I’m having an issue that’s been truly annoying. I’ve searched elsewhere and there doesn’t seem to be a solution, but I’m hoping that there is one known here.
I imported my model, which has many tools and subtools to those tools. All was well with the scale and symmetry. I think because of working on one, or more, tools with the transpose move/scale/rotate options, it seems to have affected the entire project by first rescaling it super small compared to what it was imported as, and also by offsetting the symmetry on the models by just a few millimeters (in relative terms). I’ve looked at the one video out there about this happening and resting symmetry using smart resym and such, but none of that can help this model, which is already detailed (and details must be preserved on both sides – some not symmetrical, some that are). The only option I’ve been reluctantly using is setting pivot points with local symmetry sometimes on, sometimes not (sometimes it helps, sometimes it doesn’t) on the individual tools and working on them, then clearing the pivot point so it snaps back to position on the model to check with its integration with the other tools. That’s been working for a while.
However, now that I’m trying to polypaint, I have to merge the head and upper torso together, because the character’s skin is bare and so I need to paint continuity. Doing so, however, means now the offset symmetry will not reset with pivot points, so I can only manually reset the position back to 0 with a pivot point active, then would have to retype the position it was before snapping it back to the model to see if ti works … every single time…
I don’t, for the life of me, understand why there is not a global reset on positions for things. Under geometry, I thought that I found it. Position. However, it’s completely different for every subtool you have, so you can’t reset anything there, and sometimes changing one value on one axis resets the other two.
My second problem is also decimating the model. I’ve used Decimation Master, because the parts are large in polygon counts coming in, before I merged the objects. it’s preserved all the crisp details visually at 52%, so everything looks fine. However, I notice for some reason the masking options no longer are working right. Mask by Cavity doesn’t, just masks the entire object. Anyone know what happened there, or how I can fix it?
Size and position are stored on the model, but shouldn’t really be causing you any problems. What were you importing them from/to? How were you doing it?
Mask by cavity needs at least 2 subD levels to work. It actually reads the lowest one to help it understand where a cavity is. After you decimated the model, you only have 1 subD level remaining.
Also, you shouldn’t be painting on a decimated mesh.
Any images you can provide to help us help you would be a great start as well.
Well, the size and position were fine on import. Somewhere along the line after only doing two things did things get messed up. One, I posed one subtool, the eyelids, rotating them and reshaping them to fit the new eyeballs I made; two, I replaced one part of the clothing, a medallion, with a remodeled one using a Zsphere. After that, the position and scale of the entire bunch of tools changed. I’ve read that this is something that can happen, but in a thread on Polycount a couple of years ago. I imported the model parts from regular objs.
Anyway, here are some pictures to illustrate things better (modeling an accurate Kain from the Legacy of Kain games):
First brought in with symmetry and position fine. The red dot is right in the center, like it should be:
The head and body parts append right to where they should be and line up perfectly:
This is the position of the head, which is the focal point the other tools are going off of (even though I first brought in another tool into the scene:
Now, after the minor changes, symmetry on all tools is off by just a bit (enough to make a huge difference):
And all tools still line up, so it’s a global change:
Also, cavity mask was working, even though the model parts brought in had no subdivisions since it was modeled in another program (which raised the polygon count far too high and why it’s split up in the first place, argh…)
beta_channel, I don’t know that subtool master can help. I’m not having an issue filling the objects with solid color, but with having them connected to paint detail naturally across the surface. Like, a line, skin shading variance, etc., all of that can’t end in visible seems. From what I saw, anyway, subtool master can only fill the objects with the color selected, yes?
So. How are you importing objects?
you said you’re just importing obj files. Are you appending them, multi-append. Overwriting a previous mesh, etc? Why are you importing and exporting objects over and over again (I am assuming you are).
Your position on your mesh shouldn’t be changing.
Is there a reason you’re not just using HD geo?
Why did you have to break your mesh apart?
What does your wireframe look like?
You say keeping your mesh together is what you want to do, so why not do that?
First, let me preface this as I should have above by saying that I am new to these programs and this is the first model I’ve ever done. When I asked my question above about Subtool Master and its abilities, it’s because I was sincerely asking, not to be smart.
I imported the objs, cloned them, then recompiled them in the subtool menu by appending them. Then deleted the cloned tools. I’m not importing and exporting ones over and over again. Once it was apparent symmetry was screwed up, I did import one of them again to check if it came in that way, then saw how the scale and symmetry plane had been completely changed somehow. I then checked an earlier save and the symmetry, position, and scale were fine. The original obj now when it’s appended to the rest is about 6 times larger and way off center coming in, so again, something globally was changed by Zbrush. I have not found any method of doing this internally that would rescale and offset all subtools in the exact same manner (if there were such a way, I’d be able to reverse this, but it looks like there isn’t…). Something was altered by the program.
Your position on your mesh shouldn’t be changing.
I agree. I wish it hadn’t.
Is there a reason you’re not just using HD geo?
Why did you have to break your mesh apart?
What does your wireframe look like?
As I said, I modeled this in another program, which was 3D-Coat. Let me just relay the abridged tale of that. When going into its new surface mode sculpt anything way with their newer live clay brushes, it’s a bit buggy for a lot of users, apparently. Holes, tears, explosions can happen, especially during fine detailing later. It doesn’t have subdivision levels the way Zbrush has, so it seems to just keep increasing the tessellation in whatever area you’re detailing. The polygon count on the model rose through the roof when doing fine detailing. Actually, this is the second attempt, since the first became just unusable. I was encouraged there to try again doing certain steps to shore up the model in Surface mode. It didn’t work, and here I am again. The model is much better, but still plagued with the huge polygon count that makes it very difficult to work with even in pieces in Zbrush. This is why I had to break the model up just to bring it into Zbrush. The head is over 10 million. The body is 18 million. Upper arms, 7; crown piece 5. More crown pieces are more millions, not to mention the clothing pieces… In 3D-Coat, the polys were even higher. Intact, it would be far too high to import into Zbrush, which has capped subtools at 21 million, it says, but I have issues at anything over 15, really. Also, I can’t use HD Geometry until there are sublevels, which these models will not have from 3D-Coat. Even if they did, it says the model has triangles, so it won’t do it anyway. Plus, I don’t want to further detail. The detailing, minus any last minute etching, or so, is done. I just want to paint it (eventually make a low UV low poly also for animating) and my license with 3D-Coat is educational, which only would allow for 2K maps.
You say keeping your mesh together is what you want to do, so why not do that?
Well, Zbrush can’t without crashing and telling me it’s out of Ram. If it were a 64 bit-compatible program, there probably wouldn’t be a problem. I have 12 GB of ram and a good enough PC to handle it, but since that doesn’t mater to Zbrush, here I am. I could merge them together once decimated to 50%, but it’s still very laggy, and now the problem of the cavity mask. What I’ve tried for the past two days – all day and night – is to try some working formula to get some sublevs into the mesh. The pieces I cloned, then decimated down to 15%, subdivided once and projected the high poly detail onto them. That’s what I eventually was able to do without the system crashing every time, and to a level the details are about 98% preserved, and I have sublevels. However, I cannot still merge all pieces. I get the upper body merged, but the head trying to merge that down and preserve the sublevels crashes Zbrush. Merging without trying to preserve the sublevels works without a hitch. I’d lose the ability to use cavity mask on it again, though.
Basically, I’m working around the problem that I had the misfortune of creating a super high polycount model in another program and Zbrush is tolerating it, but it’s not happy about it.
I just want to paint it, and end the very, very aggravating and stressful time I’ve had with this all (3 solid months of constant problems – I’ve never done anything in my life with this many constant issues). I want to do it right, though. Painting on parts here would be very hard to match up. If it’s what I have to do, then it’s what I have to do…
It sounds to me like your workflow is a bit off.
Your model’s head should not need 10 million points to make Kain.
Basically, I’m at a loss for how dense your mesh is. A complete human with pore level detail should really only be coming in around 10-15 mil for the entire body. So Kain’s head being 10 million is blowing my mind. Why do you feel your mesh needs to be this dense?
You are correct that if Zbrush was 64bit it could access all of that RAM, but it sounds like your workflow wouldn’t really help it any.
Duplicate your mesh, then use Zremesher to lower your polygon counts.
subD up your mesh and project your new mesh onto your old one.
It looks to me like you’re using surface noise and actually applying it to your mesh. You shouldn’t really need to do this as the amount of displacement that you’re using could easily be achieved with a bump map.
What is the end goal of your project? Is this a render, a sculpture (3D print), a game character, a character for a cinematic?
How do you plan to get your maps out of Zbrush even if you could paint it? Is the model unwrapped, you decimated it, so I’m going to go with no. Very few programs could handle the model density you have, so you can’t just use vertex color information.
Are you planning to use normal maps, displacement maps, or a combination of the two?
3D-coat can paint over multiple meshes at the same time, so you could just use 3D-coat to paint the model, assuming you can get it low enough to actually unwrap.
If I sound like I am rambling, I probably am. If you could break down your end goal, and your plan to achieve it that would probably go a long way to illuminating the solutions to the problems you are facing.
I don’t feel that it needs to be this high. I never wanted it to get this high ever. My goal was only 10 million, or so, for the full body.
As I noted, currently the brushes in 3D-Coat used to detail are causing soaring tessellation for some people, as well as other problems like holes, tears, etc. which can threaten to ruin the mesh.
To your point below, I didn’t use surface noise in Zbrush for any of this. In 3D-Coat, I had the entire head and body built up to about 5 million total polygons. Just using the crease brush (the equivalent of the Dam Standard here) to create the cracks in Kain’s skin was the biggest increase of the tessellation. Instead of being able to really subdivide (I was advised the subdivide brush isn’t so good, so to do it this way) in that program, instead the brushes simply increase tessellation right where you’re working – kind of like the spotlight feature in HD Geometry, for something to compare it to. However, it’s far higher than it should be. Using that and some custom alphas for the flaky metallic skin look I hand sculpted into his skin – the equivalent of his pores – this rose and rose the polygon count. I had two choices, which were to stop before it got higher and bring it over to Zbrush for detailing, or to continue and then chop it up at the end and bring it over that way, then fix it here. I was encouraged to try and detail it in 3D-Coat, so I did, and it wound up creating a mesh this high and one that I wish wasn’t. However, now I’m very much trying to decrease in the polygon count and to merge back together properly without losing all of that precious detail it took me so long to sculpt in.
Duplicate your mesh, then use Zremesher to lower your polygon counts.
subD up your mesh and project your new mesh onto your old one.
That’s what I have been trying. Like I mentioned, I’ve been working on the best way to do it for days. Remesh in the subtool menu just creates an unusable blob that doesn’t retain the mouth cavity, so gets nasty when trying to project detail onto. Decimation alone won’t work. I’ve used it first on one piece and then Zremesher.
Zremesher seems to be having more of an issue due to the uneven topology and that there are so many triangles it has to deal with. I’ve been working on the best blend of guides on the mesh and “use polypaint” to tell it where to keep the most detail. I’ve finally got the head as really good as it probably will get down to 9,00 polygons and subdivided up. I had to make sure some guides were inside the mouth cavity, because it wasn’t shaping the mouth right still. There were little explosions right in the down-turned corners of the mouth. Now, it’s fixed. I’ve gotten the body torso piece mostly where it should be (12,000 subdivided up). If I can get the other pieces done today, then I can try to do something with them.
The pieces are subdivided back up to less than, but close to the polygon count they were in order to get the detail back onto them – so subdivided about 5-6 times. After that, I think I will have to do another round of decimation and then only one subdivide level more to make the pieces half, or less, the count that they are – but I will keep the low poly zremeshed tools saved for making initial uvs from.
And of course, I will never, ever work this way again. I can build and block models easily enough in 3D-Coat, but if I do, I’ll export them and bring them in to Zbrush when at that lower poly, less than a million stage, remesh, then work from there on in Zbrush. This has been a nightmare to fix.
It looks to me like you’re using surface noise and actually applying it to your mesh. You shouldn’t really need to do this as the amount of displacement that you’re using could easily be achieved with a bump map.
What is the end goal of your project? Is this a render, a sculpture (3D print), a game character, a character for a cinematic?
I wanted the details to be as truly sculpted in as possible, to create displacement and normal maps from and also just to have preserved the highest model possible. To answer is this for a render, sculpture, game character, or cinematic – all of the above, basically. I want to use it for renders and for my portfolio. I also want to use it for a game character and cinematics for that, which will be the next hurdle after figuring out the proper work flow for modeling so this never happens again. My ambitions are probably exceeding reason, but I’m going to try and learn how to use the UDK creation kit to make a recreation of the entire world universe of LoK, from Blood Omen 1 on – but starting with BO1. I’ve already taken this elder Kain model, when back in 3D-Coat, and blocked out from it the original human Kain and armor for him. So, I’ll be trying to do all stages of him, from human, to fledgling vampire (and all armor and outfit sets), all the way to this.
How do you plan to get your maps out of Zbrush even if you could paint it? Is the model unwrapped, you decimated it, so I’m going to go with no. Very few programs could handle the model density you have, so you can’t just use vertex color information.
Are you planning to use normal maps, displacement maps, or a combination of the two?
Combination of the two, and after I merge the pieces back together on the body is when I planned on zremeshing again to use the actual low poly for unwrapping. I’m aware this thing would be too high for anything else to handle as is. I was going to get it down to reasonable less than half size, like I mention above in this post, and that and the low poly created from it I think should be workable for Xnormal to handle to create the other maps from.
At least, that’s the plan.
3D-coat can paint over multiple meshes at the same time, so you could just use 3D-coat to paint the model, assuming you can get it low enough to actually unwrap.
Yes, 3D-Coat could handle the texturing. Before sculpting, I have used it to texture diffuse maps for LoK: Defiance characters, which I made texture packs for people to update that 2003 game using Texmod. 3D-Coat is great for texturing, so it’s a shame I can’t use it now. As I mention, I can’t use it to create anything higher than 2k maps, though (because it’s an Educational license and I don’t have the money for an upgrade), so this is why I have opted to use Zbrush for texure painting. I’d have to probably use Zbrush for Zremesher anyway, since I didn’t like the result of the autopo remesher in 3D-Coat.
edit: your Kain looks really nice, btw.
Thanks When I get this figured out, I can show the entire model to see all the other parts in a WIP thread, like it gauntlets, cape, and greeves.
first let me say AMAZING WORK GUYS!!! ok my question is when inserting eyes into a mesh how to i fit them in correcty im having a problem moving tools i want to move 1 tool and they all seem to move at the same time?
Well. The growing pains of learning. If nothing else, you’ve learned what not to do.
A remake of the entire world of blood omen is quiet the under taking to be sure.
Good luck.
I’ve worked since I last posted day in, day out on the best method to merge and get this all working. It came down to finally a combination first of Dynameshing the merged body and Zremeshing that after a lot of work on guides that would hold the detail the best without also destroying certain areas. The head most especially because of the mouth cavity wasn’t able to work with Dynamesh, so had to be Zremeshed. I then had a ton of time trying to learn retopology first to connect the head to the body and then trying to do it from scratch on the whole form. I found 3D-Coat’s topology tools to be much better than Zbrush’s confusing ones, but the mesh still had problems due to me having rotated the eyelids slightly out of proportion along the way and the symmetry issue.
I did find out it was due to appending the tools to the new medallion tool way back when. That’s why all tools were resized and off symmetry. God, what a terrible, completely worthless feature that is.
Anyway, I decided just using the connected head and body one from before was enough, then working with projecting all the details onto that and shoring the details up in HD afterward. The brow was never connected, nor were the eyelids before, so I first assumed I was going to have to leave them be, because projecting them onto a combined mesh resulted in creases and explosion artifacts.
Then, I discovered the breakthrough that the Zprojection brush actually works. I tried it before and thought it didn’t, but I was only using it one way with the subtract on. This is further great, since the regular projection doesn’t work in HD levels, and the highest tested standard subdivision still had loss of detail anyway. Once I found this out, I then worked on using the new low poly meshes and combining them in 3D-Coat, while taking the high rez mesh pieces and further detailing them. I’m now in the process, finally, of projecting their details onto the new subdivided low poly mesh (19,000 polygons, subdivided 6 times) with all parts combined that needed to be – head, torso, upper arms, eye lids, and brow – in one level of HD subdivision, so at 80 million polygons.
It’s working well, except certain areas. You’ve got to project onto a flatter plane as you can facing you with the Zprojection Brush. It has issues going around slopes, and with crevices and creases. It will cause the mesh to warp into stretched artifacts. Plus, I’m also having issues in spots where the mesh’s polygon makeup is showing through, and smoothing doesn’t cut it to get rid of it. My question now is, what brush, or what method can I use to get rid of the artifacts if they appear, or for this low poly look? :
Even going a step higher in division shows this topology look, though it is a bit less. Smoothing it out, or using other brushes spreads out this mesh area look again, though.
When I get this figured out, I can show the entire model to see all the other parts in a WIP thread, like it gauntlets, cape, and greeves.
