Jump to content

Sculpt prims, can I get a professional opinion? (Never mind)


You are about to reply to a thread that has been inactive for 4439 days.

Please take a moment to consider if this thread is worth bumping.

Recommended Posts


Shotgun Galtier wrote:

Can I get a Professional opinion about this? I am trying to make a fedora sculpt prim but I can't seem to get it right and I wanted to know if this would work for the sculpt before I uploaded it onto S.L

.

Any advice would be helpful ^^ oh and I am using Blender 2.61 for this so if this wouldn't work could someone please direct me to a good blender 2.61 tutorial that teaches you how to make the correct photo format.

Edit the link of your post, is uncorrect. Perhaps this: http://i442.photobucket.com/albums/qq141/sexy-shotgun/untitled.png

Link to comment
Share on other sites

From the texture image you posted (you still need to fix the link, by the way), it appears you simply created an arbitrary mesh model of, and UV'ed it in the typical fashion for such a mesh.  The problem is sculpties are not arbitary, and their UV layout is not typical.  They have to be made in a very specific way.

Sculpties are like origami.  Everything is ulimately unfoldable into a flat rectangle.  The UV layout must be a perfect grid.  The topology must be a perfect, singular surface.  It's analogous to wrapping a gift.  You fold the paper in 3D space, to take on the apparent shape of the box, but at the end of the day, it's still the same old flat piece of paper it started out as.

When making sculpties, you can't do anything to disrupt the surface topology, or the UV layout.  That means you can't use most of the tools you're probably used to using for making efficient, sensible mesh models.  You can't extrude, you can't perform Boolean operations, you can't add or delete vertices, you can't delete faces, etc.  No matter how you bend, twist, stretch, fold, or otherwise distort that perfect rectangle in 3D space, it must remain the exact same rectangle in 2D.

Since you're a Blender user, I would recommend you follow the Machinimatrix Blender Trail tutorials.  The very first one happens to use the example of a top hat.  Needless to say, that's just a few tweaks away from a fedora.  I'm sure that after you follow the tutorial series, you'll have no further troubles. (No need to 'reward' me with a copy of your sculpt map, by the way.  If I ever want a fedora hat, I can make one myself, in about five minutes.)

 

That having been said, I do have to ask why you want to make this hat as a sculpty at all.  SL does have arbitary mesh support now.  You could make the same hat far more efficiently, far more easily, and in a fraction of the time, as an arbitary polygonal model than you ever could as a sculpty.

Link to comment
Share on other sites


Shotgun Galtier wrote:

Thats all good and all, but for one; I am not paying for primstar-2, and for two; i just checked the object creation on second life, it is still the same old thing as it was yesterday, I don't want a cluster of prims in just one object

You can use Blender for making mesh. If you can make a sculpty, you can make mesh aswell. A mesh can be a cluster of prims or one single object, that just depends on how you build it.

I'm not a Blender user, but if you ask in the mesh forums, people will happily point you to some tutorials.

Here's one for making sculpties though, not using Primstar:

Sculpty in Blender 2.5

Link to comment
Share on other sites

I'm still sculpting rocks and such with sculptypaint/lunapic/irfanview. Because of the amount of curvature, there is basically no wasted shape data on these sculpts, unless you consider there to be wasted curvature as well.

A also recently got the prim generator, with which I can say I am very pleased, in terms of its ability to quickly produce sculpts of more geometric objects. I should mention, though, that if you don't max out the vertices count, you'll get a gray space at the end of the sculpt map. It doesn't hurt the sculpt, but it is a waste of data by some measures. Going to half the maximum vert count (for conceptually simpler shapes) seems to avoid the gray area while producing a more resolved over-all shape. My octagonal grate, for example, is octagonal because it is cooked down from 32 vert count, whereas my circular grate is more arguably circular only because it is cooked down from a 32 vert count; the generator prims used are actually of the same shapes. 

I do not dismiss mesh, but the viewers that show it continue to ruthie my avatar, and I still can't find a 3-button mouse I need to start using blender (yes, I can get a 5-button wireless mouse here, but the price!).

The prim generator seems to me an easy way to set limits for the complexity of projects because it's a matter of making them as complex as I can within the set data limit. I suppose there's no reason I can't later import the data to belender or other applications to cut out some tiny number of wasted triangles for a mesh version, but what's the hurry? Mesh versions of such things will be charged at more then one prim for the sizes and LOD at which they are likely to be used anyway. 

Link to comment
Share on other sites


Josh Susanto wrote:

Mesh versions of such things will be charged at more then one prim for the sizes and LOD at which they are likely to be used anyway. 

Hmm, not going to restart the discussion on this:)

But since you mention grates, I can guarantuee you a mesh grate at any usable size will probably be just one prim aswell, with a lot less geometry and a lot better behaviour on LoD.

Link to comment
Share on other sites


Shotgun Galtier wrote:

I don't want a cluster of prims in just one object.

I did not suggest making a cluster of prims.  I suggested making a mesh model.  I'm not sure where you got "cluster of prims" from that.

If you're trying to say that you already uploaded your mesh fedora, and that its land impact came out higher than you'd like, then I'd submit you just didn't make the model very efficiently.  A fedora is a pretty simple item.  It doesn't require more than a couple hundred polygons at most for the highest LOD, and just a fraction of that for lower ones.  If the land impact on any well made mesh fedora were to come out any higher than 0.5, I'd be extremely surprised. 

It's worth noting that for wearable items, land impact isn't really what you should be concerned with anyway.  For wearables, the display weight is what's most important.  The display weight for just about about any well-made mesh model will almost always be considerably lower than any comparably shaped sculpty model.

If you need tips on how to make your model more efficient, by all means ask for help over in the mesh forum.  There are any number of us over there who would be more than happy to offer some guidance.

Link to comment
Share on other sites

>But since you mention grates, I can guarantuee you a mesh grate at any usable size will probably be just one prim aswell, with a lot less geometry and a lot better behaviour on LoD.

My grates are the same number of triangles as the other sculpts I earlier began to load through blender and into SL's mesh uploads process just to see what happens, so they would cost 8 times as much to load unless I could cut out some number of triangles, but that would be a lot fewer than half of them. That is: way more than enough triangles, still to add up to one prim, if it's mesh.

The difference in physical behavior is essentially the difference between a flattened cylinder and flattened sphere. At 1/8 the loading cost and a lower land impact, I can live with that for a while at least.

Of course, there's nothing to prevent my customers from exporting the data and making a mesh version. It would be a lot faster and easier than starting from scratch, and I welcome them to do it. 

 


Link to comment
Share on other sites

>I did not suggest making a cluster of prims.  I suggested making a mesh model.  I'm not sure where you got "cluster of prims" from that.

I probably confused him with my broader discussion of sculpts

Since none of his links work, it has been difficult for me to understand what he's trying to do.

How about this?...

1) I sculpt a fedora and post the map here for free.

2) Chosen or Kwak loads it into some piece of software that costs more than a car, uses that to suck out the extra triangles, and charges whatever they like for the resultant mesh. 

Link to comment
Share on other sites


Josh Susanto wrote:

My grates are the same number of triangles as the other sculpts I earlier began to load through blender and into SL's mesh uploads process just to see what happens, so they would cost 8 times as much to load unless I could cut out some number of triangles, but that would be a lot fewer than half of them.


Yes if you upload your sculpt as a mesh and don't bother to make custom LoDs, you'll get a high landimpact. Something virtually flat, like your grates, can use imposters or billboards for probably Lod medium to lowest, which means 4 triangles. A while back someone posted a similair object, a near flat billboard or scaffolding thing. It was one sculpt and the creator was sure that was impossible with mesh. A lot of people dove into it with various techniques and the landimpact was indeed 1 for mesh. I got it low enough to be able to link two together for a single prim even. For that grate I'm 100% sure it will be the same. You'll lose the shiny on the lower  LoDs, but that won't be very noticable at a distance.

EDIT... Yes both maya and max will cost as much as a nice second hander, but you can use a free 3D tool to make these meshes aswell.

Link to comment
Share on other sites

My grates are not flat. But they are shiny, and they stay shiny at any size. 

if I were selling flat plane alphas, that would pretty clearly be a waste of data when compared to mesh. 

If an alpha grate is just going to be recessed into a floor, I suppose you might as well just put it on a larger single mesh triangle.

No?

 

Link to comment
Share on other sites

"...so they would cost 8 times as much to load..."

This is 576 triangles. Inspired by a sculpty grate in your marketplace store*. Made in Blender (if you can get a used car for the price of Blender, I don't want to ride in it!). It stays at 1 LI until it gets bigger than 2.1 m diameter. It uses the same mesh for first two LODs, so there is no distortion. The lowest LOD is two octagonal planes with an alpha texture with a picture of the grate. This is used for the physics too. If you make the lowest two LODs like this, it is one LI up to 4.2 m diameter.

 octograte2.jpg

I think your sculpty work is outstanding. Your methods are uniquely creative and your products have artistic qualities that I could not aspire to. You have nothing to worry about. People will always want to buy good stuff.  I do think you could make it even better with mesh, but that is your choice. I have seen many pure prim builds that are far better than anything I can make wiuth mesh. Technique is secondary to artistic ability when it comes to making good content.

*No need to use you sculpt map. Don't worry. It only appeared for the photo on Aditi. I am not going to sell it.

Link to comment
Share on other sites

Oh are we going to play the "we can build you a mesh that is better than your sculpt " game again? hehe

Well anyway that is what I had in mind.. well I think Josh's sculpt has 288 visible faces, all squares as cross section.

Oh and I drive a car that cost me as much as Blender actually and there are no issues with it whatsoever, it's a 30something year old well maintained classic!

@josh..I said virtually flat, not flat, what drongle built is what I ment. One could play around with the LoDs maybe, the geometry is pretty much as it is...

Link to comment
Share on other sites

My Blender car is in top notch condition, passed the yearly test two weeks ago with zero fails... And actually your mesh is better since it has less visual load and less physics load and going by your numbers more shape on the cross section.

I fully agree on the comment you made about Josh's sculpts btw..and building in general...it's not the tool or method, it's the creator that makes the biggest impact on how good it is.

Link to comment
Share on other sites

Ok, my two cents on the mesh version of Josh's grate....

Grate.jpg

The profile of the grate is different, but the shape as close as I could get it to Josh's. I think we can all agree all grates like this look just as good, so it's the rest we need to look at. I uploaded this one with full geometry for LoD high and medium and a 4 triangle imposter for LoD low and lowest. As the picture shows it will have a LI of 1 up to 5,5 meters. Physical shape on an object like this can be convex hull.

If you want bigger sizes for an impact of 1, set LoD medium to the imposter aswell. At the bigger sizes it won't show very soon. With that setup the size for a LI of 1 can be 21 meters. Bigger than that should also be possible, although then you'd have to split the object into two halves.

The only disadvantage of a mesh over a sculpt with these objects is the mesh will need different objects for different sizes. The sculpt will be worth a single prim no matter what.

Link to comment
Share on other sites


Josh Susanto wrote:

I probably confused him with my broader discussion of sculpts


That may very well be, since I too am confused as to why you interjected (advertised?) mention of your of rocks and gates at all into this discussion about a hat.  Whatever your reasoning was, however, that particular confusion could not possibly have had anything to do with the "cluster of prims" thing.   That was mentioned a full three posts BEFORE your first appearance in this thread.  Not sure why you thought otherwise.

 

 


Josh Susanto wrote:

How about this?...

1) I sculpt a fedora and post the map here for free.

2) Chosen or Kwak loads it into some piece of software that costs more than a car, uses that to suck out the extra triangles, and charges whatever they like for the resultant mesh.


I'm not going to speak for anyone else, but for my part I'll say no thanks, Josh.  If I want a fedora hat, I'll make one from scratch, and it'll take me five minutes to do.  That's a heck of a lot easier than going in and reducing a nother model after the fact, whether the other model be a sculpty or anything else.

I thought we had gotten you over this limited thinking of yours, this stubborn instence upon sculpties as the foundation for everything, months ago.  It's disappointing to see you've reverted back to your previous unenlightend state in this regard.

 

 

 

 

 


Josh Susanto wrote:

II'm still sculpting rocks and such with sculptypaint/lunapic/irfanview. Because of the amount of curvature, there is basically no wasted shape data on these sculpts, unless you consider there to be wasted curvature as well.


Uh, thanks for the announcement?  That's about all that comes to mind to say, since I'm not sure what your comment has to do with the topic of this thread.

 


Josh Susanto wrote:

A also recently got the prim generator, with which I can say I am very pleased, in terms of its ability to quickly produce sculpts of more geometric objects. I should mention, though, that if you don't max out the vertices count, you'll get a gray space at the end of the sculpt map. It doesn't hurt the sculpt, but it is a waste of data by some measures. Going to half the maximum vert count (for conceptually simpler shapes) seems to avoid the gray area while producing a more resolved over-all shape. My octagonal grate, for example, is octagonal because it is cooked down from 32 vert count, whereas my circular grate is more arguably circular only because it is cooked down from a 32 vert count; the generator prims used are actually of the same shapes.


Again, not sure what this has to do with this thread.  Perhaps you should have made this its own discussion, Josh.

 


Josh Susanto wrote:

I do not dismiss mesh, but the viewers that show it continue to ruthie my avatar,

Really?  That strikes me as strange, considering I haven't heard anyone else complain about that particilar problem in months.  When was the last time you tried, and which viewer(s) was it?

 


Josh Susanto wrote:

and I still can't find a 3-button mouse I need to start using blender (yes, I can get a 5-button wireless mouse here, but the price!).

Again, I thought you had gotten past that, months ago.

First, a 3-button mouse can be bought for as little as $4 on Amazon.  I realize economics are somewhat different in your country, but would four dollars really break the bank?

Second, I did offer to send you one, if you recall.  I've got a closet full of old keyboards and mice.  Remember we talked about that?

Third, even if if every three-button mouse in the world were to disappear tomorrow, it wouldn't matter.  You can use Blender without one.  Did you miss the fact that there is an "Emulate 3 Button Mouse" option right in your user preferences?  Turn that on, and then you can just use alt+LMB in place of all MMB functions.  Also, depending on what mouse driver you've got, you can probably enable MMB emulation globally, via your mouse settings in your Windows Control Panel.

Somehow I suspect mouse availability has little if anything to do with your lack of progress since last we spoke.  If you had been serious about wanting to make it work, you would have.

 


Josh Susanto wrote:

The prim generator seems to me an easy way to set limits for the complexity of projects because it's a matter of making them as complex as I can within the set data limit. I suppose there's no reason I can't later import the data to belender or other applications to cut out some tiny number of wasted triangles for a mesh version, but what's the hurry? Mesh versions of such things will be charged at more then one prim for the sizes and LOD at which they are likely to be used anyway.

Sigh.

We talked about this, ad nauseum for weeks on end, a few months back, Josh.  Did none of that discussion sink in? 

First, do you not recall that after having come to understand how unfairly under-charged sculpties are, in terms of land impact, you emphatically stated that you had a problem with that in principle, and that you were no longer planning to argue in favor of taking advantage of that loophole?

Second, if land impact is now once again your only concern, after all that discussion about all the other supremely important factors that make for a good user experience in any 3D simulation, that's incredibly unfortunate.  I really thought we had gotten through to you.

Thrid, as others have stated, there's no reason a simple item like a rock or a grate needs to have a land impact greater than one (unless it's quite large, in which case it DESERVES to be more than one).

 

Link to comment
Share on other sites

I appreciate your efforts not to rub me the wrong way on this.

Your analysis of the grate makes perfect sense to me.

Most of my items, I make as 10m on one axis, partly because they're easier to photograph that way, so I tend to think of everything being potentially used at very large sizes. 

I recognize that some items, people will mostly want smaller. If that's the case and they can get the LI down in the process, there's no reason I can think of not to use mesh instead (except that one might want everyone to be able to see it without getting ruthied, which I know is a weird stupid problem that should not matter, although, in fact, it does).

My recent sculpt work is just incredibly stupidly easy to produce, actually. Not to insult anyone, but people who are impressed with it probably haven't tried the tool I'm using in order to produce it. 

If someone would make such a machine that spits out meshes, I'd already be using that.

 

Link to comment
Share on other sites

You are about to reply to a thread that has been inactive for 4439 days.

Please take a moment to consider if this thread is worth bumping.

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share

×
×
  • Create New...