Psmith

Many Wishes Granted I just noticed that many of my requests were "made real" in this last update, (October 7, 2008), but not obviously. First of all, I could not believe the performance boost I was getting in all areas of voxel sculpting, even on my very low powered laptop. I have been assuming, because of my first experiences with the alpha, that my machine was just too slow to work with the resolution of the initial sphere, especially with my favorite tool, the move tool. To my surprise, working with this sphere was fast and much smoother. Unbelievable! And, as if that was not sufficient, the "Clear" tool not only clears the screen of all geometry, (including that high resolution sphere), but introduces a smaller, lower resolution one in its place, already centered on the axis of symmetry. And sculpting this one is lightning fast with all tools. The universal "Smooth" tool does a fine job over the entire mesh and can be used in multiple passes. The "Skin" shader is fantastic. The fill tool behaves like a "super" smoothing/filling tool - really nice. Try the new "Knife" tool with a very high "depth" setting and the hard-edged "circle" alpha shape. "Airbrush" does surprising things with an alpha shape like one of the "dot" alphas. (You want skin pores?) This release is so entirely stable and functional and fast that I am beside myself. And all this on a cheap, Compaq, 32 bit dual core processor with a flimsy Nvidia graphics card with 128 megs of shared memory, and only 1 gig of RAM. Stupendous! Thank you Andrew! Psmith

2 New Alpha Videos Added A moderately detailed human hand is sculpted from a framework of curves. New suggestions for improving the "Snake" tool. Watch Here: The Incredible Curve Tool I hope you like them, Psmith

"Curve" and "Snake" Tool Improvements I've made a video to show difficulties encountered with using the "Curve" tool. Suggestions for possible improvements are also included: Improvements To The "Curve" & "Snake" Tools Scroll down to the bottom of the page, last video. What do you think? Psmith

Spelling Errors in Interface Text Andrew: Should we start a separate thread on this issue? One error I noticed is the word "Stright", where is should be "Straight" under the "Curve" tool's submenu for "endcap" shape. Psmith

This is a very good idea, but I have no idea how difficult this would be for Andrew to include. Psmith

Voxel Export & Retopologization Am I right in assuming that the proper, (current), workflow for merging a voxel sculpture into the regular module of 3D-Coat, for detailing and texturing, should be as follows: After merging, create a retopolgized mesh. Merge that mesh into the scene Create new UV mappings for this retopologized mesh Uwrap Start painting and detailing Is this the right set of steps, given the current, (October 6, 2008) state of development? Thanks, Psmith

Here is a site dedicated to demonstrating new Alpha/Beta developments as they are introduced. Each video is relatively small and streams quickly in QuickTime format. I hope you enjoy them. Watch Videos Here: 3D-Coat "New Features" Videos Psmith

3D-Coat Feature Videos Updated Regularly Here is a site dedicated to keeping up to date with Alpha/Beta improvements. All videos are relatively small and stream quickly. I hope you like them. Watch Videos Here: 3D-Coat Alpha Videos Psmith Edit: I didn't see a category that really fit this post, so forgive me for posting it here. I also posted it under "Volumetric WIP's", but that doesn't really fit, either. Feel free to move this.

Akira: That's the one, alright. Thank you for remembering. Everything on the page you linked is incredibly interesting and revolutionary. I might also encourage you, Andrew, to look at the video demo on that same page, entitled, Direct Manipulation of Subdivision Surfaces on GPUs Absolutely stunning technology which would be the solution to so many character animation troubles, and would also increase productivity among character modelers and animators throughout the industry. I have no idea what the "intellectual property" ramifications would be. Psmith

Dot Based Posing and Animation Andrew: I cannot find the 3D Siggraph paper demonstrating the method I believe to be the best and most intuitive way to pose/animate a 3D character, but here is a developing Japanese product that uses a similar algorithm to quickly pose and animate a 2D character. Just click on the pictures, (ignore the Japanese text, unless you can read it), for a thorough demo of the concept: UrumaDelviPaint Your dot system of posing is already very close to this system, but their system needs only the placement of approximate "IK" handle positions, with no need for anchor points. The algorithm seems to be smart enough to know where to begin the bend of a limb and how much to bend and distort it. What do you think? Psmith

Not So Urgent Request Since a person could place the initial sphere anywhere in space, could the symmetry plane be aligned with the local sphere center, rather than the center of world space? Thanks, Psmith

Urgent Request Andrew: I know I keep harping about this, but could you add a voxel tool that places an adjustable resolution sphere (defaulting to low) either in empty space or indented into the surface of whatever geometry is already present in the viewport? Like the "2D" tool, right clicking could set its place in space or on the surface of existing geometry. Or, left clicking on an object could set the x/y/z surface location and right clicking and dragging could "indent" the sphere, (aligned by surface normal). by varying degrees based on amount of drag, into the surface of whatever mesh is already present. Then, if you really wanted to, you could eliminate the long waiting period for bringing in the initial high resolution sphere by starting voxel mode with an empty scene. Thanks, and take your time, no rush, Psmith

Sounds fine. Lead me to it. Psmith

Time for a Hardware Poll? It might be useful to Andrew to run a "What Kind of System Do You Use?" poll. I honestly don't think there are that many high and mighty hardware moguls among us. I'll bet most of us are using slightly antiquated hardware or newer "low powered" hardware. So, some of us might be reticent to admit to realtime rendering failures, slow performance, etc. At any rate, it might be nice for Andrew to know these things. Of course the poll is anonymous, so there will be no shame. Psmith

3D-Coat as the Future of 3D 3Dioot and others: I do understand your feeling and respect it. We all can't possibly want the very same functions from a program like 3D-Coat, yet, at the same time it is so exciting to see improvements implemented with such lightning speed and proficiency, (by Andrew, I mean). That excitement makes it hard not to request the things we each, individually would love to see happen - especially since it seems like it is so easy for Andrew to add them all in. So, I guess that is what we are doing here, asking and getting - and that at record speed. And Andrew seems all too willing to please. Then there is the ever-present aspect of wanting the program to be streamlined, efficient, attractive - (especially the UI), and acceptable to the Pros. This aspect of development makes us all lean back, take a deep breath and consider what is most needed, and how soon. But, it also seems like 3D-Coat is redefining what it is, all the time. First it was just a detailing and texturing application, but now it is becoming one of the most unique and powerful modeling systems ever invented. And all so fast. What is there to stop it from becoming, eventually, the most powerful and efficient application for doing all the 3D things we long for? I really don't know how far Andrew's ambition and desire is capable of extending itself. I'm sure he will want to narrow things down, at some point in time. And, maybe sooner rather than later. So we all are speaking as emphatically as we can to make sure our set of things gets in there - before it is too late. Here's are some questions I have for Andrew: Do you have any desire or intention of extending the current posing system into something more fully functional, yet totally unique in its approach; something that might even be developed into a character animation system suitable for recording sequences? This is my 3D grail. You have already surpassed my 3D modeling and detailing expectations with the current version of 3D-Coat. To me, every current commercial system for character posing and animation is cumbersome, archaic, overly technical and just plain "not fun" to use. I've seen Siggraph videos showing emerging technologies that create character animation, without complex "rigs" and "skeletal" systems, that look surprisingly like your "pose dot" system, just a little more refined and far reaching in how they "weight" the character's limbs as they relate to the "trunk" of the character. I have no doubt that you are capable of quickly making something like this practical and real - and in record development time. The question is - do you want to include such a thing in the 3D-Coat package, eventually. If not, then I can go to sleep in peace and focus more precisely on simply modeling and detailing, and possibly painting with 3D-Coat. But, I have to say - the overwhelming need and desire for such a system still exists and would be widely received and welcomed throughout the CG community. Both for game character animation development, and for pure animation, developed for enlightenment and entertainment. Still, this is just me, expressing my growing excitement at what you are capable of producing - and how these tools might empower me to accomplish my personal goals. Well, sorry for the ramble and the many words, Psmith

3dioot: In the case of making a rigged figure with limbs, what other method would be better than constructing with a set of voxel curves as the base? Especially once the curve tool is perfected, (giving very smooth joining surfaces and rounded or tapering ends)? Using curves would simply abbreviate any other process for creating a limbed and rigged figure. If one didn't care about ever changing the pose, they could simply use the "snake" or the "spikes" tool for making the limbs. Psmith

Transpose - Pose - Skeletal Movement Personally, I am in favor of posing tools, but having a tool that works like the one in "Brand X", is far too limited in my opinion. The existing voxel "curve" system in 3D-Coat could actually serve as the basis for a kind of "skeletal" posing. Creating the base shape of the figure to be posed, using the "curve tools" - (the dots representing the joints) - would produce a complete figure already completely "rigged". Both construction and rigging would take place simultaneously, at least, in part. Dots could define either a curved juncture or a pointed one, (like an elbow or knee). As long as the joint "dots" were able to follow the sculpted and completed shape of the figure, (always remaining active), they could be used to manipulate that figure at any point in the process. "Cntrl-dragging" on any "dot", (or some other key configuration), could actually set the influence of the "connector" that is the child of that "dot" or joint, on the mesh that it was responsible for making. Other key combinations could be used to rotate the "joint", scale the children, etc. Ideally, something like Blender's system of "Auto IK" would be simply fantastic for posing - (and later, animating) - the figure thus constructed and rigged. What do you think? Psmith

Draw-A-Sphere Tool Andrew: Would it be possible to include a tool which, by dragging, draws a low resolution voxel sphere, either in empty space, or embedded into the surface of existing objects? Each sphere is able to be scaled by dragging, during the time the left mouse button is first depressed to create it. Now, if you included with this tool the option to vary the resolution upon drawing it, you could eliminate the need to start voxel mode with any sphere at all, just empty space. Sometimes the sculptor may want to start his composition with a blank slate - the sphere not being the desired shape to start with. (Sometimes I start my composition with a curve). This would eliminate the need of deleting the original sphere, thus saving time. Thanks, Psmith

Move Tool Is More Than a Mere "Surface Tool", Isn't it? Andrew: From everything I have observed, the "Move" tool is a true voxel tool. Yet, I see it grouped with non-voxel tools like "Draw with Pen". Shouldn't it be grouped with the other voxel tools like "Carve"? I think this is important since it is so unique and powerful as a voxel tool. It stands out above all of the other tools I can think of. In every other application, I believe all the "move" tool accomplishes is the mere "stretching" of polygons, which, I am certain, your "Move" tool does not. When I say, "voxel tool", what I mean is one that actually adds more "marching cubes" to the model - on the fly - as you work. Is this what you mean, as well? Also, the voxel export is fantastic. You can begin retopologization immediately after the "merge". No other application on earth can do what 3D-Coat can do now, even in its "very early alpha" stage. Very exciting! Psmith

Magnificent Move Tool Again Sorry for the 1 channel audio - (bad input setting). Here is the improved version: Watch here: Magnificent Move Tool Thanks for your appreciation, Psmith

Phil: Yes, that's me, no phony speech here. It's O.K., I'd like to see some bandwidth usage over there at SimplyAnimate.com. I will soon be transferring everything to "3DCoatings.net" where I have unlimited bandwidth. That site is down right now. If you want, go ahead and post a link. Psmith

Magnificent Move Tool I like using the move tool the most, since it is a true voxel tool and its results are very predictable. I've recorded a short video that features using the move tool to quickly rough out a human head. Watch here: http://simplyanimate.com/HeadMoveWithTitles%26Narr.mov Psmith

Andrew: Very nice improvements. Here are some of the good things I've noticed: Much faster sculpting when using a scaled down, merged "sphere", (lower resolution for lower powered users). Move tool is even smoother, faster, more predictable than before. Symmetry plane seems to scale in accordance with "increase resolution" Navigation is faster with a mouse. Scaling the brush with the scroll wheel is faster, (but I wish it were faster, still) I did notice that when merging the "sphere", it starts out incredibly small, as compared to its initial size, (big), in the last release. All menu and dialog text is upper case and looks like the names are what you are using for variables, (at least in the demo version which has expired). What about starting the voxel mode with a lower resolution "sphere", instead of the very high resolution one present in the current release? Since "increase resolution" now produces a very good and smooth version of a lower resolution "mesh", I think most users will appreciate the speed at which they can work, at the beginning, with a lower resolution voxel sphere. Thanks for your labor, Psmith

3dioot: In defense of my requests, let me start by asking who you think will comprise the majority user base of the 3D-Coat software, (especially in the first development stages) - High end 3D professionals Serious 3D students Beginning & advanced Hobbiests Other Andrew must have a good idea who he is aiming this software towards, especially at the current price point and feature set currently usable and under development. I would argue that the professionals are already entrenched in either daily use of ZBrush or Mudbox. Both applications supply a very useful arsenal of tools for the professional 3D artist. Why, at this point in development, would they jump ship to a newly developing application? If Andrew is intending to develop 3D-Coat for the purpose of making money, then I'm certain he must pay attention to who he perceives are his initial adopters as well as potential adopters. Also, I'm pretty sure Andrew has a very good idea of what directions he intends to develop 3DCoat, of course, based on his observations of the market and user requests. Comparing the development of 3D-Coat with the creation of any masterpiece - generally speaking, the artist proceeds from the general to the specific. In the case of software creation, I would think Andrew is working from general functionality, (all the time refining existing features), right down to the extremely specific details that you, yourself, have requested with fervor. From my observations, the ability to retopologize general mesh structures is a highly requested feature and now available in almost all popular 3D applications, including Blender. I believe this is due to the growing trend toward game content development, (where attention to topology is extremely important - especially with regard to economy of polygon usage). The inclusion of topological functions, it seems to me, fits the category of "general" 3D software functionality. 3D models designed for the ultimate purpose of game content, 3D animation and even rapid prototyping applications, (at this stage in hardware commonality and affordability), must be represented by means other than raw hundreds of millions of polygons, (where detail is concerned). 3D-Coat offers a very nice compromise, (utilising displacement mapping) for this hardware limitation. As you know, this already exists and is usable in the "regular" 3D-Coat module. For voxel sculpting to become completely practical, for the most popular uses stated above, the model must be translated into polygons to be capable of being used in the "regular" 3D-Coat module. An automatic conversion algorithm is very difficult to implement, and not usually pleasing to the end user, anyway. That leaves manual conversion as the only alternative. This, in turn, translates into user defined retopologisation, (not really a word, but I think you know what I mean). You have expressed the need for very specific and, as you have said, "perfect" brush functionality for use in the work that you do. Not just in one or two of the existing brushes, but in all of them. I would suggest that this request, if taken totally seriously by Andrew, would be the equivalent of asking Andrew to put most of his energy and time into "painting pores" into the skin of his rapidly developing masterpiece before he has even roughed out the general proportions of the overall application "figure", so to speak. Regarding user ability to improperly use content developed with both voxel tools and then regular tools, for commercial usage: the ability to export from voxel module to regular module could be limited to the demo version, only. Psmith

Andrew: Regarding production of base mesh within the voxel module: (thinking in parallel terms relative to ZBrush functionality), roughing out a base mesh with voxels quickly, for re-use as a retopo base, approximates using ZSpheres to make a very rough, low resolution (adaptive skin) for re-use as a sculpting/texturing base. Your voxel tools already provide more functionality and refinement than do ZSpheres for producing a base mesh. I know that creating an "automatic" algorithm that produces a usable and proper low resolution cage is a very difficult thing to accomplish, (especially one that pleases professionals and hobbyists alike), but many people are bypassing this step, (since nobody seems to be pleased with how programs like ZBrush produce an adaptive skin), and heading straight for the retopologizing approach. And your retopo tools are faster and more efficient than anything on the market, even at this early stage of development. I encourage you to implement the translation of volumetric "base meshes" for use as a "retopo base mesh" as soon as you can, seeing your algorithms for retopo are so good, and, for the reason that most users will probably be wanting to retopo manually rather than rely on any kind of automatic remeshing function you might so aptly invent. Lots of development time will be saved and 3D-Coat will be launched as the 3D artist's "application of choice" much faster. And, for users of low powered hardware, this is also the most practical approach. Thanks for all your work and enthusiasm, Psmith