put your cache into the same folder as Firestorm in expectations to improve performance or save space, maybe you've copied Firestorm settings over, maybe you did something other incredibly stupid and are now trying to make it my problem. Second Life is a complex system. Sometimes it works for weeks better than it did ever before and then there's a day it just doesn't work, shit doesn't load fast enough, disconnects, random crashes. A good example is your "most textures stay blurry and there is no way to refresh". This one single sentence alone tells me that you have simply no idea what you are talking about and have completely wrong expectations. "Texture Refresh" does not fix textures staying blurry. Textures stay blurry because your texture memory is full, how about you CHANGE the texture memory settings, ah i forgot, you expected the Viewer to do this automatically for you... oh wait... it does already! To take the math off the user and don't bug him with texture memory stuff i added a feature that is enabled by default which automatically increases/decreases the texture memory settings to allow the Viewer to take whatever it needs while at the same time never taking more than it must, that is given you have enough texture memory to give. So once again you are either on an outdated version that doesn't have this feature or your GPU is simply not capable handling as much textures as required by the scene you are currently looking at. Refresh wouldn't help here as explained in my previous answer to Jackson, refresh just reloads the texture which would just get trashed again when the next texture cleanup cycle comes due to low texture memory. Lastly, Firestorm is not 1. User Friendly, its simply a dumbed down interface made specifically for people who need help finding the way to the bathroom. The way they word their tooltips, the way they label their options, the way crucial settings are named. It's made so even total retards at least have a general idea what they are doing, that's not user friendly, that's lowering everything to the lowest common denominator, which is the intellect of each and every individual person. Firestorm is not lighter on your hardware either, it's a massive memory hog in comparison to any other Viewer, besides how "hard" it is on hardware solely depends on your settings. If you turn on shadows and set 512m draw distance and expect anything usable then once again that is your own issue. Also a note: In my Viewer you can always find me at the very top of your friend list, guess what that could be used for? Right, you can ask me if you need help, i suppose you didn't figure that one out either. Some people.
@Jackson Texture Refresh was never in the Viewer, i can't put it back. It's also rarely if ever needed, see below. Animation Sync can be done via the Animation Manipulation window found in Dragon - My Useful Features - Animation Controls. Select all people you want to sync and reset/restart their animations, this will sync them all up. Hide Lookat Target is a thing i do not support. There are official ways of doing it, you can find the official tweaks in the Second Life wiki if you go here: Fly override is unneeded as you can override fly by enabling the Admin mode. Preferences - Viewer - Show Admin Options, then Dragon - Develop - Show Admin Menu. Texture de-rendering/re-rendering is called texture trashing, it is a widely known issue and is infact not an issue but a feature. The Viewer drops textures out of memory because your video card memory is full otherwise your framerate would drop massively. I have already done everything necessary to allow more texture memory to be usable, you can (if your GPU has so much memory) set texture and scene memory higher, up to 8gb total, that is far more than any other Viewer and prevents the texture trashing issue. It's not a fix but it's a "throw more hardware at it" workaround. If you want a real fix, tell the texture creators to use smaller textures where possible and don't use giant textures on tiny items.
@Lizzy Ignore the note from LL. BD's arc calculation is much better at determining your actual avatar complexity than anything LL has to offer right now. LL's arc calc is a joke. I'm not going into details yet again, you can read the reasoning for changing it on my blog. But to give you a short TL:DR LL's ARC: Optimized Avatar vs Unoptimized 80,000 ARC vs 40,000 ARC 150,000 polygons vs ~3,000,000 polygons BD's ARC: Optimized Avatar vs Unoptimized 250,000 ARC vs 3,000,000+ ARC 150,000 polygons vs 3,000,000 polygons The numbers are arguably big yes but the calc does a much better job at getting rid of bad avatars (unless you set it way too high or disable it of course).
@Bastilla The KTris display is a bad way of figuring out how much polygons something (especially a single object type has) it is highly inaccurate and prone to randomness as the rendered amount of triangles can change at any time for any reason, do not use them please. There are several (and in Firestorm probably too) ways to get an accurate measurement of how much polygons a single selected object has. The Object info for instance should show this in both Black Dragon and Firestorm. Black Dragon also has a dedicated Advanced Complexity Window made to show these and i'm sure Belleza has much more than just 76k triangles when Maitreya alone has 100k+ for the body only (no head, no hands and no feet), considering that Belleza looks at least twice as dense in triangle count visually and has at least 4 invisible layers too.
As previously explained, complexity is meaningless. I honestly wouldn't give two cents about it in any other Viewer other than my own its far too forgiving, random and punishes tiny things a thousand times more than real framerate killers. I suggest hand-picking people to jellydoll/derender based on their VRAM usage and polycount as well as perceived complexity. Wether to use one bigger texture or multiple smaller ones is a complex topic. Smaller but more textures allow you finer control over how detailed and resource heavy every single face is but they will incur texture allocation impacts, which are arguable how much they really impact the performance. Textures are loaded and decoded once unless they are released either because the Viewer has to or because you left the place (or looked away long enough). So technically the impact is a one time thing which would mean having more textures but smaller ones sound like a great deal but having one bigger texture with everything included isn't exactly bad either, it saves you a lot of texture allocations but might waste a lot of space depending on how much of the texture area you fill in for actual use. In general i'd rather say use smaller but more textures because it also allows the Viewer to mipmap them easier and separately making texture memory optimizations algorithms more efficient in doing what they are supposed to do.
@everyone Pulsar is right. Complexity is not what you should look at. It is essentially meaningless. I'm here to give you some Viewer-insight as to why. The Viewer calculates complexity based on what feature is enabled, how much textures and how big these are as well as how much polygons something has. The last thing, the polycount however is badly weighted. It favors mesh content immensely over anything else, disregarding any impact bad mesh has. From the original calculation i learned that it roughly counts 15+ (up to 80) polygons as a single complexity point. In comparison, adding just a single light (create a prim and set it to cast a light) will add 512 complexity. That is an equivalent of 512*15 = 7680 polygons at minimum and in worst cases (taking 80) 512*80 = 40960 polygons. That is unrealistically disproportionate. We are essentially saying that a single light, which doesn't even do anything at all to your framerate is as heavy as a mesh with up to 40.000 polygons, one that when seen will absolutely have a visible impact on your framerate (a small one but still). On the other hand, simple items such as prims have a minimum complexity (i think it was something around 1024 or so), even if that minimum was as low as just a single complexity, it would still be disproportionate. Taking the 15 polygon example and comparing it to a cube which has 6 sides, each side having 2 polygons meaning 12 polygons we would still have 3 polygons left to spare but still get 1 complexity. This disproportionate calculation drags through the entire way complexity is calculated not to mention it basically completely ignores rigged content and treads it complexity wise as static mesh which is completely wrong, rigged content has a lot of additional, very resource heavy stuff going on which is completely ignored here. All of this together caused a single avatar with 3.5 million polygons to have roughly half as much complexity as my own avatar which i spend a major deal optimizing all over the place (including textures) Example Picture So lets break down what is happening in this picture. Look at the top right, the purple avatar, that's me, i'm completely naked, no clothing, no extra attachments other than hair and lights and i get 68000 complexity with 160000 polygons. Sounds fine right? Now look at the bottom right, there are three avatars, one of them has 3500000 polygons but only 40000 complexity. This avatar has over twenty times more polygons than me, many times more attachments (clothing and all that) which use way more textures than me too and yet she only gets 40000 complexity a bit more than half of what i get for much less. How is this possible? Why am i being punished? I've reduced texture sizes. I've reduced texture amounts (combined textures where possible and re-used the same as often as possible) I've completely removed all invisible extra states i don't need. (even replaced the hand states with bento hands) I've reduced my invisible light prims to a single small pyramid (down from a sphere) and gave it an empty white texture, even made them visible to prevent them being calculated as slower alpha objects. What has she done? Nothing. Spammed everything possible on her avatar as she saw fit, probably incapable of doing any such modifications and optimizations anyway (due to no-mod) and yet there she is, having less complexity. It drove me nuts. The explanation is above. Here's the breakdown. My avatar itself is ~8000, my flexi alpha masked (once again an optimization so its handled as solid, not alpha object for faster rendering and no alpha flickering) hair weights 27000 alone, my lights on my back and head weight roughly 31000 thousand, the whole rest. Most of my complexity comes from a few tiny lights that most users can't even see properly because they are not running Deferred Rendering (aka Advanced Lighting Model) and are unable to see more than 6 lights at any time, even if they were, these few lights basically don't touch your framerate in any noticeable way. I went absolutely apeshit when i saw this and set out to change this faulty complexity calculation to heavily punish these high poly cases. I even went to LL to make a huge drama and start a big discussion about it with the other TPV devs which finally got LL to rethink it, they will be looking into re-evaluating the complexity values soon but don't think it's going to be good, their answer "we're not going to punish bad content" basically explained this whole project as dead before it even began so don't hold your breath. They are most likely going to do minimal number shuffling and at the end your complexity will be the same, 3 million poly avatars will still run around with 40000 complexity and heavily optimized avatars will still get punished for things like flexi or lights. TL:DR complexity is pointless, it makes so sense and favors heavy avatars over optimized ones and mesh over prim regardless of their actual impact. For anyone interested, i've changed the complexity calculation in BD and had much better luck with my own version of it, doing a much better job at jellydolling heavy avatars and keeping my framerate decently high even in heavily crowded places.
@Tankgirl Are you trolling or just that retarded? Boxy furries? Really? Just because you can see the vertice edges in some places doesn't make them bad. The avatars shown have stricken a good balance between details and optimization and you clearly have no idea what you're talking about nor what's actually good for anything. Brands like Maitreya and Belleza sell well because they have a monopoly on the market and its nigh impossible to bring something new up nor do most people care enough to do so. Add the unwillingness of people to change, their ignorance as well as simply being uneducated on how everything affects Second Life. Besides, i'm not working for Linden Labs and nothing of what happens in SL pays me my days. I'm working on my Viewer and in SL for free. I'm offering my Viewer for free, i'm offering help for free and i'm offering my very few select products and solutions for free. Unlike YOU i'm informed about things and unlike YOU i care about Second Life, i've spend a whopping 12 years in SL so far, 6 of which i dedicated to Viewer development for free. Neither LL pays me nor anything LL or the Market does. The only thing that pays me to some degree are the people that support me on Patreon. And. The Marketplace is but a tiny fraction of the money LL makes that keeps SL afloat. Almost all of what keeps SL afloat is land/region renting. Upload fee and Marketplace transaction fees are but peanuts and before you even think about it: No. If these brands didn't exist, SL would still be fine. We'd have different people with different brands. There will always be people doing something somewhere. If Einstein didn't find out all the things he did, someone else would have, who knows if someone thousands of years didn't already know what he did, just no one knew he did because he would have been declared crazy. Here, i made you a few pictures to show you the visual and the actual difference of the same snake avatar vs Maitreya. Snokra VS Maitreya The visual difference is laughable. Wireframe however shows that Maitreya has roughly 5 times as many polygons in the same space. Maitreya Complexity Snokra Complexity In case you don't believe me it's Maitreya. And just to make sure you actually understand what loads of shit you talk. Here's a TL:DR and ELI5 for you: You are trying to argue that Human bodies are fine as they are because they need to be perfectly smooth, so smooth that you can basically zoom inside them and still see no edges, this at the cost of many times more work for the rendering engine. I'm trying to show you that most Furry creators have long understood and figured out how to make mesh bodies that look decent and still be 5 times less work for the rendering engine than the Human equivalent. You are trying to argue that Furry bodies are bad because they look blocky to you when the actual difference shown is so marginal that its not even worth mentioning, but here, let me mention how marginal the difference is. It would be a trivial act of adding a few more polygons at critical bodyparts of said furry body meshes making them perfectly smooth while still being almost 5 times (4 times at worst) less work for the renderer than those human mesh bodies. You seem to think that these furry mesh bodies are absolutely, unchangeable and the way they are delivered is fixed forever. If and ONLY if it was really necessary to get rid of these minimal problems of eventually visible edges you could just go to their website, download the goddamn FREE FULL PERM .blend file, actually do the fucking work of adding a few more polygons where necessary without spamming subdivide onto everything (thus adding unnecessary polygons everywhere) and then re-upload that body for you complete with rigging and everything. But no, you rather come here, spew your uneducated bullshit and try to argue without any reason against optimization and i bet you are one of those people that get the loudest when it comes to SL's horrible performance. "Oh oh, SL is so shit its so slow, it needs a new rendering engine". I bet you say the same about every other engine out there because your brain is incapable of processing the thought that maybe, just maybe something else might be the root for the performance issue. You probably think the render engine is some magical device that can travel back in time and somehow magically can render everything no matter how much you throw at it. But hey, it's just the render engines fault. Surely its not the fault of those 20 million polygons, 3gb textures and incredibly badly designed objects in every scene.
@tankgirl I showed one and just ONE single body, all of these shots were the same body and all of them except the gfy was even the very same avatar. None of them had extreme proportions in any way. Have you ever even seen extreme proportions? Have you even taken a look outside? Seen any of the human bodies? They have much extremer proportions, so many humans go by these stupid "sexy" standards (bee shape or "90 60 90" comes to mind). What i've shown was by far one of the most normal shapes you'll find in SL. And "sexy" that's super subjective. I find these very sexy, on the other hand find most human shapes super unattractive. How about you show me a "sexy" human one that i can agree that it is sexy.
Not as far away from the truth probably. As far as i know most human mesh creators don't directly make their own stuff but rather "commission" their models by third party modelers, these modelers obviously not knowing much about Second Life and it's needs. They are then shoehorned into SL. Furry creators as you said need to create their own stuff from scratch for what they are doing, they need to have some knowledge and skill. It does not explain why they are doing it in the first place, they could just ignore optimizing and not give a fuck. They could just do it like any greedy money maker...sadly... luckily they don't.
@tankgirl How is there no comparison between furry and human avatars? Quite some furry avatars have this "beautiful, curved and sexy" just as much as humans do, furry avatars come in many shapes and colors. Just because an animal person is unrealistic doesn't mean it is not comparable to humans. You're forgetting that furries are basically furred humans with animal features, they do need a "human" body just as much as you do and they want quality and good look just as much, but compared to humans they get it right off the bat and full free modding on top of that. In fact furries often need technically seen a higher base poly count due to their special bodily features, tails, big fluffy ears, body/face fluff, legs and feet, they all offer special attributes that make them need more polygons, things that humans don't need or in altered, much smaller shape. Making a raptor leg/foot, compared to a human leg/foot for instance. Yet they manage to keep poly counts low and the avatars still smooth. Besides quite a few furries use Maitrya/Belleza etc too for their bodies and they have a noticeably bigger impact than anyone else in the room. Examples of good furry avatars from my discord channel: Mild NSFW: NSFW: This is an adequate level of detail. Sure looking close you can see that its not perfectly smooth but it doesn't hurt (besides, looking close you'll see it on any mesh body) LL making better replacement meshes wouldn't have changed a thing. It's an open market and there will ALWAYS be someone who wants something different. They would have made these meshes regardless. @everyone else The reason rigged meshes are so important, much more than static meshes you attach or rez on land is because the Viewer has to do a lot of additional work on rigged meshes, a static mesh does not need to be transformed at all, while rigged meshes need to have a lot of extra calculations applied. Imagine it like this: You get a huge list, this list details every single vertice in a model, vertices are the dots of which three form a polygon (triangle), now with this list you start building a statue, just like building your IKEA wardrobe (and having a few screws and nails left over). You're like "phew finally done". Your statue looks nice and is fully built. That's a static mesh for you. Now you get another list, this list contains information which of these vertices are rigged and how much and to which bone in the skeleton. You go through that list and color your statue according to that list. Blue for not at all, green over yellow to red for more rigging weight. You also write the bone name to each vertice. You think you're done but then suddenly your chef kicks your door in and tells you that you need to calculate the movement according to the rigging weight of each bone on this statue and the movement applied via an animation, he gives you a list of poses which if looked at in order look like an animation. You start going through the entire list of all vertices again, move and readjust every vertice, checking against the pose on the picture and the weight how much the vertice should be "moved" towards the desired pose, then check whether the vertice is weighted to another bone, if not you continue with the next vertice, you do this until you're done with the entire list... just to start over again and do it for the next pose. You continue doing this until you die of heart failure in high age. That is essentially how this works, now imagine this list to be super small. Just ten entries. You'll be done very quick, if you're really good you might even be done in just a minute. What if this list contains 1000 entries? 10000 entries? 100000 entries? You'll take longer and longer to do everything mentioned above, everything slows down. A GPU obviously is extremely fast and can do this process in a matter of milliseconds many many times with many hundred thousands of polygons. But adding hundred thousand polygons here, another hundred thousand there, each one increases the time necessary to work through everything. That's why high poly avatars are ESPECIALLY bad. Note that this above doesn't even account in post processing or any extras done with the models. Features like Motion Blur, Volumetric Lighting (based on shadows) and Shadows can (depending on their implementation) add a whole lot more time to this if they are based on the object geometry and need to sort through it in any way (not even going through all vertices but even with a simplified shape). Adding a few thousand polygons here and there can cause a lot down the line and you should take that into account which is why it is so incredibly important that rigged meshes (avatars) are optimized first and not just them but to an extend entire avatars too. However rigged mesh bodies are the highest priority as everything you make for them is often based on it, they inherit their base complexity in most clothing parts. You automatically create higher poly clothes for higher poly bodies to match it, right?
@Imagin And have you tried Sansar? I didn't bother trying it but a friend who did with a top of the line Intel Six Core and two, TWO GTX 1080TI's hardly got 60 FPS and it was buggy as hell, shadows looked ugly water had ghosting artifacts and the new lighting didn't look that much better than what you can do in SL with the right tools. Sansar is a fail. It's a long road till Sansar is anything "good". @Iggy You are indeed an idiot. Half of what you said is complete bullshit. This discussion is about client-side performance. Scripts do not impact client-side performance at all, at least not directly. Scripts can spawn particles which in turn will tank performance yes but that's particles being bad not scripts. Laggy furniture scripts are laggy for the server and while you're kinda right, they are shit, they are a whole different topic for another day. @Voodoo I sure do hope that Bake-on-mesh will reduce the usage of onion layer skins. It would cut massively into polycounts down to 1/4 or 1/5 depending on the body used. That's a lot, not to mention it could improve texture memory usage as well which in turn will show a bit in performance too. @Fred I'll take a look at it the coming days, from what you say it sounds promising. @Imp It is sad that people do not care. Everyone is digging their own graves and then complain about it. Also LL doesn't want to do anything about it because it would be against what SL made SL. "Your world your imagination". Yea but if your world is a lagging pile of shit for no reason, you can keep it.
@Amanda Sadly i don't know any examples of better human bodies. Sadly this means that it is currently a "pick your poison" situation. I've only ever worn Furry avatars so far and we have had the luck to have creators who know what they are doing. @just my 2 cents Correct but mesh bodies are the worst as they define the minimum. All clothes you make for them are based off of their structure and poly density (most clothes, especially skintight are often just "copies" of the body that are widened a bit, this means you copy the already high polycount and use it for your clothes). @Gattz The Material system is fine as it is, sure it is not as advanced as that of newer games but it is perfectly fine and enough to add a lot more details to something without inflating its polycount, people just need to learn to use it properly. Look at Orange Nova Avatars His last few avatars were absolutely amazing and made good use of the material system while keeping the polycount in the 20-30k's. Also check out Beev Fallen's stuff here: He does a really great job at tickling the last bit out of materials. His stuff looks absolutely gorgeous. I like visiting his place from time to time simply because its a very simple little house that looks many times better than any other house and runs much better than them too. The only thing we still need is a way to add cubemaps for reflections. @Silent Benefactor Havok is the physics engine it runs on the server to calculate all physical interactions, updating it won't do anything for SL other than maybe fix physics bugs or improve performance of the server when lots of moving stuff is going around. I'm running a GTX 1060 and an AMD FX 6300 (Six Core with 3,8ghz) on SL, i'm not sure how much more hardware you can throw at SL besides an Intel (or Ryzen with more instructions per clock and hyperthreading to hack more performance into the two actually used cores). My framerates are incredibly bad around people if i don't limit them via jellydolls and even then i'm often only running at 20-30FPS due to the place itself and the rest of the avatars. That's with Deferred Rendering and Shadows and SSAO on at all times. SL is simply at its limits here. Only two things can help SL's performance still, optimization in content and splitting off more processes into more threads (Multithreading) and the latter is very hard to do on an already finished product especially as complex as SL and is also a highly risky thing, increasing the crash rate tenfold if not done absolutely perfectly. SL is very unstable, we've seen many issues just because a tiny single then in rendering or some other system went out of order (the texture pipeline for instance) causing a supermassive chain reaction and ending in completely corrupted textures and finally a crash. Multithreading can and will have similar effects when things start to be split and run in parallel rather than in order. Everything would need to be rewritten to allow running "out-of-order". Much easier would be optimizing content, it's a matter of following a few simple rules and doing a bit of extra work once. In all the time these creators have already put into their items, a bit more wouldn't have done a big difference and it would have been an investment for the future. @Blaise Glendevon I kinda understand you, yes, old hardware is annoying and i don't like them pandering to it either, it has halted SL's development in many regards over the years since LL always needs to make sure that it runs for everyone, this means you need to handle extra cases, add extra code, make changes, make cuts and ultimately not make the feature at all sometimes. But as said above, i'm on the far end of the hardware ladder, there is not much above me and i'm absolutely struggling to maintain a decent framerate. You can't just toss better hardware at it and use it as excuse for worse optimization. Make a game that runs on no one' machine except those that were lucky enough to run an Intel 32 core with 128gb RAM and 4 GTX 1080TI because you spammed 20 million polygons on a small finger ring that's being hidden behind a wall and then tell everyone "your PC is just shit" is pretty ignorant. That pretty much sums up what these creators do. I had firsthand experience with them when i contacted the Maitreya creator. "Some PC's just can't handle our meshes" was her answer. Hell if right clicking with my hardware makes me drop into less than single digit framerates, what the hell are people supposed to do with slightly lower hardware and what about those with higher hardware? They don't have much more FPS either. @Stephen I highly disagree, they are not. They are a technical disaster and your argument is an excuse. I've seen LEA art installations with many times less polygons and still looking like art. More polygons doesn't mean better, game or not, it does not matter. This is a platform independent thing, you should never disregard performance. What's an image editing program worth if it runs at less than 1FPS and you get to see your changes every few seconds? It's useless. Same for SL, how can you even do the most basic things in SL if every letter you hit takes 2 seconds to appear in your chatbar. It is annoying and hard to use. This needs to stop. Many people could enjoy SL much more again if people knew what they were doing. Many friends and strangers i've asked who are just standing there in clubs or public places brought up the same answer, they are standing there, almost never moving because they literally can't, their framerate is so shit they would bump into anyone and anything in miles if they tried to move. They hardly even look around, they just stand there stare at the ground to at least be able to chat. @Pulsar I've already altered the complexity calculation several times and at least in my Viewer it does a much better job now both identifying heavy avatars and getting rid of them if you keep the complexity limit at a reasonable level. With it i can go to some pretty filled places and still get 20-30FPS where the default calculation would just seemingly randomly pick avatars, jellydoll them and still leave me with an unusable framerate. @Gingir Lowering polycounts does not mean you sacrifice beauty. These items have so many polygons too much that removing them would visually change nothing. Again look at the above mentioned Orange Nova Avatars, they have a lot of detail (especially the Snokra Snake) while having only a third of these bodies's polycount. For some of them you could easily add a bunch more polygons in critical areas without ever leaving the 20-40k polygon "greenzone". Also, LL has always been competitive because there is nothing else like SL and those that tried simply failed. SL is too old, too filled with content or simply too advanced already to be rivaled, you'd have to create a copy of SL where everything is better and fill it with tremendous amounts of content to even have a chance of a success... and at the end people will just go back to SL anyway. SL will go on for a good amount of years still and as such we should care for it, improve it, not just sit there and yell at LL for not doing it. They are already doing a lot of things many things of which are a direct answer to our complaining about our issues we caused us ourselves. I give LL a shitton of flak too when they break or fuck something up but it's not all bad and at some point you just have to accept that they are just trying to keep this self-destructive environment running with all means necessary. They are literally catching crumbling wall pieces and glue them back on while we sit inside and smack the walls with a sledgehammer. @ItsemeMario You're just ignorant. Your behavior is the very reason nothing has changed. I AM complaining, ME the one who made a very Viewer for high end hardware, the one with some decent "high-end" (if GTX1080 is HIGH-END, the 1060 is pretty much almost the max) hardware himself, I AM complaining that this is bad. Not some lowie laptop hardware dude. I tell you a little secret. Those with crap hardware don't complain. They are used to it, they know they are sitting at the lower end of the spectrum, they know they should be lucky to even still be able to run shit, they have accepted the fact that they are "unsupported". It's not the low-end people complaining, it's those with decent hardware, those that would expect much better performance from their hardware as seen EVERYWHERE else. Pull your goddamn head out of your ass and look up WHO is actually complaining before spewing your BS. You would have known that it wasn't just some rando-guy saying something.
@Confused Did you login to the Jira? You have to click "watch" on the right side where it shows the count of people watching it. That's essentially "voting". Although it's already through and LL is now seeking help in their Forum to confirm this.
@Sirhc As Amanda again pointed out. You would face this issue too the moment Singularity would update to animesh if they don't catch this issue either. LL would be the first one to face this issue (besides my own users, although i already fixed it in my latest release and it seems like LL accepted it and will be fixing it too, hopefully before release). There's one thing though... Cool VL Viewer was always faster than me, i'm sure Cool had Animesh for a while now... i haven't noticed Henri saying anything about it yet, i wonder if he found it... i mean if he did he'd have probably made a big vocal thing out of it just like i did so i'm sure i would have heard about it in the opensource mailing list. If not even our almighty Henri noticed this (that is if he even could, considering that he often rewrites large parts of everything coming from LL because everything they do is trash to him and he has the skill to completely rewrite it)... So many questions. Anyway, it's accepted and soon going to be fixed. Phew.
@vwfan That's correct. Worst case (as it is right now) no one will notice, since no one hasn't noticed yet even with the huge amount of testing of animesh. Then it will go to FS, FS will be blamed for their shitty stuff once again and boom everyone got a smack in their face because one single tiny thing slipped through testing. It might get fixed someday or it might not. Problem is Lindens often don't use the Viewer they are developing enough to make such concrete observations, they are way busy with other stuff like getting this stuff to work in the first place. So it's up to testers to see this stuff. Problem is because most people use FS they will disregard the possibly lower framerate as it simply being "LL's Viewer", besides most people testing the Viewer for what it was made = animesh, rather than normal usage. This causes no one to notice until it hits their own Viewer which in case of FS with its extremely long release cycles could be a problem. Even if they were to notice it shortly after release, you most likely wouldn't see another update until another half a year. Most users will assume that it's simply FS's fault (and as much as i like the thought of that) and their usual degradation of quality and performance over several updates, which if you followed FS's development, you'd think is "normal" meaning you start ignoring the issue, no one might investigate and this thing will have sneaked into all Viewers reducing everyone's framerate for no reason. And this is how fast shit can hit the fan.
Lagomorph7 pretty much brought it on point. But i'd like to say something about it anyway. While the Viewer may obviously not the best choice for most people (especially with low hardware specs) i've done a fair share of improvements all over the place, everything i touches has put a lot of thought behind it and often this is not enough, many things will still need feedback which is why some things look and feel rough at first but become better over time as i come up with better solutions. A fast development cycle is a huge boon for this as it allows me react quickly to issues and feedback. However what most people seem to miss is that the Viewer is simply not made to emulate any other Viewer (for several reasons), if you want to use the Viewer you'll have to get used to different approaches, while these usually are limited to smaller changes, they might still look overwhelming at first because there are simply so many. The biggest problem with the Viewer for most users seem to be that the Viewer is simply not hand-holding you by any means, it throws you into the water cold and lets you figure out the rest while occasionally just giving you hints for what you're doing. For people who are simply not trained get along with this "exploration" styled approach which has been slowly driven away in gaming as well (games often being dumbed down more and more) may have serious issues getting along with having to "explore" on their own without help. I realize the Viewer is not for everyone but that's probably for the best as a big userbase would not only cut into my free time but also would eventually get LL onto me and will force them to take action on certain things i'm doing that i wouldn't be able to do with a lager userbase, such as Firestorm is basically not able to do certain things because LL is looking very closely at what they do. This goes as far as Linden Labs loosening some rules for others (such as me) to the point that we have a much more open field when it comes to developing new features. Best example seems to be the very thing that makes BD so special, it's graphics, i was told from Jessica directly that they are literally not allowed to implement any of my graphical changes as it "technically" violates the shared experience rules which they have to follow strictly unlike me, so the moment my Viewer would become the most popular i'd have to commit my changes to LL (and they will never ever take them, they are basically impenetrably resistant to graphical changes because they believe that changing even the slightest thing about SL's look would make users immediately run rampage like they did several times in the past), give up on them or face backlash and potential de-listing of my Viewer from the TPV list which as you can imagine wouldn't look all that good.
@Hamlet I'd love to but i'd prefer anything else over Twitter, i'm using the Twitter account just to write here because it was the easiest to log in with. Preferably directly inworld would be the best. @Marta You might have updated from a non AVX version to an AVX version without your CPU supporting it or from 32bit to 64bit without your OS supporting it. Did you change computers (from PC to laptop?) when you were on holidays?