Yoda - because I can

[Polygonhell]

0 infill, 4 perimeters no support, I just ran it through Slic3r and printed.
Printed surprisingly well considering some of the overhangs, when you watch it print it's clear that support would be nice and at least some infill would correct some of the flaws.
I couldn't get Slic3r's support for this model to stick to the table, it's so small, and skeinforge generated extremely odd infill, so I printed it without.

[michaellatif]

That looks really good for no infill. I may try that on my next large print to see how it goes.

[Polygonhell]

The only places that really suffer for lak of infill are the very top of the head, and the top of the shoulders, and the inside of the ears, but I was stunned how well it came out considering
He even has nostrils, and without the infill it was a relatively "fast" print.
Certainly the biggest thing by volume I've printed.

[PartDaddy]

Oly and I were talking about the fill angle. My idea was for Slic3r to have an angle threshold to control step over. Oly's is to have a min / max threshold. It's pretty impressive what these guys are doing to the software. So they should be able to implement some type of threshold angle or min / max concept that will eliminate the shallow angle step over problem once and for all.

[ALeander]

Nice Print!

What flaws? for no support of infill that looks fantastic.

[PartDaddy]

Don't get me wrong. This is a good looking print. Better than I can do! Perhaps you can post your Slic3r settings? (step z, speed, etc)

Basically what I'm talking about is how the step over becomes too great as the angle gets shallower. Then gaps begin to appear. If the program detected a threshold angle, it could adjust step up and step over so no gaps appear. I think this limiting angle (or whatever you want to call it) would allow a more perfect print the first time.

Maybe you guys can give me some advice on my Slic3r settings. Look at this Skull picture from the top. As the Z continued to step up, and the part surface angle got shallow, the step over becomes too much.

top of skull

[Polygonhell]

About all you do AFAIK in Slic3r is increase the number of perimeters.
Skeinforge has a bit more control, but I don't pretend to understand the majority of Skeinforge settings

[Leeway]

Right.
That skull looks like about two perimeters. Maybe 4 or 5 would get better results as would changing the layer height to be smaller.

[PartDaddy]

Yes, the number of settings are impressive. And mind boggling. However, it's nice to have the control when you want it.

I picture a setting like the Z step for the X Y. Maximum XY step over. That's fairly simple, just don't ask me to write the algorithm.

In the mean time, I'll continue to add perimeters. I don't think I've done 4 yet.

[PartDaddy]

OK. Tomorrow, I'll print the skull again from the shop live. I'll bump up to 4 perimeters on the skull, add support, and lower my infill.

[bvandiepenbos]

Yes, the number of settings are impressive. And mind boggling. However, it's nice to have the control when you want it.

I picture a setting like the Z step for the X Y. Maximum XY step over. That's fairly simple, just don't ask me to write the algorithm.

In the mean time, I'll continue to add perimeters. I don't think I've done 4 yet.

Steve,
that sounds like a really good idea. Hope that function can be implemented .
~*Brian

[Dave_Sohlstrom]

I printed a horizontal half cylinder 40mm dia 40mm long with a 20mm hole in it. Slic3r .25mm layer 3 perimeters . I had the same problem with gaps as the angle reduced at the top of the hole and at the top of the print. I agree it would be great if we could set Slic3r so it reduces the step over as it reaches the top of the part.

[K.Gerard]

There are several recent bug fixes that may at least partially address this:
https://github.com/alexrj/Slic3r/commit ... 23d408bd13 (some nearly-thin walls were discarded.)
https://github.com/alexrj/Slic3r/commit ... 019daaa7b4 (New automatic detection of additional required perimeters for better quality with domes and sloping walls.)
https://github.com/alexrj/Slic3r/commit ... 14cd8317fe (Fixed regression causing poor adhesion of rectilinear infill to perimeters)

Except for regex I not much of a hand at Perl ( http://xkcd.com/224/ ) so I won't try to do a code trace; and I'm still not printing (no working controller, and waiting for a Printrboard to drive) so I can't test myself.

These fixes are all after 0.7.2b, so you'll either have to build it yourself to test it or with for the next release of an installation package.

Klee Gerard

[Polygonhell]

He releases pretty much monthly, so we're about due for another release.

I don't know if it's just what I've been printing lately, but I've been using Skeinforge more than Slic3r recently.
There aren't actually that many "important" settings, so once you have something working it's actually not bad to use. Having said that I still usually run a model through Slic3r and look at it before I resort to Skeinforge.

There are a few things still missing in Slic3r decent support, rafts, and to me most importantly the equivalent of the Skeinforge comb plug-in, this orders fills and head moves between sections so that the hot end doesn't run over already printed perimeters, which almost always results in a blob.

My printers dead until next week I managed to put my finger in the running 40mm fan on the print head and it tore off a blade :/ so maybe they'll be a new version of Slic3r to try when I get it running again.

I did look at some of the Slic3r code and he does have some concept of angle between layers, so limiting the step over should be possible, though it would likely have to be a complete pass over the sliced model. You should file it as a feature request on the github page.

[PartDaddy]

Because I was supposed to post a Yoda here...

yoda (again)

[@CiDG]

PartDaddy,

Did you ever reprint the skull? I have one that is printed beautifully (on someone else's reprap...). I haven't been able to replicate the quality on my H1. My prints look a lot like yours. I'd like to try changing some settings and wanted to know how the print looked if you gave it another shot.

Thanks

Gabe

[PartDaddy]

sorry for late reply. no, i didn't have time to print it again yet.