Infill missing, preview doesn't reflect this #2474

Closed
fpiesik opened this Issue Jan 1, 2015 · 15 comments

Projects

None yet

3 participants

@fpiesik
fpiesik commented Jan 1, 2015

Slic3r version: 1.2.4
OS: Ubuntu 12.04 64Bit

Expected result: Infill should be everywhere, gcode-preview should reflect the actual gcode.
Actual result: A part of the infill is missing in the actual gcode, but not in the Slic3r-preview.

Slic3r preview:
image

actual gcode (displayed in octoprint):
image

files (didn't found a way to attach to this issue):
https://www.dropbox.com/sh/qc9rtwu8h7oxyep/AACV_R1dOt3Qy7JVWS1N_mmsa?dl=0

@a4jp-com
a4jp-com commented Jan 2, 2015

How are you trying to print this? Do you have supports turned on?

screen shot 01-02-15 at 01 19 pm

@fpiesik
fpiesik commented Jan 2, 2015

it's rotated 90 degrees along the x-axis

@a4jp-com
a4jp-com commented Jan 2, 2015

I rotated it as you said.

2015-01-02nd

Maybe you just need to increase the infill percent or lower the size of perimeters or your nozzle setting. I can't see any problems. Is your nozzle really 0.8mm? That seems big.

@fpiesik fpiesik changed the title from Infill missing, preview doesn't show to Infill missing, preview doesn't reflect this Jan 2, 2015
@fpiesik
fpiesik commented Jan 2, 2015

The thing is, that the slic3r preview doesn't show the missing infill. look at my screenshots, in the slic3r preview the infill is there, but in the actual gcode, there is some infill missing. Both screenshots show the second layer.
Yes, I actually use a 0.8mm nozzle as it prints much faster and the resolution is still good enough for most of my print jobs.

@a4jp-com
a4jp-com commented Jan 2, 2015

Maybe Octoprint has the problem. Have you tried printing using another program? I would also have at least 2 perimeters if possible. Go into the advance settings for the time being and get some more. See if that makes a difference. It might not fix Slic3r but it might fix the G-code for printing.

@fpiesik
fpiesik commented Jan 3, 2015

nope, it's not a problem with octoprint. here is a screenshot of the second layer made with pronterface:
image

With a 0.8mm nozzle there is no need for two perimeters, that speeds up the printing process significantly. I have made many good and fast prints this way.

actually, the infill is only incorrect in a few layers which is not a very big problem in this case.
I think the main issue here is the indifference between the slic3rs gcode preview and the actual gcode.

@a4jp-com
a4jp-com commented Jan 3, 2015

That's okay but please change the settings to see if there is a bug.

@fpiesik
fpiesik commented Jan 3, 2015

sorry, but I don't get the point of changing the paras to something i don't want. I have used many different parameters and never noticed this beahviour before. I provided everything needed to prove that under THIS circumstances there IS a bug. the gcode differs from what the preview shows and from what it should be...

@a4jp-com
a4jp-com commented Jan 3, 2015

Okay. If you can't try then I can't help you. Sorry. I'm just trying to help in my own free time. I'm not the programmer that made this program.

@alexrj
Owner
alexrj commented Jan 3, 2015

@fpiesik, good catch there. I'm able to reproduce this issue, and the problem is not in the viewer since the missing segments are actually missing from G-code.

@alexrj alexrj added this to the 1.2.4 milestone Jan 3, 2015
@alexrj
Owner
alexrj commented Jan 3, 2015

Ouch, I'm suddenly unable to reproduce this anymore... :(
It can't be reproduced by saving the rotated STL file, but if I follow the entire procedure (load + rotate + export) I can reproduce it.

@fpiesik
fpiesik commented Jan 3, 2015

ah, good to hear that and how it can (not) be reproduced:)

@alexrj alexrj added a commit that referenced this issue Jan 3, 2015
@alexrj Bugfix: Douglas-Peucker used perpendicular distance instead of shorte…
…st distance, thus clipping more than it should. #2474
d8be67c
@alexrj
Owner
alexrj commented Jan 3, 2015

Very tricky issue, combining two distinct minor bugs that I managed to isolate into regression tests. One of them is already fixed.

@alexrj alexrj modified the milestone: 1.2.4, 1.2.5 Jan 4, 2015
@alexrj alexrj modified the milestone: 1.2.5, 1.2.6 Jan 9, 2015
@alexrj alexrj modified the milestone: 1.2.6, 1.2.7 Jan 19, 2015
@alexrj
Owner
alexrj commented May 22, 2015

Fixed!

@alexrj alexrj closed this May 22, 2015
@alexrj alexrj added Fixed and removed Verified bug labels May 22, 2015
@fpiesik
fpiesik commented May 23, 2015

Great!
Am 22.05.2015 14:26 schrieb "Alessandro Ranellucci" <
notifications@github.com>:

Fixed!


Reply to this email directly or view it on GitHub
#2474 (comment).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment