@Timberwolf : VWD is an standalone application. it communicates with Poser through Python scripts. Thanks to the competence and the work of @Philemot, the Carrara and DAZ Studio versions will operate on the same principle.

@piersyf : I think there are two problems in the rendering you show.

• The mesh appears to have very large meshes.
• The arm of the character seem to pass through his body. Here are the methods to correct this simulation.
• Correct the pose if possible to avoid interpenetrations.
• Refine the mesh of the cloth using the "make compatible" function.
• Increase the character's collision distance. If the pose can not be changed:
  • Use the functions of the "interactions vertices" to prohibit collisions between vertices of the cloth and those of the character that are problematic.
  • Use the "Nail to collision" so attach the parts that have tendency to behave badly on the character. A bit like the stockings vidéo.

@NolosQuinn : You can further improve the stockings simulation by adding a slight elastic function (95% for example) on all the mesh.

@NolosQuinn : This new version will be available at the end of the week. I prefer to do some additional tests to make sure of proper operation. The simulation with the stockings, does not need a new feature of VWD. You can do it with the current version.

@Smaker1 : Thank you for this beautiful rendering. You are right, the dynamic deformation feature solves many problems.

@shedofjoy : You can buy directly VWD via PayPal. Its price, in this case, is 50 Euros. It will be very quickly sold on the Renderosity's Market. Its price, in this case, will be 69 Dollars.

I just released three new videos on YouTube showing the new features of VWD.

These videos are unfortunately in French for now. I will try to republish them in English at the end of next week.

Physical protection is actually a problem for the distribution of VWD. After careful reflection, I have decided to remove it. This will allow to see very quickly VWD on sale at Renderosity. This will also simplify, for me, the distribution of the program.

@MrFiddles22 : I understand your question about the future support due to the material protection. For now, I'm in full development and I do not intend to stop. In the event that I can no longer send a code corresponding to a particular machine, I can tell you that when I create a new version of the program, I actually creates 3 versions:

• Normal protected version
• Demo version
• Protected against reverse Engenering release without material protection. If I cannot send more code, it is this version that will be sent to Renderosity. I hope I answered your question.

I have some problems with the new version and its dynamic movement. I will probably take one week or two before to finish it.

I give just some news to say that I just wrote the function of dynamic modifications of a cloth or a hair using the mouse.

I think that this feature is more interesting for people who wish to make renderings poses, but I wrote this function so that it can also be used to correct animations, ie locally modify the dynamic object in a particular frame.

I will certainly do a video showing all the possibilities of this new function.

I can say that VWD will work soon for Carrara and DAZ Studio using a collaboration. I also think I can say that the sale of those other versions will also be made in Renderosity.

@shedofjoy : The next version will include a checkbox that will mean that this is a new animation. The choice of default value is not easy, because in case of chaining simulations, it is imperative to retain the previous animation.

I think I can tell you, without errors, that VWD should soon be available on Renderosity. ;-))

There are the first discussions of a collaboration in order to also provide a version for DAZ Studio and Carrara. For now, I have not found a way to satisfy people working on Mac, but I think we must go step by step. Sorry!!!. For the moment.

I have found motion capture files which include the dance animation that I had published on Youtube. (At the time, I did not know). There are some changes to be done in the animation, but there are no major inconsistencies.

Here is the address: https://sites.google.com/a/cgspeed.com/cgspeed/motion-capture/daz-friendly-release

I have made a new fun video using one of these motion capture files that I will publish today.

@tomyee : I don't know if you have watched the two demonstration videos that I have made in French. The first one shows a simple simulation of a dress and hair. The second one shows a sitting pose with a deformation of the seat. These videos are unfortunately in French, but they will allow you to understand two basic functions of VWD.

@tomyee : I think it is difficult to compare VWD and the Poser ClothRoom, because they do not do the same things. To be able to animate conforming clothes, I have had to add completely new functions. All this makes it more difficult to use since it is not possible to work with it successfully, without understanding its operation. In fact, I completely agree with you, for now VWD lacks terribly of demonstration videos. It's my job of January. I want to make at least a twenty videos showing all the features of VWD. I could have done them before, but they would be in French, which would not help many people. I do not promise a perfect English ;-((

@headwax. : Thank you for your comment, I find you indulgent regarding the documentation. I want to match it with the latest version including the corrections made by "carrarators" that I thank for his help.

Concurrently with the videos, I would like to work on two functions. I am waiting for your comments.

- A .RIP file editor, to create, delete functions in a .RIP file but also to change the selection of vertices.

- An editing tool for a recorded animation before sending it into Poser. This tool could use a dynamic without inertia which would be a correction more consistent  than just moving vertices.

@piersyf : I did not quite talk about distance modifications between the mesh layers, I talked a distance of collision. This may allow to define different collision distances (with the collision objects). This already solves many problems of collision between close objects or close layers. It would indeed be possible to define a distance of collision between the different layers, but, for me, this is no longer collision functions but self-collision functions. We can think about it. With regard to wet hair, it is necessary to experiment.

@piersyf : yes, you can use the selection of materials. This vertex selection method is very convenient in many cases. The lack of memory is a problem due to the 3D visualization library that I use and that is 32 bits. It is true that this limitation may seem boring, but we must remember that the clothes we work with, are conforming clothes that have not been created for the simulation. Maybe one day, I completely rewrite the visualization of the program and I will rewrite the program as 64-bit program. Simulations that have this error, already largely exceed 30 million springs, making them slow to simulate. I just bought "Wave hair for G2F" because I find these hair very realistic. I have a memory error at the tenth of the number of vertices (450,000). If the program were 64 bits, I think it would generate about 300 million springs, making the simulation... endless. For this hair, I had to set the distance limit to 4.85. This means that the springs were generated between 4.85 and 5.0. Yet the simulation passed without worries. Do not hesitate to use this function limitation because it reduces the number of springs and therefore the computation time. It also helps to avoid memory errors.

I released a new version of the program. This version allows to make a mesh subdivision. This function allows to simulate more conforming clothes, but it also allowed to make the simulation of the pleated skirt because, with a mesh generation at 0.5 and retaining the spring generation to 1.0, it is certain that each fold will be stiffened.

I also just wrote a function that lets you assign a distance of collision by vertices. This is actually very useful for multi layers clothes or to push out the bottom of the hair to prevent interpenetration with a cloth.

@Vestmann : seeing your message, I've realized that I had made a mistake. Indeed, the old files do not work anymore since I have added the new mesh subdivision function. The mistake I have made, is not to have added them to the end of the parameters, but in the current order. If you want, I can write a small executable that makes the transformation and thus make the old .RIP files, compatible with the new version of the program.

@piersyf : rereading my message, I told myself that I was wrong to adress you personally my post as it has nothing in what I said that is personally addressed to you. I always protested against this way of thinking that makes that people believe it exists everywhere "magic" buttons. They never exist, and if you are able to make quality things, it is because you have worked hard and nothing else. I always rebel against this infantilization of people.

@Vestmann : could you tell me if the new version asks you the same problem as the previous version.

@piersyf : I will answer you on the two subjects of your post.

The selection of vertices has been a great difficulty to write. In addition to the methods that I had integrated in the program to its beginning, some users have asked me to add new methods. I think that, in the current state, the existing methods allow to make any kind of selections. The operations of the calculator type are the only one able to offer much flexibility. Some clothes selections require a rather complex selection but it should be remembered that .RIP files allow to repeat simulation parameters.

As regards the types materials: cotton, wool, leather etc ... I tried to explain why I have not done it and why I will not do it.

I tried many cloth simulators and I've never found that the type of material corresponded to that proposed. I'll try to give you a technical explanation.

First, more than thirty years in the field of development of programs linked to the numerical simulation, have taught me that the operation "click button" never works. What for :

To successfully assign a behavior of a certain type of fabric to a mesh, we need three conditions, it is necessary to know: the mesh size, the shape of the mesh and regularity of the mesh. Apart from some adaptive quadrilateral meshes that generally yield poor quality simulation results, there is the quadrangle and the triangle. The quadrangle does not achieve meshes with regular size. The only solution to obtain a regular mesh is to work with equilateral triangles. To be able to create a pattern of behavior, the mesh size must be all identical. The best mesher program to achieve this, is called HyperMesh and costs tens of thousands of euros per year. There is also a free mesher that often gives very good results and is called GMSH. In both cases, it is necessary to start with a geometry file (.STEP, .IGES). If you meet these requirements, then you can claim to affect a repeatable behavior to a mesh. These conditions are hardly met in dynamic clothes. For conforming clothes, that, remember it, were not created for the purpose of simulation at the time of their creation, these conditions are never met. I refuse to have a cloth to which I claim to assign a cotton behavior and that behaves like leather as seen too often.

Please excuse me for such explanations that may seem too rigid, but I have not written VWD to look like the others, but only because the others did not allow me to do the simulations I wanted to do.

@Vestmann : Yes, you use the Recorded imports with the right way.

I use very often the Recorded imports for hair and I have never had this problem.

In first thinking, I see only one reason you have problems with the hair and not with the clothes.

With a cloth, the selection of vertices that are attached to the character, is made from the cloth itself. With hair, the selection of vertices is done first on the character, it is then transferred to the hair with a remote control and it is then transferred back to the character for generating the springs. You might have this problem if you try to affect hair on a character different from that which was recorded in the Recorded import. Is that the case?

If this is the case, you would have to rename the Recorded import with a very descriptive name for the hair, such as "MyHair For V4", for example.

@piersyf : This problem of animation that does not match the desired one is recurring. When importing a collision actor, the program looks at whether there is a cache file corresponding to the number of vertices of this actor and having a number of frames equal to the current animation. When doing a simulation using a collision actor and that we try a different animation than before but having the same length, the program thinks that the animation is the same.This possibility is very useful when the simulations are linked, ie when a dynamic actor becomes a collision actor in the next simulation.

In this case, there are several solutions:

- Delete files in the "Exchange" directory, as rightly says Biscuits.
- Press the "Record" button for the concerned collision actor.
- Change the number of frames of animation in Poser (set the number of frames to 101 instead of 100)

I thought putting a check box called "New Animation" at the time of the creation of a collision actor. This checkbox will not be checked by default . Pressing this box will remove the cache file of the actor. Tell me what you think of this solution.

@estherau : I understand your problem and it is part of those on which I want to work in the coming months. The ultimate goal would be to transfer fully VWD under Mac. I just ask you a little time.

I just released a new video on which I had spoken. This is a simulation of a pleated skirt. Not necessarily very easy. Before the new mesh subdivision function, this type of simulation was almost impossible. Now, it is rather simple. Here is the link: https://youtu.be/9KpvPAlkAdQ

@honzu : The numerical values problem is now solved. Now even if you keep the comma as the decimal separator in Windows, it will be points that will appear in all numerical values. This will be corrected in the new version that I would (finally !!) send next week.

The subdivision of meshes works fine. For some clothes with initially very long meshes, it is necessary to increase the number of sub-samples. I will publish on YouTube, an animation of a pleated skirt that would not have been possible before this technique.

In Python buttons, I also added two functions respectively for copying and pasting of materials. Once a conforming actor is transformed into dynamic actor with VWD, it is no longer possible to assign materials using poses. It is now possible to apply a material pose on the original actor (invisible), copy its materials, then reselect the dynamic actor and paste the saved materials.

Things are moving forward with Jenn and I am hopeful that we finally will found VWD in the Renderosity MarketPlace in the early months of the next year. While this is not immediate, it is a great advance because I really thought, at one time, that it would not be possible.