Getting Closer - but Help on Textures for Mesh

[Toysoldier Thor]

Thanks to those on this thread that gave me a lot of hints and clues and ideas on how to get the basics of bringing in my first MESH from my 3D tool. 

Since then I have been really strucggling to grasp the development pipelines.  Not only trying to learn how what and how to best reduce the topology on Voxel type 3D modeling tools like SCULPTRIS, or how to best create a proper UV Map.  I also put a lot of effort in trying 3D Coat which seemed like an amazing tool but I ran into some basic functions that didnt work.  Turns out I discovered that the TRIAL version of 3DCoat has a lot of these valued features not available so I had to abandon 3DCoat.  I am not sure why a company releases a TIMED TRIAL and then also removes leading edge functions from the trial.  Now I have to walk away from it.

Happily, yesterday I went through my first successful (sorta) end-to-end MESH creation, texturing, and import into SL with pretty good results.

Here is the result of what I created.

 

The title summarizes it, but...

 

• I used SCULPTRIS to hand sculpt this model from scratch out of my head (took about 4 hours)
• After the model was done, I reduced the 300K + vertices down to about 6000.  I might have gone a little too far as there is slight boxiness on a couple parts of the model in SL and not sure how this effects textures either.
• After I reduced the model to a low res, I exported it to OBJ and imported it into ZBRUSH where I used the pretty cool and easy to understand UV MASTER zplugin tool.
• I then Exported this UV'ed model to OBJ from ZBRUSH and imported it back to SCULPTRIS
• I went into SCULPTRIS Paint mode at 2024x2024 resolution and painted the model (with some korny blue rock-textured look)
• Once I finished painting, I exported back out to OBJ and also exported the TextureMap, Normals, and Bump images.
• I imported the OBJ to Blender and exported this model to DAE format.
• THIS PART I AM STILL NOT DOING RIGHT BUT... I took the texture map, bump and normals images into Photoshop and blended them together to get that similar texture skin that I could see in Sculptris.  I cant come close to getting the same look but I created a resulting TEXTURE of the skin of the model.
• I went into SL and imported the model.  Since its a statue to be seen up close, I set the model's med to 100 vertices, lower to 10 and lowest to 10.  I set physics to lowest.
• The upload fee was 11 and the LI was .6
• I imported the texture for the model too.  Althought the texture from the model fit perfect with no visible seams, it just does not have convey the same 3D texture look like in Sculptris.

 

So can someone please help me with what I am doing wrong with the texture.  Whe Zbrush exports an OBJ it seems to even include the texture skin that was on it.  I suspect if I could paint in Zbrush like I did in Sculptris, the OBJ export would include my texture and it would come into SL via the DAE.

In Sculptris all I seem to be able to do is create the 3 critical texture components but not have them baked into the exported OBJ.  What is the right way to get my textures baked onto the mesh?

[Ann Otoole]

and after all that was it worth it to even bother? The fact of real life is what will kill SL. Just not effing worth the effort anymore.

[Drongle McMahon]

Until they do the materials project, you can't use normal or bump maps properly. So you need to bake the texture from a rendering that has these effects. Since their effects are dependent on lighting, that means that ;ighting will be baked in. Unless you are putting them in an environment that replicates that lighting, that will conflict with the inworld lighting.

The picture you show looks like it has inworld shiny applied. This is a very crude approximation. It ;ools smooth because the mesh does not have any bump or normal variation. It will usually not work with baked lighting. If you do manage to get the baked lighting, don't use shiny. It will tend to hide the baked effects.

You can bake in the normal and bump maps, under particular lighting conditions, in Blender. This combines the effects of the colour, bump and/or normal maps so that you can simply save a single texture to apply inworld. I can't believe the same thing isn't possible in the software you are using. Hopefully someone who knows will soon be here to tell you exactly how.

[Toysoldier Thor]

Well at the moment and also at this time of SL's MESH evolution - NO since I have already noticed that when I brought the two models I have created into SL, easily more that 50% of my friends that I wanted to show my newest creation and accomplishment ... they couldnt see it. 

The majority of my many friends in SL still refuse to update their viewers to a version with MESH.  The latest Phoenix 1.6 is very laggy.  Firestorm seems to have more bugs in the latest release then they fixed.  And well, myself and the majority of SL residents just dont like LL's own Viewers or wont ever give it another chance after we got such a bad taste of V2.

But... has the effort been worth it... YES.  It has been a very painful process and very technical and complicated and with countless options on how a Mesh can be efficiently created and brought into SL.  But, I have always loved working with the freedoms of 3D modeling - freehand.  Thats why I love the voxel type 3D tools like Sculptris and 3D Coat.

Another thing is that if I can start really getting my mind wrapped around UNDERSTANDING how all the components of mesh models work (i.e. UV, retopologization, hi low rez polys, texture mapping....etc etc.) and become truly effective in it, I can release the technical shackles on my art and creative talents and make some amazing things in SL.  Ohh BTW... things that can sell in SL.

The massive technical complexity of Mesh creation for SL is not only a disadvantage - its a blessing in the future.  In the past 2 years I have made over $10K US from the creation of only 6 packs of Landscape Sculpties.  The last pack I put to market was in August 2010.  These packs continue to sell about $70US a week and I never go to my store in SL or have to spend more than 5 hours a month being an SL Merchant. 

Can you imagine with the complexities of MESH how much potential new revenue there will be once LL decides to relax the LI for MESH objects that grow in size?  Right now the biggest limit for me making MESH landscapes is that it cannot compete with Sculpty for large scale rezzes (rivers, mountains, waterfalls, terrain).  I am confident that soon LL will realize this and not penalize Mesh for being large.  I HOPE.

So... generally... YES this will be worth it in the long run.

[Toysoldier Thor]

---

Drongle McMahon wrote:

Until they do the materials project, you can't use normal or bump maps properly. So you need to bake the texture from a rendering that has these effects. Since their effects are dependent on lighting, that means that ;ighting will be baked in. Unless you are putting them in an environment that replicates that lighting, that will conflict with the inworld lighting.

The picture you show looks like it has inworld shiny applied. This is a very crude approximation. It ;ools smooth because the mesh does not have any bump or normal variation. It will usually not work with baked lighting. If you do manage to get the baked lighting, don't use shiny. It will tend to hide the baked effects.

You can bake in the normal and bump maps, under particular lighting conditions, in Blender. This combines the effects of the colour, bump and/or normal maps so that you can simply save a single texture to apply inworld. I can't believe the same thing isn't possible in the software you are using. Hopefully someone who knows will soon be here to tell you exactly how.

---

Yes the model I am showing is not the model that had the baked texture on it.  It looked so crude that I didnt want to show it.  The model you see was just an inworld seamless woodgrain and Shiny on LOW to show the form better and to hide the seam a bit.

Thanks for telling me this... I thought I was doing something wrong.  So the best I can do is to do what I am doing now.... take the texturemap, bump map, and normals map and try to fake the texture surface best I can.  So I have pretty much figured out the "PIPELINE" then?

If someone knows how to get Sculptris to bake all 3 texture elements into 1 image - that would be awesome!

[Drongle McMahon]

"I am confident that soon LL will realize this and not penalize Mesh for being large."

I'm afraid your optimism here is probably misplaced. The developers have a very strong commitment to maintaining the resource-consumption-based LI scheme, and this kind of size effect must always be a result of that. The only way around it is to be very aggressive with higher LOD meshes for large objects. I'll put my graph here again. It shows the relative download weight cost/saving per triangle cganged in the four LOD models and how that depends on size. The size is represented by the object "radius", (sqrt(x*x + y*y + z*z))/2, on the horizontal axis. That is the length from the center to a corner of the bounding box. You can see that for objects with 10<r>20m, reduction of triangles at medium LOD has any effect, while the lower LODs have no effect at all. For such an object, the medium LOD is not seen until you are >40m away.

[Toysoldier Thor]

---

Drongle McMahon wrote:

"

I am confident that soon LL will realize this and not penalize Mesh for being large."

I'm afraid your optimism here is probably misplaced. The developers have a very strong commitment to maintaining the resource-consumption-based LI scheme, and this kind of size effect must always be a result of that. The only way around it is to be very aggressive with higher LOD meshes for large objects. I'll put my graph here again. It shows the relative download weight cost/saving per triangle cganged in the four LOD models and how that depends on size. The size is represented by the object "radius", (sqrt(x*x + y*y + z*z))/2, on the horizontal axis. That is the length from the center to a corner of the bounding box. You can see that for objects with 10<r>20m, reduction of triangles at medium LOD has any effect, while the lower LODs have no effect at all. For such an object, the medium LOD is not seen until you are >40m away.

---

So in my limited understanding of how MESH LI engages for LOD and the size of a Mesh, this could mean two futures for Mesh being used a the best option for landscaping...

If I were to build an equivilent MESH Rock Terrain formation (like I have with Sculpty at a fixed LI=1) and I size it up to be a background mountain on a sim....

Chances are that even though the LOD of Med and Low could be set real low (because the mesh prim will be so large that chances are the Avatar will always be within 40Meters of the mesh so the highest LOD will always be engaged), the cost of a mesh that is the size of a sim mountain will be so high that my SCULPTY LANDSCAPE PACKS will always be the cheapest and best options for my customers.

Worse yet is that the MESH Mountain will be far enough away that MED and LOW LODs will need to be pretty high or the mountain range from the distance will turn into an ugly blog.  So a bad situation for Mesh regarding Landscapes will be even worse.

This issue is what I have argued about before on the weakness of LL's LI deployment on Mesh... for large terrain and landscapes, the Mesh model will never compete with the Sculpty that has a fixed LI of 1 no matter how large it gets.

 

 

[Toysoldier Thor]

PS... I cannot get real aggressive on the vertices because if i do in order for a mesh to compete details with a sculpty - the mountains will look like boxes / blobs.  Alternatively... the Mesh mountains will look great but it will cost the customer a 100 prim to rez it.

When a sim builder / landscaper is placing a forest with mountains and rivers - and has tons of high-prim trees, he she will be looking for prim savings - and mountains made of scukpties will be their best bet.

[Toysoldier Thor]

I just did a quick test - now that i know the basics of importing mesh. 

I took one of my Shores & Ridges terrains Sculpted maps from Zbrush.  Remember that these maps are already extremely low rez degraded in order to be a sculpty with 32x32.  I simply exported it as an OBJ then imported to blender and exported it into DAE and brought it into SL.

The default LI was 17 to upload and 8 to rez for a small size of only 1x1x1m (High=4096 verts, med=1338, low=349, lowest=180).  To help the poor crippled mesh out - I even reduced the LOD of med/low/lowest to 100,50,10.  This wont be good for a large landscape terrain but ohh well....

The upload cost was 14 and the LI was 1.8.

Loaded it in.  First of all it looked terrible compared to the sculpty.  You could see every line in the mesh where the sculpty was all smooth.  Next I scaled the rocky formation even to a relatively small size for a terrains - 12m x 12m x 1m

Where the sculpty was still looking good and only 1Prim.... the mesh was now... 10. 

My Customers use my landscape sculpties at sizes now of 10x50 or even greater.  You could imagine how useless MESH would be at that size.

So... seems that mesh will only be good for small high detailed items like making clothing and statues.

[Drongle McMahon]

Yes. For things larger than 45m, where only the highest LOD counts, sculpties will always win. I think they missed a big opportunity when they increased the official size limit to 64m. They could have made mesh-type accounting apply to anything with any dimension greater than 10m. That would have made huge sculpties cost what they really consume. Existing megaprims would have been affected; although simple boxes would not have increased cost, sculpty megaprims' LI would have skyrocketed. That would have been tough for people sellling or using sculpted landscape, but megaprims were never officially supported, so the sellers and users of those were on notice of such a risk.

Now the opportunity is missed for ever and sculpties will go on consuming vastly more gpu resource than they cost, and people on low end machines will go on suffering the associated lag without realising where it comes from. I think the sculpty megaprim is probably the main thing that made them determined not to repeat the mistake of resource-independent costing with mesh. All that is very well, but as you say, as a merchant you have no choice but to go on providing the sculpties (that is, if you want to sell anything).

 

[Medhue Simoni]

Toy, good to see you here, your path down mesh creation has just started, try not to get ahead of yourself. First you have to understand LODs. By the examples you are giving us, it shows you don't fully understand LODs yet. Second point is that you have not talked about physic shapes at all. This is 1 of the most important factors pertaining to a mesh.

As far as efficiency goes, using a sculpty for large area landscaping is probably 1 of the most resource intense things you could do in SL, depending on how it is done. Go load up a whole sim with huge landscape sculpties and try to run around, your pc will be going like mad loading those objects and you will see tons of hesitations and freezes. Compare this to everything being properly made meshes, and not only will it look better, but you can run around and not have your pc freeze at all, or those freezes are so brief, it is hardly noticable.

Are there limits? Of course there are, especially compared to a system that had almost no limits (sculpties). These limits are not what you think they are tho, because you don't fully understand how to use mesh. The reality is, for most use cases, a well made mesh creation is not only going to be more efficient, but also less prims with more detail. Your job, as a creator, is to figure out how to make that happen, given the tools. Beyond just 1 individual mesh tree to sculpty tree comparison, or similar, I have yet to see a situation where a mesh is not the better choice. The more complex the scene, the better mesh is for it.

[Toysoldier Thor]

Yes i do understand Physics on the mesh and that will only make the $ Cost advatage worse for Mesh for large scale landscapes that we have been talking about.  Also, seems your opinion and optimism on Mesh as a better solution than Sculpties goes against Grongles beliefs who tends to agree with me.

LL has made Mesh not feasible for SL's grid.   We are not talking about Lag impact here Medhue... we are talkin about MONEY.  When you have a sim or parcel of land with a limited amount of PRIMS... Mesh mountains are basically stupid compared to sculpty mountains.

LOL you are telling me to go to a sim and run around where there are sculpties mountains, hills, water features, terrains.....

Did you forget I have made and sold Sculpty Terrain packs for over 2 years?  Did you forget that the 1000+ customers I have that I have gone to many of their builds are my SCULPTY LANDSCAPE customers?  Did you forget that prior to mesh showing up las august that 100% of the landscape features that were not terraformed were sculpties?  I have lived in SL for 4 years and as a SL Photographer of landscapes - I have visited countless sims made up with only sculpties and prims - trees, rivers, rocks, mountains, off-sim features, etc.

I know what sculpties do to sims... and seems to be ok for the past 3+ years.

I also know that when it comes to my landscape customers buying a Sculpty mountain for 1 Prim vs a Mesh mountain for 20 50 100 prims...  $ Talks.

Until and IF LL ever levels the playing field between mesh and sculpties for large scale content like what i make... I will focus most of my mesh playing on making smaller things like statues and freaky lil things.

No rush right now anyway as most users in SL still cannot see mesh and refuse to upgrade to mesh viewers.  I now know that for a fact when I tried to get many of my friends to see my first 2 creations.

Mesh is still just a cool thing to experiment with...   but I will admit... its fun to do some experimenting.

[Drongle McMahon]

"Mesh as a better solution than Sculpties goes against Grongles beliefs who tends to agree with me."

What I said was that as a merchant, you couldn't really expect people to buy mesh versions of what you are selling. I wouldn't go so far as saying that sculpties are a better solution. I don't think that. They win out only because they are not costed according to the resouirce they use, while mesh does. For landscape stuff, there are many aspects where mesh do have advantages, especially the collision shape. If you use undecomposed physics shapes, the physics weight gets smaller the larger the triangles are. For a given mesh, there is a crossover point where the triangle physics weight becomes less than the download weight. I won't go into the details, but if you want good collisions or things to walk on, mesh can be the answer.

I was puzzled by your saying the mesh version of your sculpt was more jagged. So I did a similar test myself. One point is that the mesh has to have just 1089 vertices (planar topology sculpt)., not the 4000 or so you mentioned. When impotred the sculpty imto Blender and exported it as collada, I got a mesh that was very nearly, but not exactly the same. On inspecting the wireframes, it was clear that the triangulation was different. The sculpty has all triangles made by diagonals in the same direction, while the Blender exporter one didn't. Neither did the mesh made when the triangul;ation was done in the uploader. When I edited the mesh making the same triangulation as used for the sculpty, the two objects were exactly identical, as expected. Different triangulation can make quite a big difference when many quads are skewed, as it typically the case in a sculpty landscape. In some circumstances that might make the difference between smooth and jagged.

Of course, the LI of the mesh is much bigger, but that is a bit unfair, as one would not generally use so many triangles when you are not forced to by the fixed sculpty mesh. Also, the mesh can have a precise surface you can walk on, something that would require a huge number of standard prims if you wanted to walk on your sculpty landscape. (maybe you can use a mesh as the physics shape for the sculpty, as long as you don't link them!).

 

 

 

[Toysoldier Thor]

No one is arguing that if you take LL's artificial $ Cost disadvantage away from the equation when comparing a MESH with a SCULPTY that the Mesh is by far the better of the two models.  In the high detailed physics, the much higher level of details that can be attained in one model, etc etc etc.  And... if it were not for LL's crippled LI cost factor that is only applied to the Mesh, I would be the happiest camper to be building much higher resolution and higher quality and more function mesh landscapes for all my SL customers then the Sculpties I offer them now.

BUT.... yes there is always a more powerful BUT...

Lets put this discussion in another light to get my point across screamingly clear for all those that are so in love with the pure technology superiority of Mesh over Sculpties....

REAL WORLD...  Lets talk about the LIGHT BULB

The incandescent light bulb (ILB) has been on the market since Edison (and others) invented it.  Thats a whole lotta years for a lot of other technological innovations to knock it off its seat.

The ILB is HUGE power hog.  for what it does - it does it so poorly.  Most of the power is used generating unwanted heat - not light.  It can also be a fire hazard and safety hazard.  And they dont last that long.  But.... IT IS CHEAP ! CHEAP CHEAP CHEAP.

So... the industry has come up with many newer innovations to reduce many of this ILB's huge weaknesses.  Lately, the industry has come out with the LED lightbulb.  I know - I am a rare customer that bought one.

The LED Lightbulb blows the ILB completely out of the water in almost every aspect of what a lightbulb should do.  It uses a micro fraction the power consumption of the ILB (mine 12Watt LED = the light of a 60watt ILB).  It generates negligable heat.  When turned on - the light brightness is instant.  I lasts 10 times longer than an ILB.  With new circuits in the LED lightbulb - it can now even dim using the ILB's dimmer technology.  And the LED Lightbulbs are 100% compatible to all ILB light sockets and circuits.

But guess which lightbulb continues to be the #1 seller right now?  With all these massive superior advantages that every tech geek and energy conservation nut in the world would 100% agree blows the ILB out of the water, the ILB massively continues to outsell the LED.....

WHY???

Because of one very simple yet overwhelming factor that trumps ALL the other factors...

PRICE !!

An ILB costs about $2 for a pack of 4.  I just bought 3 LED 12watt Lightbulbs for my kitchen...  $29/ each!  That was $90 for 3 lightbulbs!!

How fast do you think the LED lightbulbs will be flying off the shelf right now?

 

This is the world we are in now with MESH.  With all the advantages Mesh has over the Sculpty.... LL has crippled its quick adoption because of ONE factor.... COST!

you both can argue and debate and brag about the virtues of mesh on a technical basis and you will get no arguments from me on these facts.... but you both want to keep dismissing the most important factor that impacts the ability to sell and use mesh.... COST!

IF / WHEN LL decides to remove this crippling disadvantage on Mesh... then I will be the happiest camper in SL... why?  because then I will have an amazing technology that I could also make sales off of.

 

[Madeliefste Oh]

---

Toysoldier Thor wrote:

Thanks to those on this thread that gave me a lot of hints and clues and ideas on how to get the basics of bringing in my first MESH from my 3D tool. 

Since then I have been really strucggling to grasp the development pipelines.  Not only trying to learn how what and how to best reduce the topology on Voxel type 3D modeling tools like SCULPTRIS, or how to best create a proper UV Map.  I also put a lot of effort in trying 3D Coat which seemed like an amazing tool but I ran into some basic functions that didnt work.  Turns out I discovered that the TRIAL version of 3DCoat has a lot of these valued features not available so I had to abandon 3DCoat.  I am not sure why a company releases a TIMED TRIAL and then also removes leading edge functions from the trial.  Now I have to walk away from it.

Happily, yesterday I went through my first successful (sorta) end-to-end MESH creation, texturing, and import into SL with pretty good results.

Here is the result of what I created.

 Toy... nice you are working with mesh

You could improve the anatomy of your statue. The underlegs are

relatively

thin compared to the rest of the body,

and the size of the feet is also rather small.

[Toysoldier Thor]

Thanks and yes after rezzing it into SL I could see they look too small.  It was only a 4 hour sculpture and 100% completely out of my head of the human image - i was not using a photo to go by.

Plus, the critical purpose of the 2nd round of mesh creation was to figure out an effective Mesh Development pipeline.  I detest Blender (I used to use it in 2.48 days and eventually left it when an expert sculpter from IBM convinced me that Zbrush was far more friendly and easier to use - it wasnt by much).  I have since fallen in love with Sculptris and now I discovered moreso I love the freehand modeling of Voxel models (as 3D Coat refers to it as).  The way it works fits my way of creating models much easier than the more mathematical modeling tools like Blender and Zbrush.  Problem is that Sculptris is not exactly the most friendly tool to integrate easily to the SL model.  So I have been experimenting with 3DCoat and also trying t figure out which tools are best for what part of the pipeline to get a model into SL.

So far I think I am liking the flow of what I got now except for getting the texturing I got in the Sculptris to come into SL to the same resolution and image appearance that I see in the 3d Tool.  So its Sculptris, to zbrush, to sculptris, to blender to SL.

I am still playing around to improve the pipeline and the quality.  The easiest part ironically is the mesh shape itself !

 

[Medhue Simoni]

Being that mesh has all these unique properties, why would you not come at it from a different angle? Why would you compare 1 object to 1 object? Before I said the more complex the creation the better it is for mesh. That is false, and I meant that the more actual objects you add to the scene, or the more parts that make up something, the better it is for mesh. So, a mesh tree has a hard time competing, land impact wise, with a sculpty tree, but if we are talking about a christmas tree, with ornaments and presents below it, the mesh could easily be less prims, with more detail.

The only person that thinks this is a mesh vs sculpty thing, is you. I'm gonna use whatever is the best to use. Having said that, I will likely never make another sculpty. Why waste my time? Besides getting more prim savings with mesh, I can use mesh in any world, and any program, and it is the exact same. I could even sell it on IMVU, BM, Unity, and Dazstudio. There are probably more, and all the SL related worlds.

[Toysoldier Thor]

LOL

"I will use what ever is the right thing to use..."   as long as its mesh

Nuff said about your unbiasedness on Mesh.

You think everything situation is best solved with a mesh... this is where you are the only one to think this.

But i guess where cost is no option and love for a technology is so blind.... MESH can be the only solution to all.

Just like a screwdrive makes an awesome hammer and spear too.... right?

Anyway... unlike you ... I do think there is still for a long time... a valid place for Sculpty and Mesh.  I DO use the right tool and product for the right situation.

[Porky Gorky]

---

Medhue Simoni wrote:

The only person that thinks this is a mesh vs sculpty thing, is you. I'm gonna use whatever is the best to use. Having said that, I will likely never make another sculpty. Why waste my time? Besides getting more prim savings with mesh, I can use mesh in any world, and any program, and it is the exact same. I could even sell it on IMVU, BM, Unity, and Dazstudio. There are probably more, and all the SL related worlds.

---

Toy is not the only person who thinks this is a "mesh vs sculptly" thing. When creating content for commercial gain in SL the number one factor for allot of merchants is prim count.  Im now at a point where I have designed a large 100% mesh build in Max, it will fit snuggly on a 4096sqm plot. Once I import it in with "acceptable" LOD levels, the LI is outrageous when compared to a build of that size made with standard prims/sculpts. After spending weeks trying to optimize the model in Max to lower the LI, I am at the point now where I'm breaking the mesh model down into smaller segments, removing lots of parts with the intention of replacing them with sculpt and standard prims once the build is put together inworld.

After months of trial and error I have to agree with Toy, large scale mesh cannot compete commercially with sculpts and prims. Im sure the 100% mesh version of my build would sell to a few people, even taking into account the huge LI. However I need to justify spending 50 + hours on designing a prefab and I justify that by creating content that will appeal to as many people as possible and will sell over and over again. Prim count/LI has to be a major consideration with any products I release as that is a major factor that drives consumer spending. 

From a creation and artistic point of view Mesh is a godsend to SL and will be fantastic if LL manage to balance the cost of it correctly. However, my continued existence within SL is driven purely by profit, and in its current form, Mesh is only commercially viable for clothing, Av's and smallish objects IMO. I would consider any large 100% mesh build that I released right now to be niche product and would not expect it to sell as well as a hybrid mesh/sculpt/standard prim equivalent as it just cannot compete on prim count or LI or PE or whatever they are calling it this week.

[Drongle McMahon]

Toy, we agree about the advantage of sculpties in your particular area, and that it is a mistake. However we differ in that you think the mistake is usiong LI for mesh, while I think it is not applying it to sculpties. You say mesh are penalised; I say sculpties are subsidised.

There are endless argumernts that can be had about the details of the size relationship, making it better or worse, but I won't rehearse these here. But any approach to correlation with resource use requires a steep size dependence (until they introduce my user-settable LOD distance).  As mesh complexity is unlimited, without this we would be seeing mesh lag that would make sculpty megaprims look spritely. The cLI is having the intended effect of forcing builders to make efficient mesh. 

In the kind of application we are discussing, no wasted triangles and no sharp corners, and if the mesh has the same 4x LOD triangle reduction as the sculpty, the rendering load is identical and the download size is smaller for the sculpty. However, the mesh can be given more effective LOD and provides a walkable surface.

I used the 64x64x16 landscape I made to compare options. The LI of the mesh, using the high LOD for physics, was 84* (at that size, LOPD meshes have no effect). The LI of the sculpt was 1 (of course), but you can't walk on it. So I linked the sculpt to a mesh that used the high LOD for the physics shape, with an invisible box for all visible meshes to the sculpt set to physics type None. That had a LI of 42*. Using a 4x decimated mesh as physics in the same way, the LI was 8. Now the sculpty alone was impossible to walk on. The one using the high LOD physics was perfect to walk on. It would have taken hundreds of normal prims to get anywhere near it. The less complex physics was not too bad.

So perhaps there is a role for mesh even in your lanscape items. They can provide a correct optimised walkable surface much more cheaply than standard prims.

PS. I left the combo on Aditi:Mesh Sandbox 3. It will get autoreturned soon.

ETA: *these are about 9% and 4.5% of the capacity of the areas they occupy.

[Drongle McMahon]

For me, sculpty trees are an abomination (although I do use them!). If you are a traveller, they fill the landscape with huge ugly spheroids that rarely resolve before they are long out of view (I'm on a 2Mb/s connection). Nearly as bad as sculpty windowframes, most of which are excruciating to see before rezzing and at lower LODs. Among the greatest advantages of meshes are that they do not appear mangled before rezzing, and that their LODs can be made to behave acceptably.

[Medhue Simoni]

Here is where we would need to see your actual build in wireframe mode, so that we can all pick it apart and tell you exactly why your Land Impact is high. We would also need to know what your LODs are, and how your physics boxes are created.

Me, and I know many many others, have uploaded hundreds of meshes, testing this and testing that. I can quite confidently state, that if your Land Impact is high, you are most likely doing something wrong. If you or Toy want to post your builds, and explain the LODs and physics, I'm sure there will be dozens of us picking your builds apart. I have yet to see a valid complaint from anyone saying mesh Land Impact is high, outside of the sculpty tree to mesh tree example.

Please, post the models, cause I'd personally love to see the models get picked apart. Where is Chosen when we need him, lol.

 

@Drongle - I won't use sculpty trees. I have a neighbor that surrounds his tiny plot with sculpty trees and mountains, so much you can't see into his parcel. Whenever I get close enough to see his parcel, there is a long freeze as they all rez. It pisses me off to no end. To each his own tho. I keep my region running very smooth, with 3 major commercials stores on the sim, 1 being a script store. My combat testing area is just at the right spot where your pc renders my neighbors trees and stuff. So, everyplace on my sim is designed to render everything reasonably, except that 1 corner of my sim. Inside my store, of course there are many textures to render for signs, but all the geometry is simple.

Someday, I will make my whole store mesh, that would totally rock, and I bet I'd save 20 or more prims with more detail. Maybe not, cause I'd need a decently complex physics model, but I bet it is still under 100 prims. Currently, just my store is about 75 prims on a 16000 meter parcel. Lazy made a mesh Mayan Pyramid for about 100 prims for a 4096 parcel. Mine would need to be larger but I don't need all that detail. Even if mine was 150 prims, it would still be worth it. Plus, if I take all my sculpty torches, that light up the store, make them all mesh, and link them to my pyramid, I'll save more than a dozen prims there.

[Toysoldier Thor]

Drongle,

As I said in the other thread.....

A rose by any other name is still a rose.  You did make me laugh though at your statement that you dont see it as a PENALTY on Mesh as much as it is a SUBSIDY on Sculpties.   Basicly all you are doing is trying to paint a pretty coat of paint on a poor LL decision that you tend to agree on.

At the end of the day ... this is true in RL and SL economics.... when one side of competing anything (like technologies) is awarded a benefit the other does not have... its a penalty to the side that doesnt have it.  A Subsidy to one technology is a Penalty to the other technology.  But if you think that is fair... then lets play a game here....

Lets say You and I are direct competitors in the SL marketplace - we both sell PRIM HOMES (no sculpty no mesh - just prim).  There is no technical advantage between us at all and we both want to see out sales grow and out products adopted. 

BUT.... LL has decided to make a "fair" pricing policy.  They decided to implement a new rule that any Merchant whos name starts with the letter "D" will now be charged a Marketplace commission of 1% for every 5 prims in your product.  And any products created from Merchants started in "D" will have an additional LI charge on their product of 10% over the actual prim count.   It doesnt matter the logic why LL decided upon this legislated policy - its the policy and it was deployed.

So Drongle,  I am sure you must hate that you and all your products are being PENALIZED just because your merchant name starts with a "D".   I tend to disagree with you though Drongle and I think LL's policies are fair.  Come on Drongle, dont think of it as a Penalty against you..... you are not being reasonable... its only a subsidy for all us Merchants with names not starting in "D"....   RIGHT??

I know you and Medhue and a lot of other Mesh Lovers that likely were tightly involved in the closed beta with LL and were possibly part of helping LL come up with the load-based policy will never see what LL did with LI as being a poor decision.  To you it will always be thought of as "the right thing that should always have been done long ago".  It was a smart Academic decision but it surely was not a wise political nor economic deicision.

It is and will always be a penalty on Mesh that no other prim type will EVER have to be faced with.

[Toysoldier Thor]

ROFL What is so ironic about how Mesh lovers completely wave the flag at how Mesh is the godsend of SL and how its amazingly more efficient on the grid than sculpties......

and yet these same Mesh CheerLeaders fully support LL's decision to penalize Mesh on upload and LI costs that we all know LL will never ever impose on any of the completely hideous sim resouce consuming sculpties.

So if Mesh is such an amazing technology (which I totally agree it is) then why would LL impose an ariticial economic penalty on a technology that if quickly adopted onto the grid - could displace all those EVIL SCULPTIES that Medhue and others hate so much?   Please Medhue - dont post now telling us all that you dont hate sculpties with a passion - your past posts and Drongles too in this thread makes your opinions on Sculpties clear.

 

If I were LL and I wanted to displace an old resource hogging / sim lagging technology with a new much more efficient and more flexible mesh technology... the last thing on my mind would be to impose an economic penalty on this technology.

In fact I would go the other way and start imposing a gradual penalty for rezzing a sculpty over the period of 2 years.  ie. as of April 2012, all rezzed sculpties will have a LI=2 not 1.  On January 2013 the Sculpties will have an LI=4.

Then I would let Mesh not have a penalty - its cost would be LI=1.

Is this academcically wrong?  YES.  But the world in SL is not perfect as it is not in RL.  Prims put loads on a sim.  Lag exists.  But LL should be doing everything to promote the technology that best reduces that load - not penalize it.

 

[Toysoldier Thor]

---

Drongle McMahon wrote:

For me, sculpty trees are an abomination (although I do use them!). If you are a traveller, they fill the landscape with huge ugly spheroids that rarely resolve before they are long out of view (I'm on a 2Mb/s connection). Nearly as bad as sculpty windowframes, most of which are excruciating to see before rezzing and at lower LODs. Among the greatest advantages of meshes are that they do not appear mangled before rezzing, and that their LODs can be made tobehave acceptably.

---

You two make me laugh.  LOL what kind of viewer do you use Drongle and Medhue???  You two make it sound like prior to Mesh showing up... SL was a horrid unusable grid to live on.  WOW... maybe you two should consider upgrading your PC's to a computer that is newer than 10 years old.

Take a look at these two Flickr photos that I created while walking around the amazingly beautiful Alirium sim.  This is a sim completely full with more Sculpty trees and rocks and scripted grass and fields etc....  To you two... sims like Alirium would never rez as they are completely poluted with evil Sculpties....

Isnt it strange how not only could I TP and rez on this sim and walk around seeing all the enormous beauty that these sculpties offer my viewer, but when I stand there I take SL photos at resolutions of 5000 x 2500. 

http://www.flickr.com/photos/11535545@N00/4988325664/in/photostream

http://www.flickr.com/photos/11535545@N00/4968289924/in/photostream

Seems only a MESH CHEERLEADER believes that the world of SL prior to mesh was a disaster of unworkable sims.

Come on guys... be a little bit unbiased for people to believe you

I am so thankful for Sculpties and they made SL a far far better place to exist in than before they showed up.  Will mesh offer even better beauty in SL - ABSOLUTELY.... but you two are being completely unfair to slap sculpties as unfairly as you do.

But again, I should criticize other SL Residents if their cannot afford modern PCs and graphics cards.