In the 90s/early 2000s, did softwares like Mirai or Maya have soft selection during that time period? I'm working on a Leviathan Axe from God of War. I made one in Maya but I wasn't all that satisfied, so I went to blender, something I'm much more well versed in. Something I struggle with for this shape though, is making sure that the dragon has proper depth. Moving each vertex one by one to get the proper shape seems very excessive and confusing, and could potentially get me the wrong shape. I'm curious as if someone were to make this in the early 2000s (Yes I know the methods are the same from then as to now) if they would have to get the depth right the first time before adding in the peaks and valleys of the dragon, or if they would be able to use something like soft selection or proportional editing, whatever you decide to call it.
I'm pretty sure the first version of Max had something pretty close to the soft selection you see now and freeform deformation tools were present in 3dstudio R4 and Imagine so the idea certainly existed well before 2000
Mirai did have soft selection, yes. You had to define its radius by selecting elements on the mesh and it was pretty fancy from what I remember. Surface- or volume-based depending on your choice, lockable too - I think. 3ds max gained these features over time. Maya had its soft selection tools revamped at some point, not sure what it was like in the early 2000's.
Looking at these vids now - one thing that the Nichimen tools had long before Max or Maya was preselection highlighting as the default. Seems I don't have that in Blender natively to this day.
In the 90s/early 2000s, did softwares like Mirai or Maya have soft selection during that time period? I'm working on a Leviathan Axe from God of War. I made one in Maya but I wasn't all that satisfied, so I went to blender, something I'm much more well versed in. Something I struggle with for this shape though, is making sure that the dragon has proper depth. Moving each vertex one by one to get the proper shape seems very excessive and confusing, and could potentially get me the wrong shape. I'm curious as if someone were to make this in the early 2000s (Yes I know the methods are the same from then as to now) if they would have to get the depth right the first time before adding in the peaks and valleys of the dragon, or if they would be able to use something like soft selection or proportional editing, whatever you decide to call it.
Sorry if question is confusing
Leaving aside the fact that no one would put this many polys into an axe handle in the early 2000's, sometimes it is/was one vertex at a time - This is why one of the top best practices is to build a shape with the least possible detail, then iterate over it. When to increase subdivs, when to add support edges, understanding how the final shape will look before you get there - these are all things you sort of have to learn by going through the process over and over again.
Leaving aside the fact that no one would put this many polys into an axe handle in the early 2000's, sometimes it is/was one vertex at a time - This is why one of the top best practices is to build a shape with the least possible detail, then iterate over it. When to increase subdivs, when to add support edges, understanding how the final shape will look before you get there - these are all things you sort of have to learn by going through the process over and over again.
That is what I tried with this attempt:
I tried a box modeling attempt, but I am still not all that satisfied with how it came out. It's difficult trying to get the silhouette.
5rettski said: That is what I tried with this attempt:
I tried a box modeling attempt, but I am still not all that satisfied with how it came out. It's difficult trying to get the silhouette.
Its a good attempt. I guarantee you that after you've made 20 things like it you will understand the process much better and how to achieve the results you want. There are no shortcuts - you've just gotta do the time.
5rettski said: That is what I tried with this attempt:
I tried a box modeling attempt, but I am still not all that satisfied with how it came out. It's difficult trying to get the silhouette.
Its a good attempt. I guarantee you that after you've made 20 things like it you will understand the process much better and how to achieve the results you want. There are no shortcuts - you've just gotta do the time.
I have made many things like this, but for some reason I struggle on this I'm sure I'll figure it out eventually.
5rettski said: I have made many things like this, but for some reason I struggle on this I'm sure I'll figure it out eventually.
Make a thread in Showcase & Critiques. I know its frustrating to feel a gap between imagination and execution, but I promise it shrinks the more you do, and the more you fail. Failing is good - just try not to make the same mistakes over and over again What I will say is that in a production setting this axe handle was probably just sculpted, then decimated and retopologized for the game mesh. Learning topology like this is really great and will serve you well, but production is all about hitting the required quality bar in the least amount of time, and fiddling with edge loops on detail that could easily be baked into normal and ambient occlusion is a waste of time.
Make a thread in Showcase & Critiques. I know its frustrating to feel a gap between imagination and execution, but I promise it shrinks the more you do, and the more you fail. Failing is good - just try not to make the same mistakes over and over again
I know that, but I still struggled with this one more than other things I've made for some odd reason.
What I will say is that in a production setting this axe handle was probably just sculpted, then decimated and retopologized for the game mesh. Learning topology like this is really great and will serve you well, but production is all about hitting the required quality bar in the least amount of time, and fiddling with edge loops on detail that could easily be baked into normal and ambient occlusion is a waste of time.
I also know that, but I was more curious on how hard surface things with sharp edges like this were sculpted. Additionally, I was trying to get a feel if someone had to make this in the early 2000s, which is why I box modelled, edge modeled, whatever you want to do.
About this new handle : IMHO it's really just a matter of looking at the reference in a more investigative manner. "How would this be built in the real world ?"
Obviously this would be achieved by carving details over a smooth wooden handle. Meaning that the surfaces that are not carved in will have continuity since they originate from the shape of the raw handle ; and the deep carving work will follow the outlines of sharp motifs. Attempting to replicate this as subdiv can be a fun exercise of course, but this would be of no interest to your lead or AD if this was made in the context of actual production (regardless of it happening in the 20XX as opposed to 2025). They just want to be sure that you understand the object you are representing. And if you don't it would be their job to tell you what you're missing of course, that's part of their duties - especially when dealing with Junior hires.
So, the important part is to understand the object. But of course this comes with experience (hence time), but also from looking at historical artefacts in museums, or even in this case here by grabbing a wooden rod (or a potato) and doing some carving yourself to get a feel of the technique and the look that comes from it. As a matter of fact a dedicated concept artist would sometimes do just that in order to produce a believable design.
You can identify the surface continuity of the original potato thanks to the dark skin that still shows. Now with this analog in mind I think you'll see what's off about your 3D re-interpretation of this carved handle. And once you see that you'll be able to bend any approach to make it work for the task. It really, really doesn't matter if you "boxmodel" or "edgemodel" or whatevermodel.
For instance here you could take your model in its current state, select the faces that are supposed to be from the original surface, and shrinkwrap/project them back onto a temporary proxy of the raw handle shape. Or as suggested for the other handle, you could just take a few minutes to sculpt a very rough version of it and then snap your model to that. There is (and always have been) plenty of ways to get there.
Replies
I'm working on a Leviathan Axe from God of War. I made one in Maya but I wasn't all that satisfied, so I went to blender, something I'm much more well versed in. Something I struggle with for this shape though, is making sure that the dragon has proper depth. Moving each vertex one by one to get the proper shape seems very excessive and confusing, and could potentially get me the wrong shape. I'm curious as if someone were to make this in the early 2000s (Yes I know the methods are the same from then as to now) if they would have to get the depth right the first time before adding in the peaks and valleys of the dragon, or if they would be able to use something like soft selection or proportional editing, whatever you decide to call it.
Sorry if question is confusing
I tried a box modeling attempt, but I am still not all that satisfied with how it came out. It's difficult trying to get the silhouette.
What I will say is that in a production setting this axe handle was probably just sculpted, then decimated and retopologized for the game mesh. Learning topology like this is really great and will serve you well, but production is all about hitting the required quality bar in the least amount of time, and fiddling with edge loops on detail that could easily be baked into normal and ambient occlusion is a waste of time.
I also know that, but I was more curious on how hard surface things with sharp edges like this were sculpted. Additionally, I was trying to get a feel if someone had to make this in the early 2000s, which is why I box modelled, edge modeled, whatever you want to do.
So ... here's some potato carving :
For instance here you could take your model in its current state, select the faces that are supposed to be from the original surface, and shrinkwrap/project them back onto a temporary proxy of the raw handle shape. Or as suggested for the other handle, you could just take a few minutes to sculpt a very rough version of it and then snap your model to that. There is (and always have been) plenty of ways to get there.
Good luck !
[edited for clarity]