Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
Almost every full time PA that does clothes uses MD. It's a market standard for what it does. I have Blender cloth add-ons, but they are both harder to use and Blender cloth sim is vastly slower on the same machine than MD cloth sim. At the moment I'm using Blender for base meshing, the paid Blender Geometry Remesh add-on for retopology, and then MD for pinning and simming the finished meshes. MD 9 has the ability to import UV islands as sewn cloth pattern pieces, which is a lovely feature, and tremendously useful.
MD works fine on my crappy CPU on both my laptop & PC for Marvelous Designer files (Zprj). Basically, CPU is slow for Objs that are high res and MD garments that are low particle distance.
MD will be slow if:
1) You are simulating your marvelous designer garment at a low value "particle distance" = high poly. You can change particle distance of MD garments easily per each pattern piece. The default is 20. People selling MD garments will probably set it to 5 because it looks better. (Youre not supposed to work at 5!) You can also deactivate/freeze pieces as necessary and only sim one piece at a time. The only time you will want to go below 10 particle distance is when you are doing your final animation. Yes it will take a long time to do a high quality/low particle distance animation but that doesnt mean its not useable on average CPUs.
I dont make animations therefore the long simulation time for low particle distance animations is not a problem for me. I only animate for the express purpose of going from Pose A to Pose B, so I keep it above 10 when animating and only drop it to 5 for final frame of simulation.
2) Likewise, if you are not using a MD garment but instead are using a high poly mesh that you imported as 'Obj to garment' then it will simulate slowly. Unfortunately you dont have the luxury of changing particle distance on the fly for Obj meshes, but you can remesh it to be lower resolution in some program like Blender, but the lower the poly your mesh the less well it will look in simulation. CPU is really slow to simulate high res objs. GPU is better for this purpose, but GPU is also garbage in lots of ways.
3) MD is slow to synchronise on the 2D window if you have lots of garment pieces, so it is better to work on a few pieces at a time and merge them together later. And only do stitching as last step, as stitching also slows down the synchronisation.
Does MD have settings for different types denim, silk, and other textiles? It that what you mean mostly be 'limitations in MD'?
Blender: Sometimes I import as an object and use it as is, other times I import and convert it to a shape key. I prefer shape key option as I set up the materials once, then just apply a shape key. It can get finicy about converting quads back to tris and quads do look better and are easy to isolate for me into material zones. It is not uncommon to find a few triangles scattered about.
MD is not a renderer therefore how a surface looks is not something they worry about. MD does indeed have different simulation settings for any type of fabric so they behave accordingly.
now Zbrush has dynamics you might see less people considering MD the industry standard
Mec4D has been showing some awesome stuffs using it with PSD templates
I have the impression that the large amount of "paperthin" clothing we see today is a result of people starting to use MD. Is that correct?
MD is also a modelling program. Its simulator is top notch though not unique, but as far creating clothes goes no other software even comes close.
Pretty much. Simulating cloth with thickness is tricky because there's no such thing as thickness in 3D space. Every simulator treats any closed 3D object as a sack, empty on the inside. "Thickness" is usually added after the fact and it's usually avoided becase it more than doubles the polygon count.
I wonder if Luthbel uses it
Strange, my experience was quite the opposite to those saying MD is slow. Maybe it was the version I tried (an early 9-something, I think.) but it sims in MD ran rediculously faster than dforce sims (1 clothing article on 1 character in scene for both). Using an I7-4700HQ processor, 24GB of memory laptop. It was so fast that I completely quit using dforce. Just exported to MD, simmed and imported the morph. I could do every article of clothing for two characters in a scene in the time it would take dforce to run a single fifty frame sim on one article. Maybe I was dforcing wrong? Guess it doesn't matter a this point. After MD showed how slow dforce actually was for me, I have never used it again even after MD expired.
No, you completely dismissed that the type of textiles a cloth is made of affects the draping of the different pieces of garment. The actual visualization of a specific textile type via shader materials is an afterthought if the garment type's simulated geometry doesn't behave correctly. The only way I can guess MD would be able to simulate specific textile weaves and materials would be the geometry density but I could be wrong and the cloth simulation engine has built in preconfigured or configuration file parameters that define all those texture types. Having a leather jacket that drapes like silk kimono isn't a very realistic cloth simulation when you get right down to it. Of course I wouldn't expect Blender, Unity's, UE4's, nVidia's or Blender's cloth simulation to have the same response to the same parameters but they might as I think some of those cloth simulations engines are actually nVidia's cloth simulation engine under the hood.
Well that's from not adding cuffs and seams properly, something MD if I've understood the intro sales pitches can actually do. However draping everything like it was silk or like it was cardboard would give the same effect, even if that added seems and cuffs.
There is such a thing as thickness in 3D space or we'd be talking about 2D space. It does have to be modeled though.
Read the second sentence I've written in the paragraph you quoted: MD has different simulation settings for each type of fabric out there. It's not a renderer. It's a simulator. Of course it can simulate silk as well as leather as well as latex therefore it has the simulation settings built into it.
I already explained it in the post. "Thickness" has a specific definition when dealing with 3D elements, but if you're not a modeler I understand the confusion. A 3D cube has zero thickness to it because thickness is a visualization trick created by extruding polygons from the initial surface. A cube with added thickness would translate to a cube within a cube visually. The computer does this automatically for the user if he chooses to but in 3D space no new geometry is created. Why? because adding thickness in the way you're thinking of "thickness" doubles the polygon count at the very least.
A simulation engine works by applying physics to vertices. It doesn't understand thickness, it can only understand vertices. It doesn't understand that what you see as colloquially the thickness of a piece of cloth is just that. It thinks that both sides of the fabric are two seperate unlinked surfaces.
thickness really should be something the render software can do
in theory DAZ studio and iray could
someone needs to create the shader script
if one has a geoshell it should be possible to have the ends bevel on the original and a mirrored geoshell using displacement to create thickness
The sharp-edged garments that have no thickness is a problem for realism and I have no idea what the technical solutions are but I do wish that more PAs would add fold-over hems and edging. I know it is possible in MD because I followed a tutorial in the early days of owning the application. Of course, when it comes to draping such clothing, I'm constantly presented with poke-through using dForce wheras MD has the ability to pull the cloth while the drape is happening. That one featue would be a huge improvement to dForce and I would have thought they would have been able to incorporate it by now. After all, dForce is no longer new.
@Faux2D I wrote a Blender script that will pin every vertex in a certain vertex group of one object to the nearest corresponding vertex in another object. This was so that I could attach things like capes and unconventional skirts for simulation without having them fall to the ground.
I wanted to do the same thing in MD, but it doesn't have a strong scripting environment. I tried to do it by hand, manually, but there seems to be no way to deal directly with individual vertices and say "I want to put a pin here. Precisely here. On this exact vertex." It's as if the metaphor of a seamstress creating an item of clothing is inviolable.
I would love to be wrong if you can show me where I am, but it seems like MD is only a modeling program in the sense that it represents real things, and not in the "polygonal modeling" sense. When I need to do precise things like that is the only time I don't use MD, and well, I wish I could always use MD :)
Well, to get dynamic clothing you're going to have to have paper thin clothing and then do shader tricks to create thickness. If you're making conforming clothing then it's pretty easy to create the illusion of thickness with a simple L-extrude.
I have issues with that exact thing as well in MD. The program was initially aimed at fashion designers but nobody cared. It wasn't until some game showed how it can be integrated in a pipeline, that's when people started noticing it. Now it's a must-have for any content creator. They slowly adapted it to fit more with the needs of modelers: remeshing, UV export, GPU simulation, sculpting, etc.
I hope MD 10 will have the feature of being able to properly select geometry and manipulate it accordingly. It will also be neat if we coul choose certain parts of OBJ's by their materials and choose different simulation propreties.
I solved this issue in MD by just adding extra geometry in the extruded garment to hold the two layers together in MD. The end results are similar to what you see in Pixar movies.
I always make MD garments (i.e., not talking about obj models) that are two layers so that they can look 'thick' even when exported as 'thin', and curve the edges using fold angle of internal lines. I realise this is wasted geometry so sue me. Also realise this is not going to work well in Dforce, but this is an MD thread not a Dforce thread.
Also i have no idea what people are talking about in this thread anymore so i should probably butt out.
Edit: As in example attached, you can see an emulation of thickness in some of the places where i bothered to add two layers.
A lot of great stuff I saw coming from MD was edited and sculpted in Zbrush for the final appearance.
There are 2 things about "Thickness" one is about the physical appearance of the model and second is about the cloth "Thickness" simulation
I tried to simulate a heavy blanket with Physical Thickness, I still had to set up the simulation for the virtual "Thickness" as it of course will not behave this way no matter if it is plane with thickness or not.
You can simulate a single plane as it was heavy leather and add later the thickness for the final appearance in rendering and that is the best option to avoid issues.
I also made exotic plants tests with natural thickness and simulate them, it works beautifully but the surface settings were a pain to set up as it requested a completely different approach and it was 100% more intense in rendering time so totally not worth the effort.
The difference between silk and linen is not only the roughness of the surface but also the translucency of the material, if you want to get crazy using SSS on each material and render it forever, go ahead .
But you can do the same just by adding an edge panel to simulate the physical thickness with less consuming time and power resources for the same effect and render faster.
There are however exceptions like windows glass or any glass that should have the physical thickens for proper IOR and light Caustics.
But clothing ? in most all cases not unless both sides can be visible in 3D space at the same time but definitely not T-shirts.
But going back to MD , I don't used the last version but before the major issue was the topology, if you use the clothing just for dynamic simulation then all is fine, but if you want to rig proper on a character you will run in issues as the fluid motion will be lost, when you look closer in 3D movies you will see that only partial fabric is simulated and not everything , close to the character body fabric types need to go with the weight maps of the character with the proper topology and not bounce like a Jojo by each movement or like a rubber .
Ask an MD designer to make a leather jacket and run it with dforce in a walk simulation in DS and make it behave like one.
MD is a great software to create accurate base clothing models using real world patterns but that is only half of the story . However it is great for a starter if you have more on your mind than a nightgown
and Zbrush is total different story as well, you can recreate the same steps as in MD but it may take a little longer since you have to weld manually all patterns for more complex clothing (create stitches)
If you watched my last video on youtube about how to create dynamic curtains in Zbrush, then it is the same way MD do it, just different approach of preparation of the pattern in Zbrush, since the stitches are not really stitches but simple edge loops connecting (welding) the patterns (poly planes) together and the simulation contract them at the Z axis into one 3D object plus gravity simulation at Y axis so there is no magic in MD but working on a simple principle all simulations do.
The big difference here is how you approach the design but the same result between MD and ZB
That is my personal opinion about, don't mean you have to follow, as always you need to find for yourself what works best for you, as nobody else can do that for you.
Not ZB or MD making clothing , You do ! and you only need to master it to make it happen ! opinions don't!
Cath