I am working on some hard objects and trying to model them by using non-destructive workflows as much as possible. Till now, I have been successful at that, even if I had to redesign a few areas to ensure that I can keep working non-destructively. But now, I believe I am done with modelling and am starting on texturing the model. At this point, I feel like applying some of the modifiers (especially slices) would allow me to better apply my materials exactly where I want. Moreover, I had been mostly unable to apply bevels to edges which were created as part of a boolean operation as I had no way of controlling the amount of bevel to apply on them.
This has put me in a fix. On one hand, applying all my modifiers would open up a lot of ways I can make some final tweaks to my models. But at the same time, I will lose all ways of modifying some parameters of my applied modifiers and this is giving me anxiety.
How do folks with similar workflows handles this? Do you ever apply your modifiers? If not, do you trade for the inconvenience and look at other ways to achieve the same result?
As a note, I was considering creating a variant of my model. This makes me feel like it would be better if I don’t apply my modifiers. One way I have read is to backup the model. Now if I backup my model before applying my modifiers and then tweak my model after applying the modifiers, then the backup is no longer the same and whatever I do from the backup will not have the manual tweaks.
Well, I have rambled enough. I wonder what you folks think about this?
Hi ! you’re describing procedural modeling, which is only partly possible in Blender until we get “edit nodes” (let’s see if the development team goes further with this, hope so). What I do is exactly what you said, I backup my models. If you want to do more serious procedural hard surface modeling, you can try Zbrush or Houdini, both work very differently.
I don’t do a lot of hard-surface so what I do may only partially apply to you. You can use Houdini apprentice for procedural modeling, it’s free. Another good solution is Modo with meshfusion : it’s mesh-based, procedural, and it’s very good at booleans and bevels among other things, but it’s costlier than Zbrush last I checked. Blender isn’t going to get suddenly good at procedural modeling (beyond what you can do with geonodes right now) so I would turn to one of these instead.
As long as you keep a save of file where modifiers are not applied, there is no reason to feel anxiety.
Incremental saves are allowing to recover previous step of your work.
Parts of model can be copied, separated, joined to a new version of model.
There are still hundreds of operators available in edit mode that you can rely on.
So, you don’t loose all ways of modifying the result.
You will just lose some parametric ways to do modify.
Probably some can be converted to shapekeys.
Probably some modifiers can be re-added to result after the apply.
Some relocation of datablocks can be scripted.
So, there is no reason to feel anxiety because of destructive workflow.
That may be less comfortable, slow down your progression, require more time achievement.
But unless that was crucial for the job to keep things parametric because of amplitude of project ; there is no reason to doubt, that you will be able to achieve desired result.
That being said, you should be able, through geometry nodes to make a selection of edges, created by a boolean operation done, using Boolean node and its Intersecting Edges socket output.
That is a recent addition of 3.3 release.
Hi, I do as much work as I can with the non-destructed version of the mesh, including putting UV seams, applying scale etc., and then create a separate file where I apply all the modifiers I have to, make all the necessary adjustments and lay out the UVs. If an object needs extensive adjustments I just take the non destructed version and do the adjustments. This means sometimes I lose a bit of time I spent on the manual tweaks, but not too much.
It’s not a perfect workflow but it works, sort of.
It really is a stretch to say I model some kinds of hard surface. This is literally my first one and there will be more to come but I do have a goal for these.
Yeah, I realize that this is just a tad out of my comfort zone. I have seen lots of tutorials where people are like, “hmm, I can’t bevel this, welp, lets apply all the 10 booleans I am using and then try again” and it just gives me the heebie jeebies.
As for the project amplitude, right now, its a one man hobby project so there are no constraints other than money lol.
I go as long as I can without applying modifiers – if I feel that I have to before being done with tweaking, it means I should probably change my initial workflow. It’s all a constant learning experience and will remain this way for a long time, I expect. Doing things over to come up with a better way is part of that process. I also keep backups so I can go back to another stage.
I’m just a beginner, so I don’t have an overall workflow figured out yet, but since I come from programming I’ve been thinking that as soon as I’ll start on a serious scene, or a complex enough asset, I might try putting the whole shebang into a version control system. That’ll ease my fear of doing something not easily reversible.
Yeah, I am already doing that . Honestly can’t live without version control. So much so that sometimes I even put configuration files of some games under VC too!
. Honestly can’t live without version control. So much so that sometimes I even put configuration files of some games under VC too!