possibly the best box modeling ever ?
vertex
I thought I had posted this about 9 hours ago, but it disappeared and there's no draft. So I'm re-writing it.
Basically I was blown away by this box modeling process right here:
https://www.youtube.com/watch?v=ac4qV2uIF3Q
What struck me is that in my novice ways, I typically try to avoid EVER creating an ngon EVER with the exception of a spare triangle here and there. What struck me are the following:
1. This artist seems to create quite a few ngons and triangles while modeling for ease of flow. But then later refines them and they all eventually turn into quads ( almost all of them ) . The artist does a lot of poly cutting with a knife which is another thing I've tried to avoid, but the whole process of doing local cuts and adding edge loops later seems to work really well here.
2. The first 15 seconds of the video are absolutely amazing to me. The way he/she blocks things out and then pulls edges out makes a really great base shape within a really short period of time.
3. The end result of course is phenomenal. But what I love about this process is that it can also be used to create low poly models as well, you would just quit before adding too much detail at the end.
My pet peeve with a lot of box modeling head tutorials is that the model looks like crap until you sub divide it. I'm modeling for games so I really need a facial structure that can stand alone without subdivision. This process seems to really be one of the best Iv'e seen. Any thoughts or advice about this particular style of head box modeling or the whole issue of creating ngons and later refining them into quads? Haven't seen it done this way before.
Basically I was blown away by this box modeling process right here:
https://www.youtube.com/watch?v=ac4qV2uIF3QWhat struck me is that in my novice ways, I typically try to avoid EVER creating an ngon EVER with the exception of a spare triangle here and there. What struck me are the following:
1. This artist seems to create quite a few ngons and triangles while modeling for ease of flow. But then later refines them and they all eventually turn into quads ( almost all of them ) . The artist does a lot of poly cutting with a knife which is another thing I've tried to avoid, but the whole process of doing local cuts and adding edge loops later seems to work really well here.
2. The first 15 seconds of the video are absolutely amazing to me. The way he/she blocks things out and then pulls edges out makes a really great base shape within a really short period of time.
3. The end result of course is phenomenal. But what I love about this process is that it can also be used to create low poly models as well, you would just quit before adding too much detail at the end.
My pet peeve with a lot of box modeling head tutorials is that the model looks like crap until you sub divide it. I'm modeling for games so I really need a facial structure that can stand alone without subdivision. This process seems to really be one of the best Iv'e seen. Any thoughts or advice about this particular style of head box modeling or the whole issue of creating ngons and later refining them into quads? Haven't seen it done this way before.
Replies
It's actually less about the technical workflow, and much more about having a very good grasp of anatomy and facial structures.
It has its pros and cons but there's nothing new about it. The assumption of "not using ngons" only comes from yourself
So maybe you can help me slay another dragon of ignorance... It's my impression that , with games, sculpting is best suited for high poly ( in other words, the best results are making a low poly model as a base, then doing sculpting on top to make a high res version for a normal map, then bake that onto the original model ) .
But that thinking might be outdated. Are there people using ZBrush and Sculptris and other sculpting programs to make relatively low poly models for games ?
http://polycount.com/discussion/146695/polycount-to-professors-ngons-are-useful
Even though I'm aware both techniques lead to final result, would a studio care if someone preferred to box model over digitally sculpting? Even taking into account their art is amazing?
Even though you can retopo the mesh after reducing tris and bringing into Blender for rigging, my understanding is that it can be a headache making the mesh suitable for animation, especially with characters that move a lot. Is this correct ?
Are you a game artist ? What's your take on this and what is your flow ? If everyone is using sculpting for GAME characters then I think it would be worth it for me to learn sculpting for that reason alone. Please advise, I'm willing to listen to all points of view .
I should add that I'm making characters for mobile games using Unity. So these insanely high polycounts that you see in console games... I don't have that luxury.
You can retopologize it manually, putting you in complete control over the final topology.
I don't think you quite understand the retopo workflow if you think that it's risky. A part of the reason that people sculpt is because they get a good start model that they can retopo any way that they like afterwards. It's risk free and artist friendly. It's also faster to push art out that an art lead can give a yes/no on, than doing it through other modeling tools.
I wish I was a competent sculptor really, I dabble but I haven't produced anything meaningful and mostly just use sculpting for details on a basemesh.
Mudbox is only $10.00 a month and appears to have more tools than Sculptris. I like the idea of ZBrush but don't want to make that financial commitment just yet. I may just purchase a monthly license for Mudbox and see how far I can go with it. When I see sculpting time lapses, it does amaze me and it's good to know that this is the way that most artists are going now. Good info, thank you. I'm liking this forum.
Japan CGWorld magazine does an annual whitepaper on the Game/CGI industry. You can see from their 2017 survey result, Zbrush hasn't taken over Japan, yet.
Q: Which tools do you mainly use for your work? (green for primary tool, yellow for secondary, grey for the rest; the circle represent the type of work respondent do, 74.2% are working on games).
Q: which tool do you want to learn the next?
So I am quite certain there are many game companies in Japan don't rely on Zbrush.
(Of course Maya has sculpting nowadays, but it's a good indicator on why you might not need Zbrush...)
ArNavart
bitinn
WilliamVaughan
Thanks for all the great replies!
Box-modeling simple tube characters or really low poly models makes sense but, for realistic anatomy it's a step back in time. Modern pipelines have also adopted scanning and projection to pre-built, organised topologies using Zbrush or a specialised tool like wrap 3D.
Also, studios will have predetermined basemeshes that will be sculpted on with the organised topology being fixed.
The simple fact is that for VFX/cinematic/AAA characters/creatures, sculpting/retop/reprojection is themodern workflow.
This is also for iterative, non-linear pipeline puposes. A hero character might takes many months to finalise, and be handed up and down the pipe for revisions. Sculpting/retop/basemesh reproject allows this effortlessly. Nobody in that pipe is re-routing edgeflow using traditional box-modeling.
P.S: I hadn't seen that from Japan and am not surprised as they have been generally a bit behind in adopting the latest western tech/workflows. But, the future projection survey isn't surprising and within the next few years Zbrush will be top of that poll.
And thanks to everyone else who has contributed to this thread.