Make Slicer More efficient---Food for thought

Hi,
I don't know whether this is done in contemporary software's or not. The idea is is this possible to make a program that can take care of minute details with spitting out Gcode for low layer height/width and for maximum nozzle layer for the areas where no details needed. Sorry for bad English hope u understand.
When we do laser scanning we use small calibration panel to catch the details and then the big calibration panel is used where resolution/details is not required but bigger scan is required.Then both are combine together.
Can we make the slicer software more intelligent to slice the CAD STL and Gcode it in the above way?
Regards

Khalid,
I understand what you are describing - variable layer height to improve resolution where needed, and thicker layers to improve speed other places.

Your idea would be very nice for the top/bottom of curved objects, and the sides would not need the resolution and could print faster.

Slic3r does this a little, by allowing multiple layer thicknesses to be printed in the infill area at once, but you are talking about perimeters.

There would be many tricky parts in programming this
- Recognizing where to do it and where not to in a complex object
- Remembering where these evolving regions are so that the following layers continue the method, and do not conflict.
- Handling the transitions and boundaries
- ...

I think it could be done by an advanced programming team, but it could easily double the complexity of the slicer, which is more than a slight barrier.

---

My printer: Raptosaur - Large Format Delta - [www.paulwanamaker.wordpress.com]
Can you answer questions about Calibration, Printing issues, Mechanics? Write it up and improve the Wiki!

you mean like what skeinforge does? [fabmetheus.crsndoo.com]

@ntoff,
That is an interesting plugin. I like the idea that the outside wall will have smaller vertical layers and so will be smoother.

It doesn't actually improve resolution however. What he is doing with the plugin is taking an existing perimeter and printing two (or more) half-height perimeters stacked exactly on top of each other (or alternately thinner perimeters next to each other), using the original outside dimensions (note: this requires printing the outside perimeter first). So this will not smooth angled or curved surfaces, nor give finer horizontal details, but it can print the inside areas faster - using a larger layer height.

A smoother vertical sidewall would be nice tho!

Thanks for posting the link.

This has been done before, long time ago..

You effectively define two extruders...
Perimeters with low Z height via e0 and internal e1 with large Z height

Then just print threw the one extruder.... (it was a long time ago, suspect gcode post processing was used)

the issue is this cross section.
____YY[ XXX ]
__ YY_[ XXX ]
_YY___[ XXX ]

Y perimeter
_ Unavoidable gap
X solid full height infill..

Things get much more complicated if perimeters are not a nice multiple of infill...

Edited 2 time(s). Last edit at 10/14/2015 06:43AM by Dust.

YES - Skeinforge Craft - Skin!
For example use a layer height of 0.2mm but skin is set to 0.1mm!

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse