New Beta

[crocky]

For Windows 64 and 32 systems today..

You have to be registered to use it.

[McSlappy]

Great, what's the improvements and what are your thoughts so far?

[crocky]

I like it but I am waiting for a Mac version

And then a repetier Mac version that is up to date, the current one does not talk to the informer app

[crocky]

Jonathan Comments from Refugee Camp!

Hi, everyone. Here is the update with the loop-start improvements. Some things to mention:
On the seam location, I put in some code to be able to weight corners higher, either concave or convex or both. BUT
this was in place before jitter, so you only really saw the benefit if using 0 jitter
so I disabled it again, but
I plan to put it back in, with another setting that lets you control the importance of corners.
but I need some help deciding how to do this...should I have a random chance select a point based on the individual weighting or each point? This would be tough to make fast.
The starting point of each perimeter is now found (according to the specified angle), then jittered at a function of total path length instead of just angle (this helps elongated loops not always have the seam in the center).
The starting point of any loop that is not a perimeter is adjusted to be right next to the closest perimeter seam. (unless there isn't one close enough, then it keeps its own jittered value).
The random number for jitter is now calculated once per loop, so
each perimeter will have its own seam angle (within the start angle and jitter specified)
the seams will not be all the same direction on a given layer, but
previously I had some code to throw out any random numbers within 1/8th of the last start point (so basically excluding the quadrant of the last layer's seam). Since each loop gets its own random number now, that is no longer feasible. So the numbers are still nice and random (JKISS32, fyi), but you may get some seam points close to each other on back-to-back layers.
One other thing to note: a while back I made the code for determining inside / outside of some points much more robust, but this eats up more RAM. If you are running 32-bit versions especially you may have the app crash on large models. You can reduce the oversample resolution, or reduce the number of threads used for slicing, or both, to help get around this issue. Once I am done with this phase, I will work on reducing the RAM consumption once again. Also, the path precision should be a bit higher now than it was (basically at the same time as when I fixed the issue when the slicing plane was sitting on a surface), so you may not need as fine an oversample resolution as you had been using.

[crocky]

Hi, everybody.

Here is a version with some highly experimental support settings, namely:
Z Band - this is how often KISSlicer switches from rough support to the interface. I had been fixed at 2 mm. This helps when you have uneven lower surfaces (which might cause been too-small patches of interface support, small enough to not bond to the support layers beneath.) You can drop it down to 0 mm. This may help on flat overhangs by maximizing the sparse support, but may hurt on bumpy / spiked lower surfaces. I'm not sure if any of that made any sense, but there it is. [8^)
Gap (Z) - this will push the head down a tiny bit for the upper interface support, to weaken the bond between the top of the support and the bottom of the model. This may be a good idea when you are using a support that bonds too well to the model (like PLA model on PLA support). This only kicks in if the gap is > 0 and < layer thickness. The volume of material extruded is NOT modified, as this is really only intended for small gaps.
Flow Gain - this just makes the interface support thicker or thinner (the gain modified the default extrusion width). KISSlicer had been running at a constant 0.75. The allowed range is 0.1 to 2.0. Note that anything over 1.0 will probably be a bad idea if you have "[X] Solid Support-Object Interface" checked.
That's basically it. As I get feedback on what works and what doesn't I will modify the settings' limits and tooltips, etc. Please provide feedback!

@frozen: Sorry, I was talking about the "Warm Time" value that is on the [Material] tab (middle right of the tab). That has a major effect on the timing estimates, especially on multi-extruder prints.

thanks,
Jonathan

Beta 1.5 - 1.16

[McSlappy]

Thanks for relaying this. I really need to get the pro version - I've used it enough to know it's totally worth it! And once I get a tri-hotend I'll need the extra extruders

[crocky]

New beta's available, you need to have pro licence for them to work....

Includes Mac and Linux...

https://www.dropbox.com/sh/3evahzu5gfel ... FeGZa?dl=0" onclick="window.open(this.href);return false;

[0110-m-p]

Another new beta (1.5beta2.7) at the link crocky posted above. Quite a few cool new features like auto-place seam at corners if possible.