New Features in Marvelous Designer 7 are announced



  • Preview is here:

    There will be a function to trace out parts on an avatar in 3D space and get the corresponding flat pattern in 2D, which then can be used as panel in clothing.
    The avatar can of course be an existing conforming clothing part. This would make it much more easy to reconstruct existing garments as MD clothing and do grading, scaling and refit.

    There are a lot of limitations and things one does not realize on first sight and it will take some iteration to get it fully smooth. The start is there, however.

    I have no idea about the launch date.
    MD traditionally is quite liberal with public betas and free 30 days trials are normally reset when a new version is launched.



  • @F_Verbaas said in New Features in Marvelous Designer 7 are announced:

    Preview is here:

    There will be a function to trace out parts on an avatar in 3D space and get the corresponding flat pattern in 2D, which then can be used as panel in clothing.
    The avatar can of course be an existing conforming clothing part. This would make it much more easy to reconstruct existing garments as MD clothing and do grading, scaling and refit.

    There are a lot of limitations and things one does not realize on first sight and it will take some iteration to get it fully smooth. The start is there, however.

    I have no idea about the launch date.
    MD traditionally is quite liberal with public betas and free 30 days trials are normally reset when a new version is launched.

    I think I saw somewhere (maybe on their site?) that release will be some time in September. This looks like an AWESOME update!



  • Looking at the video again (while drooling) ... here's what excites me.

    First I really like that you'll be able to draw the pattern on the avatar. This opens up so many possibilities. The thing I hated most about MD is getting the flat pattern pieces positioned correctly before sewing, even with the arrangement points (though that made it much better).

    Modular clothing creation. This is a cool idea but I'm hoping that you'd be able to add your own "parts" to the collection so that you can add your own personal touch to the garment.

    STITCHES!!!! Awesome. I usually add those manually in the texture. These look to be geometry because there appears to be some depth to them.

    Working ZIPPERS. Say what? Exciting!

    Those are the features that I'm most excited about.



  • The only problem with this is that I see myself increasing my already bloated runtime with even more custom cloth items and spending less time in traditional 3D modeling apps...

    "I need yet another xxxx cloth item!"
    <looks at bloated "knit shirt" runtime directory"
    "No I don't. But, I'm know I'm going to make one, anyway, so there's no using fighting it..."

    Looks nice!



  • Marvelous Designer 7 closed beta is available for MD6 owners.

    best regards,

    Bopper



  • @bopperthijs
    As I understand anyone with a valid license to MD 3, MD4, MD5 or MD6 can join.



  • A quick question for MD users - I haven't looked at it in a long while. How is its quad conversion? IIRC, it can convert its native topology to quads, but the last time I looked at that, quite awhile back, it was less than stellar. Anything on that, lately?



  • @morkonan
    You ask at the right moment.
    MD7 comes with a possibility to make offsets along a curve
    0_1505507624796_Knipsel.JPG

    MD 6.5 had already the possibility to draw lines on the garment and turn these into internal lines.

    Meshing is uses internal lines as a boundary condition. You can therefore control the mesh in full detail: draw your critical loops and detail out the folds. Then use quad meshing and make sure the particle distance is larger than the disance between the internal lines.

    Do all his however only when you are ready to export the model. In a simulation quad polygons are potentially unstable and can be a drag.



  • I actually use MD2, and I'm very very happy with it.

    When they had the upgrade program to MD6.5 a few months ago I was tempted, but the upgrade policy sounded terribly complicated, something about "3 versions back", and MD2 cannot upgrade but don't miss the upgrade promotion (which sounded conflicting to me), and I don't know what not.

    And then on top of that options for monthly, and yearly and perpetual... brrr... complicated...

    This is one of those cases where getting too complex turns the customers away. I may have made a mistake, but then again I'm pretty happy with MD2.



  • @fbs7
    You are right there is a danger of overdoing. There is, as far as I know no example or competing application so we are looking at fresh ground broken with every release. Results may vary.
    The public communications of MD always have problems getting past the language barrier. The MD2 licenses run through an older system they now need to operate manually. The upgrades from MD2 to 6.5 also were handled manually, them sending an invoice and client transferring money with reference to the invoice.
    The reduced price for upgrading if you have a license 'up to 3 versions back' is a commercial policy.
    It may be worth while though, if you are still interested, to send them an email. The saying in Dutch is: 'A no you already have. You could win a yes'.



  • @F_Verbaas said in [New Features in Marvelous Designer 7 are announced

    Do all his however only when you are ready to export the model. In a simulation quad polygons are potentially unstable and can be a drag.

    Thanks!

    So, basically, if I understand correctly, the topology is its own during simulation and construction. These "internal lines" are some form of control/simulation surface/operator in order to help simulate folds, etc. (Not sure how they would figure into some things, other than their density, in regards to cloth types, but whatever. :)

    However, these internal control surfaces are not exported as topology, correct? So, when converting to quads on export or whenever that occurs, particle distance...

    Why larger? Admittedly, I know nothing about MD, but just from the description you've given, it would seem I'd want a smaller distance if that's what controls the mesh density, so as to preserve the simulated folds/etc. ?

    Anyway, thanks for letting me know it can be done. :) I'm looking at this for simulating cloth draping and complex cloth bits that I'll then bring into another app and then on to rigging, etc. If I couldn't get a decent quad model out of it, it would be worthless to me.

    On purchase programs: They're obviously designed for two purposes - To protect the IP and to maximize revenue. In a seller's market, which most apps like this fall squarely into, "customer service" and "convenience" are rarely of any consideration. That's just the way it is. Oh, and complicated purchase/update/licensing plans suck for personal and single-users. "Managing" one's own "property", which it isn't, but whatever, to simply maintain access and functionality shouldn't be part of a individual's life-long plan for using a product.



  • @morkonan
    Simply said: the meshing fills the maze between the internal lines and panel edges.
    Maybe an example explains better.
    The image shows the right-back hip panel of a corset I made for Roxie using the 'flattening' technique, that is: take developable panels from the avatar, in this case Roxie.

    0_1505581327092_Knipsel.JPG
    In the 2D view, on the far on the right you see the panel with tri-mesh.
    On the image in the middle you see the quad mesh, near perfect except for some anomalities in the top edge, driven by the mesh of the panel above which I did not yet give the same treatment and is still 'rough' tris.

    In the 3D view on the left you see the internal lines I added in red. (I used the 'Offset as Internal Line along Curve' function.
    The triangulation or quadding that MD does fills the mesh formed by the internal lines and panel edges. Of course if a mesh is a well formed quad and the edge length is smaller than the particle distance, there would be no need to subdivide. (As I did this I see that MD7 insists on making at least one subdivision. MD 6.5 did not do that. I will have to investigate.)
    So, any iternal line you add comes in the mesh as a series of edges, which can be used to detail folds. A gathered seam can be made using a 'grass' of internal lines with fold angle setting if necesssary..
    Peserving simulated folds within MD can be done with the 'Solidify' function.

    Access to MD2 functionality is not in question. I just tried to explain why upgrade from MD2 was arranged differenty.



  • @F_Verbaas said in New Features in Marvelous Designer 7 are announced:

    @morkonan
    ,...
    So, any iternal line you add comes in the mesh as a series of edges, which can be used to detail folds. A gathered seam can be made using a 'grass' of internal lines with fold angle setting if necesssary..
    Peserving simulated folds within MD can be done with the 'Solidify' function.

    Access to MD2 functionality is not in question. I just tried to explain why upgrade from MD2 was arranged differenty.

    Gotcha. So, the "internal lines" will define a physical edge in the model. These are also useful for guiding/using MD's specialized features. Okiedokie. :)

    That, in the question of exporting a clean model of quads, "it can be done" is what concerns me. Since it can, then I'm at least minimally confident it won't be a waste of a cloth simulator. :) I keep forgetting the name of the algorithm that results in MD's sub-d type, or whatever "cousin" MD uses if it's proprietary. But, the main issue, topologically speaking, is -

    alt text

    ;)

    Do you have any refs where it has been quad'ed using its standard practices? IF I can get it out quaded, I can reasonably automagically retopo it in some ways. If not, or not usably so, I'll have to back up and punt. :)



  • This thread has reminded me of something I read some years back regarding topology for dynamic clothing. I forget the name of it (and didn't save the article) but it mentioned that MD was the best program for creating this preferred type of topology. As a serious Blender basher I was disappointed, but nonetheless looked into MD. Other real life projects took me away but I'd like to give it a go again. Does this ring a bell with anyone?
    Thanks for any replies.



  • @baginski said in New Features in Marvelous Designer 7 are announced:

    This thread has reminded me of something I read some years back regarding topology for dynamic clothing. I forget the name of it (and didn't save the article) but it mentioned that MD was the best program for creating this preferred type of topology. As a serious Blender basher I was disappointed, but nonetheless looked into MD. Other real life projects took me away but I'd like to give it a go again. Does this ring a bell with anyone?
    Thanks for any replies.

    There's a discussion on these forums concerning the Cloth Room that goes through several topics on topology and which is most suitable. Also, the Cloth Room used to be based on proprietary (sublicensed) algorithms used in 3dsMAX/Studio, so you could try looking around the 'net for topology discussions, there.

    As far as any proprietary algorithm/topology in MD, I wouldn't know. It "looks like" Butterfly sub-d, to me, at first glance. (There's another one kinda like that, I think, will have to look it up.)


  • Poser Ambassadors

    I believe you're looking for "Delaunay triangulation"



  • @bagginsbill said in New Features in Marvelous Designer 7 are announced:

    I believe you're looking for "Delaunay triangulation"

    Aha! You're right. I'll do some reading and this time, download and keep it all.
    Thanks a bunch BB.



  • @morkonan said in New Features in Marvelous Designer 7 are announced:

    Do you have any refs where it has been quad'ed using its standard practices? IF I can get it out quaded, I can reasonably automagically retopo it in some ways. If not, or not usably so, I'll have to back up and punt. :)

    I did not save that exact example. Here is side by side a tri-meshed example and a copy quad meshed. Particle distance is 20 mm in both cases.

    0_1505760689602_Knipsel.JPG

    I hope this helps.
    BTW from what I hear from those who are retopo-ing it normally breaks the flat UV's. The nice thing with MD is you do the UV's from the beginning.



  • @bagginsbill said in New Features in Marvelous Designer 7 are announced:

    I believe you're looking for "Delaunay triangulation"

    Thankee! Yup, that's it.

    @F_Verbaas said in New Features in Marvelous Designer 7 are announced:

    Here is side by side a tri-meshed example and a copy quad meshed. Particle distance is 20 mm in both cases.

    Thank you so very much for that! AHA! I've seen this topology scheme before, so now I know where it came from! (Or similar. Seeing several articles like this with "wtf" topology and seeing your example clarifies their original construction and being quaded from that.) UV is not too big an issue as I wouldn't be using it straight out of the box, anyway, just as part of the initial generation and construction of the fabric/folds/etc. Thanks again!


Log in to reply
 

Looks like your connection to Graphics Forum was lost, please wait while we try to reconnect.