A comparisons between Cura + Slic3r

I’m in the process of printing the Deltic diesel model, which requires some parts to be printed more than once. I thought that’s a nice chance to compare Cura (17.10) and Slic3r (1.29). I’m using Slic3r normaly because it offers some functions cura don’t have like FW retract, bridging flow factor, fill patterns. But also Slic3r is not perfect, there are some anoing known bugs and some printing behavior is strange in my opinion.
Basicaly, I want to compare:
.) Creation of support material
.) Surface quality
.) Infill on small gaps

I’m using the same settings for both programs as much as possible. I will continue this thread when there are new findings, hoping I can safe other people some time when they want to choose a slicer.

Infill:
Cura does some not very efficient job on rectilinear infill, resulting in a incredible huge number of retracts. This is what Cura does:
Cura infill.JPG
It’s drawing unconnected single lines, so on each lines end, there is a retract. Slic3r is better in this situation, it tries to create a connected line path where possible:
Slic3r infill.JPG
This is also faster (safe of retract time) and possible stronger as it has more contact to perimeters.

I’m wondering if this is related to some of the frequently reported extruder jams! Some people (like me) never had one, other ones can’t get rid of them. A question of slicer software used (and the settings)?

Support material:
I compared the support on the Cam part of the engine. First thing I recognised: 20% infill in Slic3r is not 20% infill in Cura. I tried to get the same density for support and infill as in Slic3r and I ended up with 15% infill in Cura vs 20% in Slic3r. No problem at all, but there is a difference in calculation here.
As soon as the print starts, there was a noticable difference. Slic3r creates a smooth oval curve for the outline of the support. Cura creates a very fine zig-zag line, which results in slow motion due to jerk limitation. But to be honest, I know that if the contour of the part that needs support gets more complex, also Slic3r will start doing that. So, only a small + for Slic3r.

There is more difference when it comes to support material removal. Remeber, I selected the same values for both programs. In this case, I entered 0.1mm distance between support and part which works very well in Slic3r.
While Slic3r creates some solid layers on top of support, Cura doesn’t. Maybe thats the root cause of all the madness that comes now…
I wasn’t able to remove the support of the Cura part without trouble, also using a knife. The support seperartes into small peaces, whereas Slic3rs support comes of in nealy one piece because you can peel the solid layers from support from the real part. You can see the mess in the following pictures, I cleaned only a small part so it’s possible to compare the resulting part surface.

Here is a comparison of the resulting surfaces where the support was connected to. Left= Slic3r, Right= Cura. The spacing between the lines in the Slic3r part are due to my bridge flow setting of 0.8. With this, there are no loose lines and the surface is a little bit smoother. But of course, this results in not realy toughing lines. It’s a tradeoff where you have to make a compromise… I was surprised that also Cura seems to reduce the flow factor, because there are also gaps:


As you can see, Cura has severe problems with the perimeters, I thinks thats because they have no solid surface to be laid down. There are also some pieces of support materials visible in the basicaly cleaned area, they won’t come of. Maybe a little bit more z-distance would be useful here, but than there should be also even more problems with perimeters.

In this picture, you can see the perimeter is loose in the front right area of the Cura part. That’s a no go in my opinion, so this part is waste. But there is more to see here.

Here is a comparison between the top surfaces. Again, Cura on right side, Slic3r on the left. No real winner here, both looks nice :slight_smile:

That’s a realy interesting one. The upper one is Cura, as you can see on the not removed support. While the big mid-section is nice in both parts, the smaller pin side surface is realy strange looking in Cura! Looks like some layers are squished in qual distances, the only reason for this a can imagine is a bug in cura. Never had this with Slic3r. Second big minus for Cura…

Seam
There’s a last surprising difference between the two slicers, the visibility of the seam (where the ends of one perimeter comes together). For this situation, Cura does a realy great job! Without the red circle, I bet you can’t find the seam on the Cura part. For Slic3r, it’s clearly visible. I whish I would have some options to opitmize this in Slic3r…

Does somebody has an idea what’s the reason for the bad surface on the pin side in Cura? As for now, I will continue to use Slic3r, but maybe I give Cura a try on later releases. I realy love the invisible seams…

For some reasons, I can’t upload the picture of the seam comparison in my first post. Second try here:


Edit: The “squished” layers on the pin may be a result of too high nozzle pressure due to fast unretract (no FW retraction in Cura) or (more unlikely, because it’s same as in Slic3r) speed differences between infill an an outer perimeter! When I have a close look, I can see the “squishing” starts where the perimeter starts and is ending after 1/4 circle when the pressure has equaliced. It looks like it’s going all around, but that’s due to Cura rotates the seam point 90° on each new layer.

Quite the comprehensive comparison! Serves as a good reminder on what needs fixing in Cura :slight_smile:

Would be nice to see the print profiles used in these tests.

You can find my profiles in the attachment. The z-offset may be different in Cura profile, but it was the same for the test. I had to change it because I printed with PETG yesterday and I don’t change it for now back again.

One last thing:

Infill on small gaps:
The current Slic3r version has a severe bug regarding infill of small gaps. It leads to huge overfill, I tested this yesterday during the print of some cookie-cutters. The Slic3r print ended up in a mess of burnt PETG around my nozzle and in the part itself because the PETG can’t flow into a already filled gap.
Cura managed the parts without problems. Therefore, until alexrj can fix this, I will use Cura for thin parts :slight_smile:
Slic3r_config_bundle.ini (7.33 KB)
Cura.ini (11.2 KB)