Variable Layer Height Algorithm

[arhi]

Changing the layer height by a factor of ten

if you can set min and max then there's no "factor of ten", if you can also configure available increments, or even better a discrete values you "allow" (for e.g. 0.1, 0.15, 0.22, 0.26, 0.3, 0.42 only) the that problem goes away ..
first layer can always be separate with separate height, flow multiplier, temperature, speed ...

[arhi]

btw a detailed explanation here: http://kalanka.de/blog/adaptive-slicing/
and a guy's master thesis (cool read, german only): http://kalanka.de/blog/wp-content/uploa ... erfall.pdf

[jballard86]

this would be amazing.

[MaxGyver]

+1
That would be an awesome Feature, capable of reducing print time and improoving print quality at once !

[mroek]

As I've said in a few other threads, I am afraid that S3D isn't really being actively developed any longer, so a feature like this isn't something we're likely to get.

However, if the .factory file format had been known (or reverse-engineered), it would be possible to create some software that would read a factory file, then add a bunch of processes (with variable layer heights), and then this file could be read back into S3D for slicing.

Such a method would of course rely both on knowledge of the .factory file format, and also some external input as to where the layer heights needs to change. There would also probably be a limit as to how many processes should be created, but I guess even with, say 20 processes, you'd get a reasonably good result, at least for models that isn't too tall.

[CompoundCarl]

As I've said in a few other threads, I am afraid that S3D isn't really being actively developed any longer

As someone who's been using the software for many years, I definitely disagree. They've released probably 20+ updates since I started using it. They just released 3.1.1 about a week ago, so it's obviously still being developed. They put out 3.0 last year which had probably 30 major features in it, so I have no doubt they will add many big features in 4.0, 5.0, etc. Just look at their release notes page. There is a long history of constant updates: https://www.simplify3d.com/software/release-notes/

But back to the topic, I think automatically changing the layer height could be useful, but as other mentioned, I typically use different settings for different layer heights. So if I'm using settings for a 0.3mm layer height, but then it changes to 0.1mm in some places, I'm not sure that would work. I'd almost prefer to just setup a separate process and do it that way so that I can control it better.

[mroek]

As I've said in a few other threads, I am afraid that S3D isn't really being actively developed any longer

As someone who's been using the software for many years, I definitely disagree. They've released probably 20+ updates since I started using it. They just released 3.1.1 about a week ago, so it's obviously still being developed. They put out 3.0 last year which had probably 30 major features in it, so I have no doubt they will add many big features in 4.0, 5.0, etc. Just look at their release notes page. There is a long history of constant updates: https://www.simplify3d.com/software/release-notes/

Yep, that list contains 26 versions, but that's not really the point. The point is that since version 3, which was a big release with a lot of new stuff, very little of any real substance has happened. None (absolutely zero) of the most requested features has been addressed, for the most part it has been simple stuff like adding new printer profiles.

As far as I can tell, the first version was released very early in 2013 (can't find any exact info) under the name Creator (or S3DCreator). Then, under a year later they released version 2, after having released 12 minor updates inbetween.

Version 2 got 7 minor updates before version 3 came along in June 2015, some 18 months after version 2. Since then there has been only 4 minor updates (and they've been really minor), and almost 17 months has passed.

In other words, almost the same time as between version 2 and version 3 has elapsed now (at the time of writing) but with fewer updates. This could of course mean that they are busy with version 4, so less time for minor updates, but it could also mean less development is happening. I am of course hoping for the former, but I am not as hopeful as I perceive you to be.

If we extrapolate the time between the previous releases, that means v4 should be released sometime around August 2017. If instead we hope for the same time as between versions 2 and 3, then we should be in for a great Christmas present.

[CompoundCarl]

Looking at the release notes, I would think that many of the features in 3.0 and 3.1 would have taken far longer than some of the 1.x and 2.x releases. So saying it was 10 releases one year vs 15 the next isn't really a good way to compare.

You also seem to be implying that they won't start working on 3.2 or 4.0 until the last small update like 3.1.1 is completely done. I doubt that's how it works. Most companies are working on minor updates at the same time as major releases.

Anyways, it seems silly to be discussing this here at all. If we really want to help the developers, it would be better to try to be constructive and add useful feedback to these topics.

[mroek]

I am a developer myself, so I know fully well that the next major version usually is the one that gets attention, and that's why minor updates to the current version are, well, minor.

However, the lack of interaction with the developers on this (which is their own forum) is disturbing. Providing some roadmap info would have benefited both S3D as a company and this community. The utter silence is (or could be perceived as) a tell-tale sign that not much is happening at S3D. And as I've said before, the software is still popular and probably selling well, so the company might not want to have to pay expensive developers to do anything new.

The amount of constructive feedback on these forums are plentiful, but with NO feedback from the developers, it's no wonder that people tire a bit.

And also, since other slicers are getting more frequent updates, at some point it will hurt S3D sales. I still like S3D way better than any of the others, but I could mention that Craftware is about to release just what this thread is about, namely a variable layer height function. This they have mentioned publicly on their own forum, as developers should do to keep their community happy.

[arhi]

@mroek, not much point discussing no dev feedback inside variable layer height topic .. if/when they want to provide feedback they probably will in their own topic or one of the "where's the feedback" topics ... we might want to keep this topic about VLH.

now, with regards to vlh, I printed around half a kilo of green abs using variabile layer height and I like it .. there are few things I'd like to be done differently (main thing - I'd like to give "available" layer heights for algo to chose from, not just from-to span as currently slic3r adaptive layer height fork does, I'd like to setup "everything" wrt those discrete layer heights - not have app use same settings for all of them ... for e.g. I want to print 0.1 with 0.25mm line 60mm/sec 222C and 105% extrusion ratio while 0.3mm layer I want to print at 30mm/sec with 0.45mm line 228C and 98% extrusion ratio .. etc etc .. many more settings I'd set for each layer height) ...

now with s3d's ability to do "process" I'd for e.g. like to be able to define X processes, set each process for "from-to" surface angle of the part and then s3d automatically selects the process according to the part geometry .. for e.g. I set from 0-5 degrees 0.5mm layer and everythig else one can setup in a process .. then I setup from 5-30 degrees 0.35mm layer process, then from 30-60 0.1mm layer process etc etc .. and then instead of "use this process from x to y Z" s3d use the process adequate for surface .. something like that..

now as for the "much faster" .. it really depends on your object, I have to say most of the parts I was printing printed longer then "normal", lot more details, nicer print of course, but longer .. I normally print "technical" parts with 0.25mm it's fast and more then good enough for technical stuff ... the "adaptive" thing was set for 0.15mm to 0.35mm and 0.15mm "cusp value" (maximum deviation from objects original surface caused by the stair stepping effect. typical values between 0.1 and 0.2, default 0.15) .. for me speed is not a big deal so printing slower does not pose any problem

so it's not all cakes and roses but it is incredibly powerful option... now in my opinion s3d has already so many technology in place that it's way easier for them to implement this type of feature then it was for slic3r (done as master work by a german student) or craftware (they say they are working on designing this feature) ... with s3d currently you can manually do this, it is tedious and hard work but possible, it is the only thing I attm use processes for .. but since it can be done manually in s3d, automating it should be easy (compared to others, not in absolute terms)