"No layers were detected" what?

So, I must be doing something fundamentally wrong here.

I've downloaded Slic3r v0.9.3 for windows and tried to generate gcode from an OBJ exported from Maya, an OBJ from Blender and an STL downloaded from thingiverse, but each time I press the "Export G-code"-button all I get is this error:

"No layers were detected. You might want to repair your STL file and retry."

I also tried Slic3r 0.8.4 on linux with the exact same error!

What am I doing wrong?

The file is bad and needs to be repaired. Try cloud.Netfabb.com

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I'm getting the same error on good geometry. Trying to print out a spool holder, Slic3r will process the left arm but not the right.

The right arm is the EXACT same geometry, mirrored from the left arm file.

Dirty Steve Wrote:
-------------------------------------------------------
> I'm getting the same error on good geometry.
> Trying to print out a spool holder, Slic3r will
> process the left arm but not the right.
>
> The right arm is the EXACT same geometry, mirrored
> from the left arm file.

Depending on what you used to mirror it you probably ended up with the mirrored one being inside out. A lot of programs will invert all the faces when you make a mirror copy. Try running the mirrored one through Netfabb and see if it solves the problem.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

It does slice in 0.7.2b. Using 3ds Max 2010, tried output from Rhino, same result. Will try a few other things.

Dirty Steve Wrote:
-------------------------------------------------------
> It does slice in 0.7.2b. Using 3ds Max 2010, tried
> output from Rhino, same result. Will try a few
> other things.

If you view it in something like Meshlab where it shows correctly oriented faces as gray and inside out ones as black you will be able to tell if it is inside out. But from my experience it is just easier to use cloud.netfabb to fix it.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Re-imported into Max, .stl faces are getting flipped on export, but display as out-facing in Max scene file. Only shows up on re-import into Max. Import into Rhino displayed outfacing normals. Looks like export-import-flip-export-fix.

Edited 2 time(s). Last edit at 10/01/2012 08:59PM by Dirty Steve.

Could someone then please walk me through exporting a model from Maya (Or blender for that matter!)

Because making a polygon cube and then scaling it up and exporting it as an obj is clearly the wrong way to do it. And yes, the normals of the model are correct.

It may be a flaw in the OBJ importer of Slic3r. I don't actually know of anyone that uses OBJ files, try exporting it as an STL and see if that fixes it. If it is the problem you can report the bug in the Slic3r repo and maybe they can fix it.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Rhino should be able to show the faces the right way.

[www.rhino3d.com]

I love Rhino3d and RhinoCam

Sublime Wrote:
-------------------------------------------------------
> It may be a flaw in the OBJ importer of Slic3r. I
> don't actually know of anyone that uses OBJ files,
> try exporting it as an STL and see if that fixes
> it. If it is the problem you can report the bug in
> the Slic3r repo and maybe they can fix it.

Thank you very much - that was it! ... In the world I come from OBJ is the file format you use when everything else fails. Perhaps it should not be listed as a supported format in slic3r.

I have the same problem with a STL file. I made the file to demostrate another problem with toplayer being very thin and cracky if the top is sloped a bit, but it turned out I could not even load it. It works fine with other slicers.

It's the file called slope_stl that will not load. The other two files is to demostrate that holes is too small. The round model is suppose to fit into the one with the hole.

TKL Wrote:
-------------------------------------------------------
> I have the same problem with a STL file. I made
> the file to demostrate another problem with
> toplayer being very thin and cracky if the top is
> sloped a bit, but it turned out I could not even
> load it. It works fine with other slicers.
>
> It's the file called slope_stl that will not load.
> The other two files is to demostrate that holes is
> too small. The round model is suppose to fit into
> the one with the hole.

Depending on the version of Slic3r you used the two parts not fitting together is most likely a result of a printer calibration issue. Try Slic3r 0.7.2b and then adjust your flow (extrusion multiplier) until they fit together.

For the slopped top have you tried increasing your perimeter count and solid layer count?

For the other one have you tried repairing it via netfabb?

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Sublime Wrote:

> Depending on the version of Slic3r you used the
> two parts not fitting together is most likely a
> result of a printer calibration issue. Try Slic3r
> 0.7.2b and then adjust your flow (extrusion
> multiplier) until they fit together.

The printer is calibrated. Did you try printing the STL's?

>
> For the slopped top have you tried increasing your
> perimeter count and solid layer count?

> For the other one have you tried repairing it via
> netfabb?

I am sure there are hacks to make things work. I could just use some more solid layers, but that isn't changing the fact that there is a problem. The STL is good so why should I repair it?

TKL Wrote:
-------------------------------------------------------
> Sublime Wrote:
>
> > Depending on the version of Slic3r you used the
> > two parts not fitting together is most likely a
> > result of a printer calibration issue. Try
> Slic3r
> > 0.7.2b and then adjust your flow (extrusion
> > multiplier) until they fit together.
>
> The printer is calibrated.

If they do not fit together with version 0.7.2b it is not calibrated.

Did you try printing
> the STL's?

No I don't download rar files but I have designed hundreds of things and a couple of complete machines and I use Slic3r 0.7.2b and when printing with a Tantillus properly calibrated I can get away with 0.1mm clearance on slip together parts. It is normal to model a small amount of clearance for parts unless you want a press fit that may require a press to assemble. For the average machine 0.2 - 0.5mm clearance on parts is normal.

>
> >
> > For the slopped top have you tried increasing
> your
> > perimeter count and solid layer count?
>
> > For the other one have you tried repairing it
> via
> > netfabb?
>
> I am sure there are hacks to make things work. I
> could just use some more solid layers, but that
> isn't changing the fact that there is a problem.

They are not hacks. They are settings that allow you to make perfect prints. That is why they are there and should be used.

> The STL is good so why should I repair it?

How do you know it is good? I have never heard of a single program that produces perfect STL files every time (unless it is manually made in a mesh modeller by someone that understands manifoldness). Try running it through cloud.netfabb and click the view button and see if it found any errors. Or upload the file here as a binary (smaller than ascii) without the rar and I will have a look.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I haven't tried Slic3r 0.7.2b. I use the latest version. I attached the files, no rar.

Slope.stl slices fine after running it through cloud.netfabb.com and the other two will almost never fit since you have not given them any clearance. The circle is exactly 16mm and so is the opening in the cube.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I just printed the slope after running it through cloud.netfabb.com and it really doesn't show the problem I was trying to demostrate. I did not see cloud.netfabb.com reporting any problems but it increased the number of triangles.

As for the inside and outside model you mention clearance? I expect that the models will fit within a tollerence of +/- 1 layer. For me it looks like the layer is placed in the middle in the out-/in-lines /and not inside the model - actually the hole seems to be 2 layer width too small..

BTW I have not tried the older v.0.7.2. Was that just for calibration or is the quality of the prints better overall?

Edited 1 time(s). Last edit at 10/14/2012 05:01AM by TKL.

They completely rewrote the slicing algorithm to get a few milliseconds speed increase but lost the quality in the * series and never got it back. There were also other bugs in the newer releases that make them less usable in my opinion. Then they started trying to make it smarter and compensate for hole shrinkage etc and their maths do not apply to all machine combinations and cause problems on machines that are unlike the ones they test on mostly. Then something happened really recently in the ( series where the layers do not line up correctly and make it look like you have backlash when you do not.

For the clearance, if you machined two parts like that on a milling machine out of steel to those exact dimensions you would have to press them together unless you gave it some clearance. On milled parts you only need a few micron for a slip together assembly but for FFF (FDM) printed parts you need to allow for the ridges on the surface. So on your model I would make the hole 16.1 - 16.5 depending on your layer alignment and flow calibration.

Also even though you have measured the distance the filament moves to calibrate the machine the maths are not perfect due to some assumptions that have to be made. This is why the extrusion multiplier is there. Most likely you will have to reduce it until you can print a solid cube with a perfect top surface.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I already already adjust the filament used because I had some big boogers in the corners of printed models. The top is actualy indenting so I am rather using too little. I do not know how they calculate the filament flow. I think it should be based on the volume of the filament. But it would then have to be trimmed down because the round plastic worm would never be squeezed airless.

The clearence would have to more like 1mm in diameter in my case, at least with the latest version. I will try 0.7.2b and compare.

There's a good way to check whether a given STL file is good or not: use Netfabb Studio Basic. It's free. As soon as you load your STL file in it, Netfabb will display a large warning triangle sign should the file be invalid. Slic3r's algorithm is fully compliant with that check. So if you don't trust Slic3r's warnings, go check with Netfabb.

Sublime, were you referring to Slic3r in your last message? If so, I should correct you: the slicing algorithm was never rewritten with the mere goal of speed. It was rewritten to switch from a numerical approach to a topological approach, resulting in full compliance with Netfabb validity checks. Previously, a fair number of files were sliced incorrectly. Now Slic3r recognizes correctly all files that Netfabb (and other tools, of course) report as valid.
I should correct you also about hole compensation: it was removed quickly, so it lasted no more than a few weeks (and long time has passed).
Also, can you point me to the other bugs you're talking about?