Mudbox 2014
polycounter lvl 16
Mudbox new features guide
http://download.autodesk.com/global/docs/mudbox2014/en_us/
Autodesk mudbox 2014 unfold videos
http://area.autodesk.com/2014unfold/products/mudbox.html#whats-new
New Features Highlight:
(Advance Retopology Tools)
- Mesh Reduction
- Remeshing
- Patch holes
- Smooth selected regions of a mesh
http://download.autodesk.com/global/docs/mudbox2014/en_us/
Autodesk mudbox 2014 unfold videos
http://area.autodesk.com/2014unfold/products/mudbox.html#whats-new
New Features Highlight:
(Advance Retopology Tools)
- Mesh Reduction
- Remeshing
- Patch holes
- Smooth selected regions of a mesh
Replies
and merge paint layer for export is also a great one...
the ideas forum is also cleaned up...
http://mudboxfeedback.autodesk.com/
Loser of this year: Max?
There's been alot of work go into Softimage in recent versions. 2014 is the first release done by the new product/dev team. 2014 has had alot of really good bug fixes, so it's always tricky to strike the balance between bug fixing and tweaking/enhancing/new features. Whilst new stuff is great, we always want to try and ensure the products are as stable and reliable as we can.
No new features stood out to me with the Max update. Looks like Maya is getting new modelling tools (please just give it all the modelling features Max has), so that's good.
I like Mudbox and it's clean user interface/experience. Hope the retopology tools are easy/nice to use.
What I really want:
I don't think that's going to happen - there's too much profit to be made from having all those 3 as seperate software suite's, not to mention people already being settled into using their chosen software. In terms of Mac for OSX though I'm surprised that hasn't already been done
[ame="
Nice!
is that automatic retopo ?
i thought you had to use curves first to set the edge flow.
this new retopo algorithm is really good... beside tricky areas its doing a 90% ready mesh...
fingers, eyes and mouth is still hand work...
I can't imagine them getting that right even if they tried.
Though an app with all the modeling tools from XSI+Max, and Maxs modifier stack, mixed with Mayas animation setup, and ICE, with all scripting done via a common language like Python, could be amazing.
auto retopo and dynamesh is a different thing... dynamesh is based on a cube grid and cutting the mesh in small faces...
auto retopo is more like qrmesher in zbrush... it tries to create a clean quad topo...
and you dont have to use curves... just hit the auto retopo button...
http://area.autodesk.com/2014unfold/live/unfoldevent/workflow.html
and i noticed that the new retopology in mudbox seems to be very odd.
*symmetry is broken in retopo
*too many edge collapsing going on in random places.
*non-uniform edge looping in random places resulting in clusters on dense retopo (see example in lips)
can any beta tester confirm these issues ?
if 2014 is like this, i will have to skip it yet again and stick with mudbox2011
ps. i am wondering whether or not mudbox2014 is worth upgrading so i posted same question in AREA forum. in this current state, i will stick to mudbox2011.
8k textures, amicable.
clear and fill paint layers, thank god.
all the other big stuff, mostly production ready useable tools.
multi-touch, its fluff to me.
lets keep the fluff down at about zero%. no need for gag-features, non-features, and bloat. no one is going to be using this on there nexus 10, i hope...
i just downloaded mudbox2014 demo...
so far i am finding the retopo pretty much useless.
the topology is very buggy, not symmetrical and guide curves dont paint on symmetry when part of that symmetry is backfacing.
combine this with previous years old bugs and screwed up move/grab brush i am just
btw, i tried it several times and i did convert the curves to hard constraint but still ended up with a messy topology like that the above.
not always. i found out that if i am continuing from another curve then it only draws curve on the painting side and the symmetry gets broken.
notice that the small orange circle that indicates connecting is only on one side too.
But I did find something cool. Zbrush make planar equivalent in mudbox.
another alternate in Mudbox has already been there since 2011 in flatten brush.
if you want to get 100% planar you can use combination of scrape, then use flatten and make sure you uncheck "update plane" in flatten brush properties. this will conform rest of your brush stroke in that plane.
As for this new version, I dunno. It's cool and all but I think it would require some bigger paradigm shifts than just this auto retopo to really bring it to the next level. As usual, features are being incrementally added but it needs a stronger vision driving it forward. The new detach faces is a pleasant surprise, but there's not much point to it if there is no way to stitch vertices back together...
At the moment it remains what it's always been - the best app to sculpt accurately from a basemesh, but still way behind ZBrush when it comes to making things from scratch.
in the example onionhead_o posted i think make planer is a better option then flatten brush. but in general flatten with a high strength is more versatile as you noted.
also 2013 introduces falloff curves with unlimited number of point. that feature is what gave zbrush most of its flexibility to make brushes. you can make a brush with a custom falloff and stamp that has the identical look and feel to claytube. and actually make brushes that are much nicer if you tweak all the brush options. that feature alone makes a huge improvement and is worth the upgrade.
also verical trays, tile plane, paint improvements etc are totally worth the upgrade.
@pior
making things from scratch in zbrush is almost always slower and more error prone then sculpting a mesh. when you have to factor in the endless revisions that always come up after you have started your animation cage. the extremely high polycount of some pieces. the nasty issues that come up with kickbacks after rigging, etc. voxels despite being fun to work with become a huge pain in the ass. its cool for design and if you have a personal project that has no time constraints but it is inferior on a tight schedule in most cases.
also map extraction based on subdivision/uvs as apposed to ray projection is a huge advantage. it ensures almost perfect error free maps in a fraction of the time it takes to do ray projection. there are many trick you can do based on subdivision/uvs that just do not exist with un-ordered non-hierarchical voxel meshes.
@ Gray
you can input unlimited points in falloff in 2011 as well. just have to right-click inside that small falloff thingy and insert points.
also, i noticed that regular brush stroke is messed up in2014 if i have a sharp pointy falloff.
it looks like dotted spacing in the stroke even though the spacing is set to 0.
also, the grab brush is not good either. they changed the grab brush in 2012 and still fucked.
i am gonna stick to 2011 for now.
if you have a mesh with 60,000,000 polys and you take a large brush and make a fast stroke across a large distance then yes, you will get skips in you stroke. the only other option would be to have the application freeze up while it tries to calculate a huge number of brush samples and tries to apply every one to the mesh. in most cases that would make the application non interactive. no sculpting package works that way.
if you have a hard time with these strokes then turn on steady stroke. that will help you make more consistent smooth strokes.
i would never set stamp space to 0. thats part of the reason why your strokes look bad. i usually have it set between 1 to 5. and if your working with stamps usually around 20 to 90.
edit:
really? unlimited falloff points in 2011it is introduced in 2013 docs as a new feature.
i was seriously bitching about that. should have tried right click. :poly127:
but also keep in mind it has a new large window that pops up now and point tabs. so it makes it much better to make a precise falloff. this is really helpful for paint brushes.
and there is so many new things from 2011. the gigatexel engine once you figure out how to tune it is just crazy. so much other stuff.
hehe, from your reply i can tell that you are either not experienced with mudbox sculpting or just lack proper information.
i been using this tool since the beginning and i can assure you that it is neither my lack of hand-eye coordination or my lack of powerful hardware.
that mesh is a sphere @ level 7 = roughly 6.2 mil quads and not even remotely close enough to lag my system. i still get ~30fps with that mesh, so no lag.
the problem is not about steady stroke, it is about steady displacement from the stroke.
also the same error persists in mudbox 2011, but it is just lot less there. it is more evident in 2014 as far as i see.
stamp spacing 0 is not a problem. however disabling stamp spacing is a problem which i am aware of and it is not disabled.
well hey, there are few people that have been around since "the beginning". and i am one of them. i was on the mudbox beta pre v1.0 back in the days of skymatter etc. i remember buying v1.0, it was an amazing time. were you on the beta for v1.0? even if you were not and you go back as far as skymatter there is no need for "old timers" to get testy with each other. so if you were offended by my post then i'm sorry about that. that was not my intention.
i will tell you what tho. when they sold mudbox to autodesk it sucked really bad. i think if they would have stayed independent then it would be a much better package then it is now. and yes the tools have changed and so has the whole architecture of mudbox. some of the old brush behaviour was better then the new brush engine and updates have introduced. but honestly i think there are lots of improvements that have come along that make mudbox better then it ever was and they have not really done to much to fu*k thing up or add stupid useless crap to copy all the stupid crap that ends up in other packagers. the brushes are still far better then anything else out there so i do not have such a harsh opinion on where the brushes are now.
really i have no problems with the new brush engine and have not had any of the problems you seem to have. it sucks. i hope you can get things worked out because most people don't seem to be having those problem. anyways i'm not going to argue if its justified to upgrade because if you want to use 2011 then that's your right. but for me as a long time user that was there at "the beginning" i'm really impressed with the new features and glad that autodesk is giveing so much resources to mudbox development. you should file some bug reports if you find them and try to get your workstation checked out.
edit:
in general, the smaller the stamp space the smaller the interval the brush has to make a round trip to sample the surface do all the calculations and update the surface given the current stroke. so setting stamp space to 0 is really not a good idea. you will never see any difference and get better performance.
i still prefer mudbox over zbrush for lot of reason that would take too long to explain.
the last time i been on beta forums i tried my best to provide some alternative vision for the software, some "thinking out of the box" if you will but it wasn't met with much enthusiasm.
seeing as Autodesk werent able to retain 2 of the 3 main mudbox devs, it tells me that Autodesk doesnt really care about innovation as much as they care about cashing in on gimicky features.
*edit* stamp spacing 0 is really not the issue here. i have tried many setting over the years. it is a limitation of the software no matter what setting or hardware you use.
hehe, well if you go back to the beta then your words are gold in my book and i consider you a fellow traveller.
honestly i have not cracked open an old version of mud in a looong time.
question: which was the last version that did not have the problem with the brushes that your talking about? id seriously like to check this out to see what the deal is and make a comparison. i have no doubt that your having a real issue but it does not seem to effect me. so maybe it is an issue that can be fixed.
edit:
also what wacom are you using and what driver version. it could be a wacom driver issue. because it looks like its a sampling issue where its not reading your stroke.
also yea watching what has happened to skymatter is painful and most people don't even know what happened. but what the hell are you going to do about it? its just money and greed etc. so you have to make the best of it. it would be nice to see someone come along with a new package but it looks unlikely at the moment.
edit 2:
i had high hopes for 3dcoat. the programmer is top notch but it just seems like they have gotten so totally off into every direction. zbrush copied some of there best features and now 3dcoat is copying some of the crap from zbrush so its just bad. very depressing.
frankly, that brush stamping issue is the least of my problems in the new 2014 release.
i have not upgraded since 2011 because they messed up the grab brush behavior with symmetry along x axis with release of 2012 and it is still a issue in 2014.
more details here:
http://area.autodesk.com/forum/autodesk-mudbox/general-discussion/revert-grab-brush-to-pre-2011-state/
wacom drivers are also not the issue here, trust me.
analysis of my own work flow is that its a situation that does not come up enough for me to have noticed the problem. i had noticed that the grab changed but i did not run into that specific issue where you have 2 thin strips close to the axis and its a pain in the ass. in most cases grab does not have that issue. and when i run into it i generally just freeze one side grab then mirror the changes. you could also put it in a separate layer etc. my guess is that you are probably one of a hand full of people that caught that one. it does not seem to effect most peoples work flow enough for them to have noticed or complained about it.
they updated a lot of things. and i know a lot of the changes were for speed improvements on super high res geo. i think the new one works base on the depth buffer and brush radius. it obviously has some issues that the surface grab did not have.
your other issue with the brush skip i would have noticed. i really do not have that issue. the only way i can duplicate it is if i take the stamp space up to a high value or when the brush is tangent to the normals facing perpendicular to the camera. and in that case its really impossible to calculate the correct spacing due to the fact of the depth buffer i think. was it better in older versions? i would have to go back and do some tests to find out. i have not seen it crop up in everyday sculpting. i cant say that its any worse then it has always been. i definitely get clean stokes at high res without any noticeable sputter unless i'm at really high res with a big brush or perpendicular. nothing like what you posted in that pic. if i did i would certainly notice. im not sure what to say about it because i have not seen anyone else with that issue and its not noted as a bug. if i come up with any ideas what it might be then ill post them. i cant duplicate the issue on my end.
my suggestion, VDM and repeat brush. my own workflow is useing VDM for every model now. also more stamps in general both for modeling and paint. VDM can often get you effects that are a pain in the butt to do in any other way because you can do undercuts etc. at some point i realized that VDM is better then falloff in a lot of ways. the other nice thing is that you can build a library which speeds things up.
Would love to hear more on this...
I find myself swapping between Zbrush and Mudbox a lot. I prefer the way Zbrush displays models in the viewport (maybe down to the variety of matcaps out there). I agree that Mudbox's ability to paint based on UVs rather than polys/normals makes it better than Zbrush in that regard. Dynamesh/Zspheres though, it can be very useful for sketching/concepting, which I do like to do in 3D sometimes.
With performance, does Zbrush handle larger polycounts better because of how it handles subtools (seperating high poly meshes)? My dense (separated) meshes always seemed to lag more in Mudbox(2012).
it is not a workflow issue. we will just have to agree to disagree there.
i think i even see the spotty displacement in your image right there.
anyways, i never ever had to use VDM for majority of my sculpting and i probably sculpt a new character every other month. i am sure as more people use 2014 release they will report this issue. it seems too obvious to overlook for anyone who does any sort of detail sculpting.
also, interestingly enough in 2014 you will get that spottyness even with a VDM. i just tried it with provided VDMs and spacing of 0.2 and 0
@ Fomori - sorry but dont wanna get into a vs debate here.
if you have any specific mudbox questions let me know.
try the 'repeat brush' and check your messages.
ive tried it... and its working fine here...
Sorry, I had to laugh a bit (in a friendly way) when I read this and then scrolled down to see your test case. As an experienced modeler, you realize of course that you're challenging the retopo algorithm with one of the most sensitive topology scenarios possible: a low resolution face. The lower the resolution of your target, the more perfect every vertex and edge connection has to be. And of course especially with a functioning human face, there is such a low threshold for error. So yeah, Mashru, Mudbox's computed retopology sucks for super low resolution human faces right now, I agree.
Set the stamp spacing to a negative number. The reason you're seeing dots now (vs 2011) is that the stroke profile is now much more accurate to the falloff curve. So you can finally begin achieving some of the actual sharpness that your falloff defines. But I'm with Gray here: I use VDM stamps all the time for this kind of stuff. Oh, and Mashru, in your VDM test I suspect you forgot to change the falloff back to the default falloff. As you know, the falloff multiplies with the stamp image to produce the final effect -- so using a VDM stamp isn't going to change anything if your falloff is still tight.
you know i thought about it but i did not try it, i got distracted.
just goes to show that just because you have been using a tool a long time does not mean you know it all. and nope, no explanation what i'm talking about. i think you can figure that out...
2 points @atac :thumbup:
you can create a lowpoly base in topogun and transfer subd levels from topogun to mudbox.
if i am going to retopo inside mudbox and i am not even able to define the topology 100% to my needs and have to stick to a dense starting base mesh then it becomes useless in that respect.
as for stamping, that might work as a around but it makes the stamp overlap and that means the falloff accuracy doesnt work anymore. it also makes the transition very strong.
with negative spacing, sharp falloff is no longer that sharp either . to eliminate the spots completely i down to -5
comparison
i guess you cant ask for everything...
with the resurf curves there are 2 different options you can 'retopologize' or you can 'create mesh from curves'. this second option from the tests that i have been doing is really quite nice and will probably be much more effective if your working on a low res cage. it creates a one to one match for your curve net. and if you dig into the options for the tool it has all the controls for merge distance etc to do the job well enough that i would consider it polished and usable. i have been able to make nets with no holes or error regions with not much fuss. the gen time is also quite fast. fast enough to do a lot of iterations and curve adjustments without a coffee break.
once you have your new base you can transfer your highres sculpt onto it.