jedwards

Thank you very much Sir!

Back to the topic and original post... I definitely have a preference to use 3dcoat at the beginning stages of a piece. Voxels give more freedom to realize forms early on and even refine them up to a certain point. At higher resolutions though I simply prefer to send the work over to zbrush for finishing. I think that as the painting and sculpting brushes improve though I may be more interested in keeping the work in 3dcoat in time. Most of the new tools we have now took less than a year to develop, so I fully understand that refinements will happen and look forward to those. I think rounding out the UV and polygon mesh editing/management may help here too. Currently I prefer to export a retopo mesh to another app for UV work, once I bring it back in I'd love to be able to merge it with other meshes and their UV sets, make tweaks etc. Consolidation of multiple meshes into one happens quite a lot when building game assets so having the capability to do those things in 3dc will make it more of a one stop 3d tool as well. Bottom line for me though is that zbrush is simply more mature as a tool for refining sculpts. That's nobody's fault. Pixologic got there first after all. Andrew simply needs time to refine his tools and build upon what he's done so far - which is very impressive to me even now. A year ago I had no use for 3dc in my work at all. It simply offered nothing I felt I could use to speed up or improve how I work. Today I try to use it as much as possible because I simply enjoy working with it.

The power to shape surface detail would be amazing if this brush worked this way AND had profile curve adjustment. The way it behaves now is part of the reason I don't try to do as much detail work in 3dc.

I pretty much agree with everything you put in requests... I've still got some good use out of it myself, but without all those things you mentioned it does take longer to get there.

I'd be willing to try that... whatever helps eliminate the distortions is good. The dropping of selections when you hit undo is a pretty big workfow killer too, I agree. I've found posing in general to be a trial and error type thing so a lot of times you end up undoing the transform when you realize the selection isn't as accurate as you want it to be. This forces you to constantly repaint everything quite often. At this point I've learned not to use transpose once I'm into higher resolutions. I try to nail my pose early when I have more flexibility with it. Even if I decide I'm not 100% happy later on, I generally leave it alone as the final piece will be fully rigged for posing/animation in another app anyway. I'm still glad to have it and do get some use out of it in early stages. But it doesn't quite offer the same flexibility as its zbrush counterpart yet. It can only improve from here though. =]

The Copy brush does this. Create a new empty voxel layer, and use it as your active layer, then whatever you paint over in the viewport with the Copy brush will show up there. I like this much better than mesh extraction, which I use a lot in zbrush.

It already is... I'm using .08C right now.

I'd like to have a toggle for transparency in the voxel tree, perhaps nested within the visibility toggle so you could cycle through them by clicking on the button. This would be useful when you have multiple voxel objects overlapping each other and you need to be able to sculpt on one that is being covered by another. Sometimes it's more useful to be able to see an object while you sculpt on another rather than just making it completely invisible. I use the transparency feature for subtools in zb for this quite a lot as often the changes I make on one subtool are in direct reference to another subtool that it intersects with. Selection and masking options, both paint, lasso and window modes - as 3dioot already mentioned elsewhere, it would be good to separate this stuff from other tools such as pose, etc, as they can be very useful for other things than doing transforms. I do lots of detail sculpting by painting masks in ZB and then applying the inflate modifier - which will raise or lower unmasked surfaces evenly across your mesh. Great for doing intricate line work, etching etc. There is ongoing discussion elsewhere in these forums about refining 3dcoat's brush and voxel tools to be better equipped at handling higher resolution sculpting. I would really like to see more focus put into these areas through this release as currently I don't find voxels easy enough to work with at higher resolutions either. My preference right now is to build up a sculpt, retopo it and then finish that off in zbrush. I'd love nothing more than to work to full completion in 3dcoat though, especially if I could take advantage of normal mapping, occlusion baking etc all from a high resolution voxel mesh. At some point I think greater focus on the UV tools is going to be necessary too - perhaps the ability to join multiple retopo meshes and their uvs into one would be good. Often this type of process is done in a 3d app, since they are better equiped to handle that stuff. A lot of my characters are built and sculpted from many pieces, which get consolidated down into only a couple of meshes with one or two UV maps for baking and final texturing. Being able to do all this in 3dcoat would save another export hassle. This would mean a need for more basic 3d editing tools eventually. Not a big priority for me yet... I'm much more interested in seeing the voxel and paint tools evolve. This is where 3dcoat has the most potential to improve and gain customers. I don't think there are many artists who are interested in using multiple sculpting apps the way I use 3dc and zb right now. I think that if voxels ever get fully realized, you'll then see a lot more people migrating to 3dcoat as a one stop solution rather than adding it as yet another tool, complicating their pipeline.

Even using the default shader with the color falloff doesn't give 100% coverage. There are often spots on a voxel mesh that simply cannot be masked regardless - even when trying to smooth over them. I agree with 3dioot in that we need to separate posing from selection. Hopefully this would also allow for performing window and lasso selections, which are often much faster than having to paint everything, and might even guarantee full coverage without spots. This is a major wishlist feature for me, not just for the posing capability but for selection/masking in general, which is a major aspect to detail sculpting as well. Being able to paint complex masks in zbrush and then use the inflate modifier to give even extrusions is something I do almost every day when sculpting details for next gen game assets. One thing I'd love to be able to do when it comes to posing multiple voxel meshes simultaneously is go into each voxel layer in the tree, make my selections with a paint or window/lasso selection and then, do my posing in whatever layer I want, knowing that all my selections in the other layers will come along for the ride. This would be especially powerful when posing a base body mesh for example, where all the other accessories, equipment and clothing are located in separate layers. Just make all your selections in each layer, then pose the body itself and have the rest match that same transform.

Would be cool to be able to browse such libraries from within 3dc... kinda like XSI's netview or modo's new preset browser, with little thumbnails that represent the brushes so you can literally choose them to work with on the fly. Probably a bit more work upfront, but a great way to keep the community involved directly in improving 3dcoat. I know I much prefer being able to sift through a set of brushes or presets and just use what I want rather than having to download and then sort through and discard the stuff I don't want. Being able to categorize them would be gravy too.

Most anticipated software purchase I've made in awhile. Upgraded both my pc and mac versions and absolutely love how it performs on both systems. Thanks for all the hard work and dedication to your product and your community Andrew, it is very much appreciated. I hope it is rewarding for you as well.

Thanks for the comments all! YES I love the copy brush! I use mesh extraction quite heavily in zbrush and this is so much better. I'll definitely need to refine the clothing more... it is barely touched from the initial copy I made from the base body. I mostly created the extra details on it like the cuffs and collar by using the copy brush as well and refining those parts first so now I need to go back and make the wrinkles, folds and extra volume that defines them as clothing instead of skin.

Another WIP, just messing around with the voxel tree and the copy brush. Started from a sphere as usual, just because you can in 3dc. =] It's broken down into about 15 pieces mostly sourced from a quick blockout of the full body.

That's actually a feature... and you don't even need to wear special 3d glasses! =]

I definitely think the resolution is there to do high rez with voxels. What I'd really like to see though is decent masking functions for being able to detail like in zbrush, but also to be used for increasing resolution locally. Voxel meshes would be much more efficient if you could localize resolution at key areas like on the hands and face while leaving the rest at lower resolution. I think the brushes really need some refinement as well though, or at least more customization for behavior by the artist. Is brush curve editing going to make it into the final release? On my iMac I up-rez'd a mesh to 18.7 million triangles. It took about 4 minutes to process but the program didn't crash and I could actually still do detail work fairly smoothly. Viewport navigation slowed to a crawl though. =] That was more of a test than anything. I'm more likely to keep single meshes at or below 5 million tris, and even my iMac can handle that. This is where controlling resolution in key areas would come in handy. For a single sculpt it gets a bit tricky, but if I break things up over several meshes I can work much as I would in zbrush with multiple subtools. 3dcoat is way easier to work with for building from scratch though. I love the fact that I can now start a character directly in this app and not have to build base meshes.

I get that error dialogue every time I load a new session on my mac. As far as I know the program is not crashing when I close it and the file I was last using appears fine. I've just been clicking on 'skip' and ignoring it.

64 build works great on osx so far! Thanks for adding in the quadrangulate and paint per pixel so quickly!

... and also just being able to use the quadrangulate and paint function in the voxel tree popup menu =]. Will be nice when I can just take my voxel sculpts, retopo them and start painting with the new tools. I know this is in the works and just takes some time. Just really excited to see this stuff coming together finally!

It would be great if each voxel object in the tree could have a transparency toggle along with the visibility toggle - or even just a global toggle that sets all layers except the one you are working with to transparent, so that you can see through objects that might obscure your view of the object you are currently sculpting. It's often handy to use the surrounding objects as visual reference rather than just hiding them completely. I use this feature in ZB quite a bit - especially when objects overlap or intersect. Now that I'm using multiple voxel objects per file it's becoming more of an issue.

Btw I did a similar thing to your proposed symmetry/copy method Andrew, I created a new empty layer, used the copy brush with symmetry on the first finger, then just used transpose(select object) to manipulate it with symmetry active and it was fine. So long as I didn't use the transform tool all was good. Multiple ways to approach this I guess, which is always nice.

I just gotta say, as someone who depends on the mesh extraction tool in zbrush, the copy brush in 3dcoat is so much better! Hadn't been able to use 3dc lately as my pc is out of commission. Finally broke down and installed it on my mac today as well and had a play with that copy goodness. Wow. Just, wow!

Awesome Andrew. I can't wait to try painting some low poly props and enviro pieces. This should also be really good for bringing in assets that already have their own normal maps and other textures for touch ups or whatever.

I'm not a fan of the Hex GUI at all. Never liked it. Too many damn buttons. I don't care how pretty they look, each one of those means more distractions that pull you away from what you are working on. UI means much more to me than pretty gfx or cramming more stuff in it to imply that you'll be more productive or take the app more seriously. Silo has the best UI in my opinion, basically because you can skin the buttons and general appearance to look however you want (everyone wins this way). But you can also turn all of that off and drive the program completely just with your mouse/tablet and a couple of keybinds. This makes it more valuable to me than any of the bloated beasts like Maya, or XSI or Max that I own and use at work. Hexagon can't do that either - which actually does put it in the realm of 'children's toy' for me. Contextual behaviour is a must also and customizing user input down to the mouse and tablet are also a must. If someone doesn't like the navigation they should be able to change it to whatever they want. That means not just the way maya or lw or modo work, but even creating a unique nav scheme of their own if they want. Again I can do this in silo. Button clicking is a horrible user experience when sculpting or painting and seeing that Hexagon GUI all the time would just annoy me, knowing that at any given moment, more than half of what is there is not needed. I agree with the idea of 'rooms' or modules as they are nice and contextual and cut down immensely on the amount of GUI you have to deal with at any given time. This aspect of XSI actually makes it the least bloated of the studio apps I've used - though the key and mouse inputs are something else entirely. =\ In 3dcoat I don't want to see retopo or paint tools when I'm sculpting, etc. So long as I can quickly toggle between the different modes I'm quite happy. For me, even stacking the key and mouse assignments so that they can have unique assignments 'per room' would be even sweeter, as it would cut down on excessive binding and allow you to reuse the most valuable binds for the most important tasks in any given room. This is a big reason why I divide my polygon modeling between the likes of Silo and UVLayout and Zbrush. They all are dedicated programs that put input priorities on the tools you use the most for their specific purpose. I look at all three as basically being separate, fully contextual rooms for the task of working with polygons. This way I get the best tools with the most focused workflows for each. It's actually easier for me to work this way despite having to send obj files flying around my hard drive than it is to do everything in Max, or XSI or Maya. 3dcoat should be no different. It is not a bloated studio app. It has a narrow focus (sculpting and painting) and so should favor a focused toolset and user experience and avoid dumping everything in your face.

Hey 3dioot, I think that would work for me. Perhaps my post seemed a bit too strong... Just wanted to make it clear where I stand in terms of all the designs and ideas floating around. It's true, my preference leans toward the more boring, but standard manipulator look, but I say why mess with something that others put a lot of effort into already when we have bigger things to worry about. =] We do need a manipulator though and that's why we're all here talking about it I suppose. I like your revision. I didn't make it clear in my first post that obviously I'm still 'ok' with the silo manipulator... I've been using it for so many years, I must be, right? =D But I should have stated that simply proper color coding the rotation handles would easily improve it imo, as your illustration shows and without a lot of fuss. My preference is for rings, but I could live with what you propose in those last images too. And as I mentioned I do like the idea of planar toggles. You've got a simple and functional design and I think that is a good enough start for this. Should also add I prefer the colored axis handles too over black or transparent. Much easier to read.

The silo manipulator is a bad design. I use that app every day for 8+ hours and it's like it was designed just for the sake of being different. I love its toolset, and prefer it for modeling over anything else, but visually, the manipulator handles don't make sense. You don't put a Y rotation handle on an X axis transform handle and then color it red when Y axis transform color is green. This is why rings around the manipulator were created for controlling rotation. The ring flows around the axis it controls rotation for, AND is color coded to that axis. Function is implied immediately both through the form, placement and color. You can also grab it from a much broader area in the viewport as opposed to a tiny handle crammed in with the other 2 transform handles (too much clutter imo). This is also why screen rotation is represented by a ring around the whole manipulator in other apps that use them - it is more consistent with the other rotation rings, and is easier to grab which pretty much all other 3d apps do barring a few exceptions not worth noting. Even though I like the screen space rotation ring in silo it doesn't make sense because it is not consistent with the single axis rotation handles. I don't know why it was 'borrowed' for use with the silo manipulator when by design it is not consistent with the handles they chose to use for controlling rotation. Same goes for the planar transforms. Placement along one axis when the handle represents 2 axes AND color coding it to just one axis is also visually, not right. Design wise, it makes no sense. There is a disconnect between the representation and the function. This is where apps like max actually get it right. As much as I don't like modeling in max, I do think the way it represents planar transforms is better than what I'm seeing here. Visually, you have a plane handle that stretches between the 2 axes it represents, immediately implying planar behavior. Color coding is neutral as well (just like with uniform scale, or screen space), which is easier to associate with planar manipulation behavior than red, green or blue. The only handles that make sense in the silo manipulator are the single axis scale and move handles because their shape, placement and color coding all imply their purpose. Ultimately for me though, planar transforms just aren't as important. I pretty much never do planar scaling, not in all the years I've been modeling, and rarely do planar movements. Not if I have easy access to screenspace transforms. I do not miss planar handles at all when they are not available to me in some apps. In terms of their importance vs the visual clutter they add to a manipulator I could do without them and not even blink. Also, I'm a big fan of the blender approach as well (as a keyboard alternative to the manipulator). Hit a key to transform, then hit another if you want to constrain the transform. I'd be more likely to use this method for a lot of my transforms if it were available, as I've grown rather fond of its simplicity in blender. In fact, I even prefer blender's manipulator, 1. because it adheres to the same standard as others, and uses rotation rings, but also, because I can toggle the display of all the transform types on and off for it as I like. If I don't want scaling I can turn it off leaving just rotation and movement handles visible. Combined with its keyboard controls this makes for an awesome way to personalize it to your liking. The ideas I have liked so far here have been to add the center area for screen translation (must have) around the uniform scale handle and the planar constraint toggle (because I can keep them hidden and probably never use them). While I like the way max implements planar handles, I don't like how it is all mouse-over activated. I'd much prefer to toggle their visibility when I want to use them, but in terms of location and visual representation I think they are a better approach to take than what I've seen so far. .02