MatterControl 2.19.2 - Release Candidate
Thanks for your reply, Lars. Then I'll wait for the macOS update.
The Mac alpha is now up
I got it.
First impressions, 15% less memory usage and 12% less CPU on this 18.104.22.16854 Alfa update.
If I encountering a crash error on Mac I will send debugging information.
For our Mac Users as well as Windows we would like to know what kind of performance increases you have seen so far if you have measured them?
I did a simple performance test comparing the estable version 22.214.171.12423 and the 126.96.36.19954 on my MacBook pro with the Activity Monitor application.
In the same conditions I compared the use of the CPU and RAM and I calculated the percentage of difference between both versions.
Nice job @aemux thanks for letting me know, hopefully we can get a few more people to report in their findings. We like all the data points!
HmBeerBrewer last edited by
Has the automatic support tool replaced the ability to mark any primitive as support? I'm no longer able to mark a primitive as a support unless I'm overlooking something. Version 188.8.131.5257 (stable release)
Any chance of future versions having tree type supports similar to Cura? It would be nice to have the supports originate all on the bed for some prints.
HardcoreMidget last edited by
Finding print leveling for an A8 problematic. Have been informed that it is designed to be used with Zstop on the TOP, not the bottom. E0 Z offset not available at this update
The new support options still works like the old one with just more options, you can still click the support button while having an STL selected but you can also click the drop down to generate more options.
You can toggle an object to be support with the button (while the object is selected).
Or you can generate support columns with the drop down menu.
mstange42 last edited by mstange42
I finally had time today to download and test the new support system. I downloaded the Release version, not the release candidate. I have spent the last 3 hours messing and testing stuff with it and comparing to the old version, 184.108.40.20616.
When I use larger pillar (3-4) settings it does not generate the needed supports and leaves whole areas without.
When I use a small pillar (1.5mm, 2 still not small enough to get good supports) setting it does not connect the pillars so some supports are useless. On one model it added in an interface layer in mid air (no support), though I get this on occasion with the old version as well. But with some slight tweaking to my model the support this version generated was near perfect (Will Print this model). The old version did about the same for supports.
Furthermore the slicing time now takes longer than the old version. At least for the few objects I tested. On average I would say about 50% longer but it is hard to tell exactly as I have to run on different machines. Smaller supports the longer it takes to generate.
And since the 'Support Everywhere' option has been taken out I can not even use this for my prints. I know I can manually add supports in to take care of the issues I have but that defeats the purpose since the old version works better for the most part.
Before the following I did a couple small test prints that didn't take too long with no problems.
(Note: This is the model mentioned previously about the supports)
Now here is what PISSED me off. I went to print a large detailed model. I Sliced it and checked it before printing. I made NO changes to anything and hit Print, and it went to reslice it. It took 18 mins to slice it the first time, now another 18 mins to reslice it. The old version took about 13 mins to slice, and it never resliced anything unless I made a change.
Finally went to print but it had pull up on the brim so it was a failed print. AND THEN IT WENT TO RESLICE AGAIN!!!
This is a MAJOR EPIC FAIL for a release version.
On the brighter side it uses noticeably less CPU/Memory, as previously my sliced parts would have issues as I moved around making sure it created supports it needed. This version is MUCH better.
Also, the slicer appears to have a better algorithm as the pattern was nicer in the new version than the old (at least the first layer before it failed on me and....).
HardcoreMidget last edited by
@hardcoremidget Yep, I'm seeing the same thing about it Slicing every chance it gets. I've used the Slice button to preslice but it still goes ahead and Slices it anyway when the Print button is pressed. Started a 732 layer print last night and wow........
Thanks for the feedback. We have an issue in GitHub for the re-slicing and will git it fixed for the next release. We are also working to improve support, now that we have this new system.
You should still be able to create support everywhere. If there is a specific model that is not generating support correctly, filing an issue in GitHub and attaching or linking to the model is the best way to let the developers know.
We know you come to depend on our software working correctly, and that it is very problematic when there are bugs. We will work to correct these things as soon as possible.
Thanks again, Lars.
Just adding to what Lars mentioned above, I will be here to take any suggestions and bugs and send them on up the chain. Any and all input, positive or negative, will be taken and built upon. Our goal is to build a good slicer software and we know there will always be a few bugs but know that we will always work to eliminate them.
Thanks again guys!
mstange42 last edited by
Whatever changes you made to the slicer in 220.127.116.1169 sped up the slicing majorly. A 1562 layer slice too me about 12 mins compared to the over 30 mins it was taking me with the OLD version 18.104.22.16816.
Compared to 22.214.171.12457 a couple test slices that took about 50% longer, it's about 50% shorter in 169 just as my extra large model. I can now slice and reslice when exporting for the moment. Hopefully all of my test slices will come out as fast.
And thank you for fixing the Memory Bug I reported!