DanielRavenNest Noe

Zalandria Zaius wrote: The possiblities are endless but there is that 8 material limit that they don't seem to be catching and erroring on. Just a note for anyone testing items on Aditi - only 6 of 8 materials are functioning at the moment, the other two remain grey. That was reported as an issue, and Kelly Linden noted that it was fixed in the JIRA, but the fix has not made it to the servers / viewer code that I know of yet. I have some 8 material objects I check with periodically. EDIT: the problem seems to have been fixed now, visiting Mesh Sandbox 28 and going back to Sandbox 30 may have done something, I had usually worked in Sandbox 30.

No, it is not wise to attach such high poly models, and it is totally unnecessary. The smallest thing you can see on your monitor is 1 pixel. If you are zoomed into an avatar so you only see half their body, and you have a large monitor, they might take up 500x1000 pixels on the screen, and the full body would then be 1 million pixels. If you have two million triangles, that means each triangle is smaller than one pixel, which is too small to render. Past experience with another virtual world that uses mesh for everything shows I can make very nice clothing items for under 5000 triangles per piece. For small jewelry that will only take up 50 or 100 pixels *at most* on the screen, you can get by with far less. Anything more is vanity or for taking marketplace photos. Here is a pair of jeweled stilettos I made for under 5000 triangles, and even at higher magnification than most players will see it, you can't see any defects from triangulation: Jewel Strap Stiletto @Leliel - Anyone wearing over 10 million triangles would (a) grind their own graphics card to a halt, (b) have to pay an arm and a leg for upload costs, © be abuse reported like any other method of bringing a region to a standstill. So while 600M triangles is theoretically possible, it is unlikely to happen. On my land I would eject anyone wearing over a million triangles for the same reason I would eject them for carrying 500 scripts: excessive lag. Oh, when making your absurd worst case scenario, all 40 avatars would have to be in a group hug, so they are all at the max LOD, otherwise the triangle count goes down, and in that case occlusion cuts down what has to be rendered, so it can never happen.

I shrank my palm tree down until it's just a garden plant size, then linked three copies plus a box prim to make a decorative planter. In separate pieces it costs 7 prims (2 each for the plants, 1 for the box). Linked together it costs 3 prims, which is not bad. The plant model is around 900 triangles, with a billboard for the lowest LOD and physics. The prim costs still get outlandish when you use mesh for large objects, but for small things they can be reasonable if you link things together the right way and make a sacrifice to the prim gods: Photo

Use the 2011.3 FBX exporter. The 2012 exporter creates a Collada 1.4.1 file, and the SL upload can only understand Collada 1.4. Missing bones is the most likely problem if you did not rig the full list of 26. If you weghted skin vertices to only some bones, add the missing bone names into the collada file bone list after the ones you actually use, and set the array count just before the bone names to 26. If you did use all of them, the most likely issue is the bone names have to match what SL uses internally or the bone hierarchy is different. As far as export settings, the only thing I have to make sure is the axes are Z-up, other ones I leave at default. I use Max 2010 mostly, and learning 2012, but a lot of plugins and compatibility is not ready yet for the latest version (including SL).

Just make tiny human mesh avatars (like 1/4 normal size), and scale down the houses and furniture to match, it will cost a lot less prims that way.

It's a small palm (and now it costs 12 PE). About 900 triangles for the high LOD. Small Palm Tree To show what can be done with a more relaxed triangle budget and a good graphics engine, here is a scene I set up with the CryEngine 3, with 750K triangles per frame. There are two kinds of grass tufts, short and tall, of around 600-800 triangles each. The ones in the upper left corner are turned into "sprites", meaning a single flat plane with a picture of the object. A sprite faces the camera, while a billboard (as used in SL) has a fixed position. The ones in the foreground are rendered in full. All the plants are copies of just 3 objects, so "streaming cost" would be very low if the data was being sent online. After sending the 3 models and their textures, the copies only have to specify position, rotation, and scaling, which amounts to about 16 bytes per copy. The terrain texture also has two levels of detail. At long range the snow and rock texture on the mountain is displayed. At close range a grass and clover texture is displayed to fill in between the 3D grass tufts. Note that this has a good frame rate with shadows turned on (35 fps in the statistics top right) on my GTX-260, and that is with post-processing (the bloom from the overexposed snow), and *animating the plants*, they all move in the wind individually. Cryengine 3 Scene

Vivienne Daguerre wrote: What do you mean by billboard like mesh? It means replacing a 3D model of an object with a flat picture of the object on a two triangle flat rectangle. billboard example It's called a billboard because it resembles an outdoor advertising sign. It's also called an alpha plane or sprite sometimes. You can get away with it at longer view distances because a model of the object and a picture of the object (with transparency for parts) end up looking about the same.

It works like this: with sculpt maps, in theory you can get 2048 triangles for one prim. With mesh I can get around 100-300 triangles per prim for moderate size objects (it depends on a lot of factors), so roughly mesh starts out with a ten times higher cost for the amount of detail. So you have to work very hard to find cases where mesh wins on land parcels. For avatar attachments it's a different story, prim count does not matter there.

We have been charged upload costs for as long as I have been using Aditi, but the L$ balance there is "fake money", and you can get more by asking a linden staff person. And yes, the unfinished state of costs is annoying since I don't know how much I need to optimize things, or even if they are feasible to make with mesh. Nobody will want my mesh palm tree that costs 9 prims when they can get a nice sculpted one for 1 or 2 prims.

Upload Enablement

The only proper way to test render speed in SL is to actually use slower graphics cards in suitably filled regions. Unfortunately I gave away my 9 year old PC with the Geforce 3 that I first started in SL with. So I don't have a slow card to use any more. The best I could do was fill up a Mesh Sandbox with high triangle objects and see what happens with my current card (Nvidia 260), and then estimate how that scales to other cards. I'd be happy to set up a bunch of objects as a test area for other people to visit and see what they get.

Actually, I know the builder of that region, and he has done even more amazing work for Blue Mars, which is an all mesh-based world. So I think it is likely given the opportunity, he would rebuild in SL. Better tools make better results. That has been my personal experience also.

Your results are in accord with what I see in SL and other graphics engines. A reasonable graphics card can render a lot more than 250K triangles without slowing down. From my own tests, I extrapolate that an Nvidia 7600 class card (relatively slow these days), should be able to handle 800K triangles on the SL low graphics setting. For comparison, the CryEngine 2 recommended limiting scenes to 3-4 million triangles, using a more complex shader with multiple texture maps/triangle (typically 3) and would perform well with a GTX 8800 class card. That would translate to 6 million triangles in SL because we only use 1 texture map/triangle. And whatever the parcel costs of using mesh objects, if avatars are not charged for heavy use they will load up on high detail meshes just like they do on prim and sculpt heavy items now. Therefore people with weak graphics will still experience lag in busy areas. The bottom line is the "cost" being calculated for mesh is so much higher than for prims and sculpts that people will not use them on parcels. A penalty of ~2x on prim cost would let the other advantages of mesh compete (custom UV maps, 8 textures vs 1 for sculpt), but with the current penalty of up to 10x they will simply not be used for many items.

Must you always interject stupid comments? They could set up a new grid and allow all existing content, merely with new cost calculations. It can even be the same grid, just some regions would run a different server and different cost calculations. This is done already, where Homestead regions get a different prim allowance calculation per parcel area. So it is not beyond the realm of possibility to have other differences in calculations between regions. For people to choose the "new" regions rather than "legacy" regions, there has to be some advantage to do so. The advantage could be in cost, performance, or some other features. But if there are no advantages then nobody will choose them.

According to the Grid Status report page, there is a rolling restart today, but there are no release notes saying what what changed.

I'm not sure if Ann considers me a cheerleader. I don't consider myself that. I think of myself as a reporter of mesh status, a helper when I can, and a tester of the software. Whether I will build with mesh myself depends on how the final version turns out, including costs. If the Lab's goal is to add a new building tool, my opinion is Mesh will suceed at that. Mesh uploads and display have mostly worked the last 6 months, and although creation has it's quirks, several 3D programs are able to create compatible outputs. If, on the other hand, the goal is to improve the looks and reduce the lag within Second Life, Mesh will not suceed given the current state. Terrain, landscaping, and buildings are discouraged from using mesh objects by the cost formula. So the general look of SL will not change much. Meanwhile, avatars will be able to load up on complex mesh attachments, so the avatar-induced lag from current complex attachments will not go down. I don't make a recommendation to other people to use or not use mesh objects. I leave that up their creative interests and parcel prim budget. But I do expect people to continue to behave in SL the way they have in the past. That includes wanting their avatars to look good, and to care about parcel prim count. So I base my expectations on that behavior.

As a large trader, I can say those fees are the same for everyone, large and small. The majority of it is due to credit card transaction fees which Linden Lab has to pay when people buy L$ with a credit card. If they did not assess a fee on the exchange for transactions they would actually lose money running it. As far as I know, the buy fee is charged if any part of a limit buy is filled, but not if you cancel before any L$ are delivered. For a market buy, those usually get filled right away, so the question does not come up.

Drongle McMahon wrote: Second, I have tacitly assumed that the developers strategy would be satisfied with the users of low end machines receiving satisfactory performance with the low or medium settings. They can quite easily counterv the argument by insisting that low end machines must perform with higher settings. I think this would be misguided, but it would be a logically consistent view. I tried rezzing multiple objects until I had 1.2 million triangles in the scene, with the default low graphics setting I am getting 160 fps. That is for an Nvidia GTX-260. So scaling that performance down to whatever you consider a "low" graphics card to target for should give you an acceptable frame rate and triangle count combination. Unfortunately I don't have a low end card any more to test with.

Atashi Yue wrote: Apples to Oranges though Daniel - your scene isn't streamed live is it? It was built on my PC, but it could be set up to be streamed from a server. The functions exist in that engine to deliver files incrementally. The data for that scene amounts to 17 MB. 12 MB for the terrain map, which is 2x2 km (64 SL regions), and 5 plant models and textures for the other 5 MB. There are only 5 kinds of plants, but they cover the entire map using "instancing", so there are roughly half a million copies used, mostly grass tufts. Each copy only needs to record the location. It comes to 265 kb per SL map region of area, so it's actually a pretty efficient way to landscape. Of course, if you have more different kinds of objects that will be more data. What I was talking about was PC rendering though, which does not care when the files arrived. In Second Life the objects and textures are delivered after you teleport to a location. Most game engines deliver them ahead of time on a disk or download. Either way, when it comes time to render, the geometry and texture files go into the graphics card, and what matters is how many triangles and how complex the shader is.

Drongle McMahon wrote: If the triangle budget is motivated primarily by the aim of controlling the resource consumption for the users of lower end machines, then it may be reasonable to suggest that the calculations should represent the settings they are likely to be using, rather than the relatively high level settings implicitly used. This would result in streaming costs about five times lower than the current calculation (without correction) yields. This is what has bothered me about the cost formula. I have been testing the Crysis 2 / CryEngine 3 MOD SDK and set up a scene with terrain and plants.. The attached photo is on the "Ultra" graphics setting (the highest they have), on a very demanding graphics engine - all the plants are animated for example, to move in the wind. It's a bit hard to read the statistics in the upper right corner but there are 500K triangles, and I am getting 30 frames per second, which is very acceptable. Basing the SL formula on 250K triangles for a much simpler rendering engine is much too restrictive. In effect you are preventing people with decent PCs from actually using them, for the sake of people with slow PCs. http://lh5.googleusercontent.com/-hkqEOf35KMg/Tg-WkPDtOII/AAAAAAAACkc/nvqCsYfiZu4/s1600/DemoLevelSunset.JPG

I have hundreds of models I created for the Blue Mars world, about evenly split between house parts and clothing, and I have tested a variety of other item types (rigged avatar, terrain pieces, plants). At this point I am testing new viewers as each point release comes out (2.7.4, 2.7.5, ...), and reporting problems when I find them. I'm waiting to see how the cost issue works out to see which of my items will be even feasible to market. Right now that looks like avatar attachments only, since parcel rezzed items are heavily penalized relative to sculpts.

It's not only hard to understand, it's dangerous to users if merely resizing or linking causes object return. People do not expect that and will be very unhappy when it happens. Although it may not be a perfect model of the cost of a mesh object, a formula like "300 triangles = 1 prim" is easy to understand, both for designers and users. It will eliminate the danger of linking or resizing, and still penalize complex meshes so that we will try to make efficient ones.

If you only used one material, that doesn't cause a problem. Try a cube with a different material ID on each side. That fails to work properly if you export direct from Max 2012 --> dae format.

Vivienne Schell wrote:. Allow fixed sizes, no mod, and done. Easy, fast, fun. So now you want to break the permissions and build system? What happens when such a no-mod mesh is linked to non-mesh items? Does the whole thing become no-mod? You are just exchanging one kind of worthless (it costs too many prims) for another kind of worthless (Even the original creator cannot edit the item)

When you do your review, please consider a few things that I don't think went into the original calculations: - The original formula was based on a target number of triangles in a full region. Many users have slower graphics cards (even Intel integrated), and so do not see a full region. They have to use a low draw distance to get a playable speed. - I don't think occlusion was factored into the formula. Some objects will likely be occluded by other objects or the terrain, reducing the number of objects to be downloaded and drawn. Larger objects, and especially ones you go inside tend to occlude more, so should not be penalized as heavily for size as they are now. - Aside from basing a formula on technical performance, some consideration should be given to the *business* benefits of mesh. A better looking world, or one where you can do more with a small land parcel, should attract and hold more people. If mesh is costed such that very few people use it, you lose the business benefit. - Moore's Law (the continuing improvement in computers). Both servers and PCs at home should continue to get faster each year. Mesh will not displace 100% of grid content in a day, that will take years. So the target cost formula should consider that by the time mesh is widespread, computers in general will be faster than they are now. - A cost formula which is difficult to understand, which the current one is, and can accidentally return objects by exceeding the parcel limit when you link, resize, or change parameters, does not encourage it's use.