Yeah, it would make sense that a model like the second and third can't be spiralized, but that's not the kind of model in my case. My model was created specifically with spiralization in mind. I am using a 1mm nozzle and my model's wall is exactly 1mm thick. In the settings, I've set wall count to 1. I insist that this implementation is wrong and needs some solution that works in vanilla Cura, because it is literally workflow-breaking and requires I move my work elsewhere to get it done. I ended up using orcaslicer for this model. All it took was checking "vase mode" and it sliced perfectly immediately. I genuinely do not understand why that cannot be the case for Cura. If the model isn't made for spiralize, then of course you can't expect it to work, and the people who do should have to deal with the troubleshooting until they realize that what they're trying to do isn't feasible. But if it is specifically built for this print method, there is no excuse for it not working. Again, I am begging you to either revert this to the way it was, or give us more parameters so that it can work as intended (like the "start spiralize at layer" option I suggested for example).
Edited by michaeltzan5 hours ago, michaeltzan said:I genuinely do not understand why that cannot be the case for Cura. If the model isn't made for spiralize, then of course you can't expect it to work, and the people who do should have to deal with the troubleshooting until they realize that what they're trying to do isn't feasible. But if it is specifically built for this print method, there is no excuse for it not working. Again, I am begging you to either revert this to the way it was, or give us more parameters so that it can work as intended (like the "start spiralize at layer" option I suggested for example).
Programming rule #16: no matter how hard you think something will be, it will be harder than that.
Orca has a completely separate development history to Cura (it's based on Bambu Studio, which is based on PrusaSlicer, which IIRC was based on the original Slic3r). Cura introduced a new slicing engine (Arachne) with 5.0 and while generally it's better than the old one (if the people who disagree with me could form an orderly line, I'll get to you one at a time) it's still not perfect, but it's usually edge cases like this where you'll get problems.
I never said it would be easy. What I am saying is that it is necessary. I don't know nearly enough about slicing engines to be able to tell if the transition to arachne would matter in this issue, but I do know that the other slicers (orca in this case) use arachne as well and, again, have no trouble figuring out how to slice a model as simple as that in vase mode. What is dumbfounding to me is that this issue has been reported for the last two years at least, and the reply is always the same from the support: "it's the way it works". Then change it! What kind of stupid excuse is that? I'm sorry for being pissy but this is genuinely pissing me off because now I have to go and set up all the print settings again for each of the products I'm trying to print in another slicer.
9 hours ago, michaeltzan said:but I do know that the other slicers (orca in this case) use arachne as well
Not that simple. Neither just pass the STL straight to Arachne, they both process it first and that part is unique to each slicer.
9 hours ago, michaeltzan said:I never said it would be easy. What I am saying is that it is necessary.
This is a fairly edge use case which very few people will encounter. The Cura team is stretched as it is and things which will affect a lot more people are considered more important. Even if only 1% of the people who have this problem post on the forum asking about it (you're not the first) the whole number represents a very, very small percentage of the total Cura user base.
9 hours ago, michaeltzan said:I'm sorry for being pissy but this is genuinely pissing me off
I fully understand. There's times when I just can't get something to slice properly in Cura but I usually end up beating it into submission in Cura (brute force and occasionally writing a post-processing script just for that particular thing) but I have a lot more free time than most people.
I understand, that makes sense. Well, I really hope some better solution comes up eventually!
Recommended Posts
GregValiant 1,454
I always use the SmartAvionics fork of Cura for spiralizing models. His build is based on UM Cura 4.x.
Sometimes you just need to grab a different tool.
This is with MB 4.20.25 and there is no Cutting Mesh required for the hole.
You can't look at just one model to determine how it should be properly spiralized. There were always rules that had to be obeyed. Your model has a defined "Wall Thickness" of that outer cylinder. The earlier versions of Cura ignored that fact. In my (never humble) opinion it was wrong in doing so, and the current implementation is correct.
This is the model in 5.9.0 and yes, it is a mess and not at all what I want. But it is a true and accurate slice of the model. It is not ignoring the fact that the outer cylinder is 3mm thick.
This model cannot be spiralized and yet a lot of people have put in bug reports because it can't be.
This sort of model will always generate travel moves (and consequent Z-seams) as the nozzle moves from the outer feature to the inner.
Link to post
Share on other sites