SYN7AX3RR0R

proflooney, here's a couple of thoughts on the retopo for that model. - what do you inted to use this for? if just to texture and do a still render, i would run an auto reTopo on it with general strokes, i wouldn't go crazy on it. pretty much what ever comes out will be good enough. the thing you have to get right is the density. i would go at least as dense as what i got in this example. i would even subdivide it in retopo one time to get extra detail. the mesh will still be 'light' enough for a still render. for instance my retopo mesh is under 29 000 polys, with an extra subD it's under 115 000. that's nothing for a still render. this will hold your silhouette very good. if you're struggling with the new auto reTopo algorithm with guides and density paint, just revert to the old auto retopo. for this you will need an extra subdiv in retopo room to hold up. after the retopo u need to UV it if it's for animation, i doubt you can make it work. u should start with a more rig-standard symmetrical pose. hope this helps

michalis, you really pulled my smart-ass string LOL no offense, but i don't get what you're talking about. are you hoping for a feature quality forground character no-tweak-auto-reTopo tool? i doubt you'll get that any time soon bro. you have an issue with a couple of ngons and such, so what's the alternative? manual retopo? retopo at all, or would you rather old school box model it LOL i've already used the old auto reTopo algorithm in production on commercials for props both in foreground and bg. stuff coming from sculpting packages that we'd retopo and use normals on (we managed to avoid displacements). i've also already used the new algorithm in production on a documentary, and it rocks. it's working amazingly good. YES, ofcourse there is lot of room for improvement, naturally. BTW, how long since andrew released this as a public beta? a month? have some faith dude. so back to the question at hand. i need to finish this sculpt pronto, i'm done with the concept sculpt in sculptris and want to get this in Z for posing and detailing. for a 15 min 3DC, and 20 min silo cleanup and polygroup setup job i'm more than happy with this. too bad you're blinded by a couple of ngons to see the production value in this. the re-topoed mesh was subDived 2 times and projected to get back the detail, polycount ~460 K

here's my second go at auto retopo. this time with an actual purpose. i'm doing a challenge ovet at CGSculpt, my initial sculpt is in sculptris and i wanted to take it to some other package for posing detailing, probably ZBrush. i wanted to share this as there's still not too many auto retopo screenshots out there. sorry for big images, i'll edit the post when i get time.

i run the updated version through a file with the same guides (i had to repaint density as it is not saved [this should now get saved with the file ... hint andrew ]) the new version of autoreTopo v3.3.12 is following guides much better, yes. here's the same model with same guides as before

proflooney, yes you can control the poly count; also poly density by area (that's the first image in my post, you paint the poly density map) the poly count is dependant on several factors, it takes a bit of time to get a hang of. the whole process needs a bit of playing with to i just read this on andrew's twitter: Also I improved following guides-strokes in auto-retopo. 3DC will follow guides much better. 8:53 AM Aug 11th looking forward to try the new algorithm out. i've had success using guides on broad areas but not in tight areas with lots of detail here' s a suggestion for andrew in terms of reTopo polycount. i would love to be able to 'hint' the poly size. it's a much more visual way of dealing with polycount. - paint density hint - draw guides - use something like 'points and faces' to create a polygon (leaning on a guide) in certain areas to hint polygon size/density and flow. make any sence?

I've just tried the new auto retopo algorithm and wanted to share some of the stuff i got. I'm looking forward to playing with it some more, learning how the algorithm works and how i could control it better. Here's my first test, top images are from 3DC, Density Paint and Strokes edge-flow hints. Bottom images are from Silo, a comparison of the original mesh (from Sculptris) and the 3DC auto reTopo mesh.

hey polyxo, thanks a lot on the quick reply and all the info!

Hi, Anyone know if there's a tool in 3DC that is something like the MRGBZ Grabber in ZBrush? Basically I want to capture the elevation detail from a region and use that as an alpha to clone it on surfaces. Thanks

i've got a BFG 8800 GTS OC 320MB, i have no issues with voxels in 3DC CUDA 64 (OS: win7). the issues you guys are experiencing is that something specific to the 470/480 series and 3DC? i was thinking of upgrading soon, and this is valuable info, thanks in advance.

When in PAINT mode; go to OPTIONS > Pressure Sensitivity (max it out!) this should fix it. I have to add that this has developed into an amazing tool. It has quickly become my favorite app for quick jams. Here's a couple of models I did while beta testing it: I've captured the second one. Links if anyone is interested:

awesome stuff, whole thread is packed with cool examples of what 3dc is about. very good sculpting love The Noble Chin sculpt and concept.

amazing stuff. this whole thread is packed with great sculpts. some of the best 3dc stuff i've seen. your anatomy is very good! the latest one is looking neato!

it's a 320MB BFG 8800 GTS

Hey, I thought you might be interested to hear about how far 3DC went on my system. I had to convert a model that was modeled in regular methods to be 3D-printer ready. Basically have it as one watertight mesh. So I figured, I'll import everything in 3DC as voxels, perform the needed booleans, and export. In order to meet the required specs some boolean work was required . The largest piece of the bunch was 21.7 mil tris. While I was performing merging of the voxel layers and exporting the mesh I hit 7.6 GB of RAM usage. 3DC was the only open app I had. Neat, this is the most I had gotten from a single app so far. I've spent 4 days on this project, and got just a couple of crashes. The stability was great, speed of processing was also admirable. 3DC 64 CUDA 8 GB RAM 8800 GTS Q6600 win7 64 Ult

got some time to finish this sculpt, merged the eyes in and made some minor tweaks. i'm calling it done as far as the main volume goes. anyway, here's the latest 3dcrender and the links to the vidz. 360 from 3dc, divx

cool sculpts. all of them look good, congrats

hey GED, last 2 are really good. love the venom sculpt, it's got good volume an silhouette!

in my personal experience there is no perfect (G)UI. everyone is different and has different workflows. the best UI is the one that is easiest and most efficient to customize. the best workflow is the one that adopts to the user. in light of this i think you're going down a dead end street trying to revitalize the UI, you just can't please everyone. i think it's better to put it off for the time when you can open it up (open it up completely) to the users to do their thing. in a couple of months you'll end up with a bunch of GUIs here on the forum. i think it would be more productive to start talks about what we want to have open to customization, and in what ways one could actually change stuff (like setting keys from the menu, hotkey editor, CTRL+LMB on a button, etc [] being able to drop PNGs to screen as buttons and drag them arround freely; being able to snap them to existing ones, etc) for instance this was a 'begginers UI' i had created for silo. basically with most of the commands all on screen, this is something you'd toggle the visibility when needed (cool with sticky keys), BTW IKHandel you should have downloaded this when you were learning silo i agree with IKHandel's remark on blender UI and works on 2.5. look at ZBrush and how they have alienated tons of users with their stubborn 'we know it best' attitude to customization, one of the reason why mudbox was developed was because pixologic did not want to implement some stuff WETA was asking for, i'm gonna go out on a limb here and guestimate that navigation/camera, brush adjustements, and layers were big here.

instead of scaling the proxy you can also do this: VoxTree (pop-up) > Select Layer > Increase Resolution of Current Object (button on the bottom) you will get a multiplier number to indicate rez

i'd love this idea. i'm not a fan of the way this UI looks, but whatever, it's not a big deal. i had made a couple of UIs for Silo, it was kind of a begginers UI with most of the commands availible on screen. something one would use before setting up their keys. i remember some friends i demoed silo for kept the UI as it provided them with quick access to all commands. anyway, what's even more important than the graphicUI is the shortcuts! i tried customizing 3DC shortcuts, but it's a mess. if you assign a customKey, the old command that was bound to that hotkey is not dropped! this makes it next to impossible to create custom hotkeys. this forced me to revert back to the default and basically have to use 3DC with just 2 hotkeys (SPACE; and 'T') that's all i use to sculpt in voxels. works ok, but only because i can not setup custom hotkeys. this is a big issue andrew, i hope you look into this. silo's hotkey system works really well, also the fact that they opened pretty much every command and setting to be bound to a hotkey is great. these are a couple of things i'd love to see in 3DC: - ease of setting up hotkeys (must be efficient) - all commands, settings, windows, etc to be opened up to hotkeys - easy way to add custom setups to presets (another thing silo does well). as an example i worked in a studio that had 3 artist use silo for modeling. now, the fact it was so easy to customize meant that it was next to impossible for us to use each other's systems because it felt like a different app. here's where having settings in a drop down preset was great. for instance for 99% of my voxel sculpting i use only 2 alphas, i'd love to be able to drop those to hotkeys, this would speed me up incredibly as i switch between the 2 very often. (i do have to note that the only reason i would use this is because you can not create a custom brush that has the alpha 'stick' to it, this would be a better idea i believe) as i mentioned i'm pretty comfortable using 3DC as is, using just the 2 hotkeys i mentioned. it's usable, it's fast and there is not a lot of mouse travel due to mouse related menu pop-up. if it was not for those 2 shortcuts i probably would not use 3DC. ease of use is king in my books, and there is no way u can please everyone's workflow, the key is to have the application comfort to the user. creating a hotkey system that reflects this would go a long way.

hypercam. thanks, i was thinking of adding some voice over if i ever finish this and get the time for that.i've been working on this a bit more when i have the time. updating the vid as i get new footage. this is where i'm at the vid clocks at 30 min now. same link: youtube (10 min cunks, this will be updated more often) Pt4 is latest: http://www.youtube.com/watch?v=sdvjuC7NCmk

SSDs are still pricey. 3.1.16 64 CUDA boots in ~3 sec on my system. win7 64 Q6600 8GB RAM HDD: WD Raptor 10,000 RPM

congratulation on making it into the gallery! this piece turned out great.

voxels > Adjust > Pose excuse me if i'm doing something wrong here and this is already part of the tool functions, i'm still pretty new to the app. the behaviour of the pose tool is buggy in the way it's implemented currently. what makes it hard to work with is that the tool stays planted when doing translations. when using the Move function both points should translate, this would allow for quick translate-rotate-scale operations. as it is now, after each translate move, you have to readjust the pivot, this makes no sense. in rotation the gizmo should should also 'stcik' (one point planted, other following). in the Params window there should be an option to 'Reset Gizmo to World' and 'Center to Object'. even more important is a snap option where the 'floating' end of the Gizmo could be snapped in relation to the 'anchored' point. this should work with a modifier key, for example SHIFT. in case i'm not clear, i could draw a couple of sketches to illustrate this

thanks folks. here's a minor update on the eyes. i've started capturing this session from scratch. other than the speed and this a 1/2 size vid, it's the whole session unedited. you can see all the struggles and bugs LOL. it's quite sped up, the idea is to try and keep it interesting, i'll need at least as much time to finish this thing up. i'm interested what you guys think of the speed of the vid?