In some tests I did, I saw that the perimeters printing is done for each shell separately. So if you have specified 3 shells, it will print the 1st shell for everything and then the 2nd, etc.
So if for instance you have a rectangular box with 20 holes on it, it will go to each hole for every single shell.
This is a very wrong and inefficient way to do the printing not only because the time is much more due to excessive travel, but it increases the risk (especially for first layer) to drug some of the shells away.
All other slicers I have worked with deal with each separate area for all shells, so for a box with 20 holes it will print all shells for each hole before going to the next. It's much faster and much safer because it's harder to drug a 3 shell hole than a single shell hole.
This was the reason for some prints, with a lot of separate areas at the first layer, to change slicer because the print was failing every single time.
Re: Perimeters Print Logic Is Wrong
Actually, i like and dislike this, lets call it feature. It is good in a way that does not overheat one area and thus letting material to settle before applying the next perimeter.
But as with all things there are down sides to this also, and as you mentioned, as can help in one area, it can make s**t in other. Id be cool if one has control of this logic and can user be selected.
But as with all things there are down sides to this also, and as you mentioned, as can help in one area, it can make s**t in other. Id be cool if one has control of this logic and can user be selected.
Re: Perimeters Print Logic Is Wrong
This is better for cooling. You don't want to focus heat into one area for too long.
Re: Perimeters Print Logic Is Wrong
+1rsilvers wrote:This is better for cooling. You don't want to focus heat into one area for too long.
Re: Perimeters Print Logic Is Wrong
This would be a great "Option".
Re: Perimeters Print Logic Is Wrong
I disagree, this behavior does not improve parts due to "cooling". I print parts with very small details - lets say rods 2mm in diameter that protrude above a rectangular plate by 5 mm or so (standoffs). When the single perimeter prints on these standoffs and the next internal perimeter is not immediately printed (lets say 3 perimeters total) there is 1) a lot of wasted movement and 2) much worse bonding of the standoff to the sub-plate it's attached to. I printed parts like this from Kiss, Cura and Slic3r with none of these problems, and more importantly, there was no problem with "cooling". And the standoffs printed with these other slicers have much stronger bonds to the underlying plate, the S3D standoffs snap off very easily. To make things worse, when printing the 1st layer there is a reinforcement effect by printing the next perimeter(s) immediately. More contact surface area with the bed results in less likelyhood for the perimeter to come loose. Single perimeters are much more likely to detach than multiple perimeters. Printing all the perimeters before moving on is a tried and true method that every other slicer uses to advantage with no adverse effects. If anyone has a part where printing, let's say 3, perimeters all together results in "overheating one area" then please post an example and I'll retract my statement. Otherwise, I put this in the category of "the developer really has no real world experience with 3D printing". If he did, this "feature" would either 1) not exist or 2) if it provided any real benefit would be called out and marketed as "the next great thing". The fact that it is neither tells me its a bug at best or boneheaded implementation due to lack of domain experience at worse.
~CatDog>
~CatDog>
~CatDog>
(Alone in the world is a little CatDog!)
(Alone in the world is a little CatDog!)
Re: Perimeters Print Logic Is Wrong
Wow, you really wrote a lot of text that only really proved that in one specific situation, this "feature" didn't help "you". For a lot of other people, this is actually really useful. If you printed in PLA with any regularity, you would understand that stacking 2 or 3 perimeters on a small feature, will yield serious warping of those layers. Not in ABS, but in PLA.
Especially if you use software like MakerWare and this small feature ends up being the last area needed to finish a layer. In that case, you end up getting 4-6 perimeters in the exact same spot, saving you a precious 1-2 seconds of print time but giving you a really ugly pile of warped plastic. I prefer to get better prints, not get worse prints faster.
Try to step outside of your one specific situation and realize that there are many other situations that may prefer this feature. I don't have an issue with making it an option so people like you can turn it on or off, but I think your statement about the authors not having any real world experience is pretty far off. The fact they specifically have an option to "print islands sequentially without optimization" pretty much tells you that they understand cooling of small details and shoots your comment out of the water.
___
The print islands sequentially without optimization option changes the order in which the slicer prints "islands", or isolated areas of a particular layer. Usually, it will print all the islands of a layer in order, then do the next layer in reverse order so that time isn't wasted in traveling back to the original first island to start over. This is good for reducing print times, but for small objects where each layer doesn't take a long time, it is bad for cooling. The hot extruder warms the plastic near it, and laying down new plastic on top of warm plastic leads to deformation. The islands near the start and end of the sequence get a double dose of heat because they get two consecutive layers at a time. Printing them sequentially (i.e., in the same order every time) gives every island the same amount of time to cool between layers.
__
This software is light years ahead of the other garbage you reference. Why don't you use one of the other alternatives if they are so much better?
Especially if you use software like MakerWare and this small feature ends up being the last area needed to finish a layer. In that case, you end up getting 4-6 perimeters in the exact same spot, saving you a precious 1-2 seconds of print time but giving you a really ugly pile of warped plastic. I prefer to get better prints, not get worse prints faster.
Try to step outside of your one specific situation and realize that there are many other situations that may prefer this feature. I don't have an issue with making it an option so people like you can turn it on or off, but I think your statement about the authors not having any real world experience is pretty far off. The fact they specifically have an option to "print islands sequentially without optimization" pretty much tells you that they understand cooling of small details and shoots your comment out of the water.
___
The print islands sequentially without optimization option changes the order in which the slicer prints "islands", or isolated areas of a particular layer. Usually, it will print all the islands of a layer in order, then do the next layer in reverse order so that time isn't wasted in traveling back to the original first island to start over. This is good for reducing print times, but for small objects where each layer doesn't take a long time, it is bad for cooling. The hot extruder warms the plastic near it, and laying down new plastic on top of warm plastic leads to deformation. The islands near the start and end of the sequence get a double dose of heat because they get two consecutive layers at a time. Printing them sequentially (i.e., in the same order every time) gives every island the same amount of time to cool between layers.
__
This software is light years ahead of the other garbage you reference. Why don't you use one of the other alternatives if they are so much better?
Re: Perimeters Print Logic Is Wrong
Ah, yet one more pile of scat left by our new friend, Scott. You know nothing about what and how much I print. Let me add some clarity so you don't make an ass out of yourself again Scott. I print exclusively in PLA. I have printed well over 1000 kilos of PLA over the last 4 years. I have contributed at least 30 things to Thingiverse, two of which were featured. One of these things had very small diameter standoffs so don't tell me about geometries and printing dynamics. I'd love to post photos of these things but then you'd know who CatDog is and I don't want you to know. It's bad enough we know your name Scott.
And to answer your question, I do still use KISS, it is the only slicer that will create the quality output I require on a few of my parts. Don't get me wrong, I'm not saying I use it for everything, otherwise I would not have laid out my cold, hard cash for S3D, but I have a few parts where aesthetics are very, very important and KISS is the only slicer that will pull it off. You see, I like to have several tools at my disposal. Folks like you Scott fixate on one tool, so every problem looks the same. I'd love to see some of the work you produce, pedestrian I'm sure (oh my, was that personal? Sorry, I'll try to be less judgmental in the future).
And to respond to your comment, I contend that printing all perimeters in succession rather than one at a time serves the use cases of significantly more users than the converse (by orders of magnitude). The converse can lead to inadvertent pealing of the first layer and aborted prints. I've had that happen multiple times now on parts with thin footprints that never happen when all perimeters are laid down. The needs of the many... Rule #3 in usability. What sort of engineer are you Scott? Or should I just google it?
~CatDog>
And to answer your question, I do still use KISS, it is the only slicer that will create the quality output I require on a few of my parts. Don't get me wrong, I'm not saying I use it for everything, otherwise I would not have laid out my cold, hard cash for S3D, but I have a few parts where aesthetics are very, very important and KISS is the only slicer that will pull it off. You see, I like to have several tools at my disposal. Folks like you Scott fixate on one tool, so every problem looks the same. I'd love to see some of the work you produce, pedestrian I'm sure (oh my, was that personal? Sorry, I'll try to be less judgmental in the future).
And to respond to your comment, I contend that printing all perimeters in succession rather than one at a time serves the use cases of significantly more users than the converse (by orders of magnitude). The converse can lead to inadvertent pealing of the first layer and aborted prints. I've had that happen multiple times now on parts with thin footprints that never happen when all perimeters are laid down. The needs of the many... Rule #3 in usability. What sort of engineer are you Scott? Or should I just google it?
~CatDog>
~CatDog>
(Alone in the world is a little CatDog!)
(Alone in the world is a little CatDog!)
Re: Perimeters Print Logic Is Wrong
Printed 1000's of kilos of PLA but doesn't know the difference between a gcode viewer and an STL. Seems likely.....
Perimeters Print Logic Is Wrong
Quick (pre-sales 
) question: is this behavior (printing the perimeters for all features before printing the infill for all features) still active in the current version of Simplify3D?
) question: is this behavior (printing the perimeters for all features before printing the infill for all features) still active in the current version of Simplify3D?