@@hajosmulders I feel like you missed the joke. the Joke binge that if you got lots to choose from... its not much of a standard you got, regardless of how open they are.
Good enough files formats tent to stick. 3MF sounds like a really good idea. The compression itself could be what is needed to make people using it. Especially for large models.
not really. jpeg has been replaced by heif this year ( within ios11 and android 9). All iphones and samsung phones (maybe even more) by default take heif photos and hvec videos. Only when clicking "send" or "export" the phone converts it back into jpeg to prevent incompatibility issues.
In Prusa repo there are also .step and .scad files to download, you can easily export .step to any format, but .scad is a bit more tricky (and problem is that some parts are available in .step only, and the rest is only in .scad)
The print settings really shouldn't be saved in the file. That's like saving the print settings in a PDF. Makes no sense considering people have different printers and quality requirements.
Stuff like number of perimeters, infill, custom supports, etc. are pretty much independent of the printes, it's pretty nice to have those set-up for you by the model creator. - Mikolas
I'm pretty confident that how slicers implement this feature is going to determine the adoption of 3MF. As an example: let's say the only supports I'm comfortable cleaning post print are tree style. Hypothetically if I have to override that setting for *every* print, it'd be easier for me to just stick to an stl where my user settings are respected.
I'm not an expert on the current PDF specification (I get the feeling no one is, these days, considering even Adobe has trouble transferring PDFs between their applications :P), but I'd be very surprised if it doesn't let applications store print settings and other custom data.
@@Prusa3D They're independent of the prints, but not independent of the *printers* (either the people or the machines). Not saying it's a bad idea, but I'd definitely want Prusa Slic3r to have some sort of prompt for whether I want to use the included settings or not (in a: yes/no/always ask sort of format). Custom supports in particular might be very dependent on the capabilities of an individual's printer.
The 3MF Consortium library reference implementation is a disaster - unbuildable for Node, with next to no documentation and missing bindings file despite claims that it exists in the V2 release. It’s going to be a long time before this proposed ‘standard’ becomes even remotely reliable
TROPtastic I went looking to review the AMF file format specs defined by ASTM International which is purportedly an ‘open’ standard Additive Manufacturing Format but when I attempted to review the spec I found it was only available if I purchased the PDF. This is not the open standard you’re looking for.
Thank you for your feedback. We actually have put a lot of energy into making proper Node bindings. Can you contact me directly to sort out your issues and fix them? It might be as simple as a version issue as discussed here: github.com/3MFConsortium/lib3mf/issues/181 Regarding documentation, Node has never been a core focus, but the C++ extension is pretty extensive (lib3mf.readthedocs.io/en/master/source/Cpp/LandingPage.html) and gives you the same class hierarchy.
Especially when STL is simple (LR/LL parse). If there's anything to learn about that being widely used.. Use simple formats and glue it together with some archive format. XML=autofail.
@@henrituhola no, it isn't, if there are proper schema for the XML files. Then there are tools to generate proper code. XML without schema, is auto-fail though, like any file format.
3MF does not prevent errors. You can convert every STL file into a 3MF file. If the software makes mistakes they will end up in the saved file no matter what format you use.
All cool but there is a reason why STL is still around: It's well tested and does the job in 99.9% of the time. Adding complexity and specialization (like material, print settings etc) is honestly not the right way to go. Keep it simple.
@@cocotug0 "STL is dead" is typical unserious advertising gossip.The industry and many enterprises use STL. I, for example, mill highly exact teeth of zircon, gold or steel. Companies wich have to earn money with it pay attention only to its well functions, companies change it only if the new format is better and it is industrial standard!
@@nothahn yes, I agree. but.. considering that SIEMENS, Dassault, autodesk, Stratasys and 3dsystems among others are interested in this. It means that it is going to become an industry standard. I mean... among those first 3 you have more than 50% of CAD marketshare...
It does! Fusion has 3MF export. You have to go through a quick setup though and it looks like it's not very well setup for 3D printers. There is also addon for easy 3MF export, but it has the scale off by a factor of 10 :/
Also, it really doesn't matter whether Fusion360 has it or not. The workflow should be CAD - Slicer - Sharing 3MF anyway, so the fact that we still use STL during the transition from CAD to Slicer is fine :)
@@my.luminaire apps.autodesk.com/FUSION/en/Detail/Index?id=9199586017740700165&appLang=en&os=Win64 hint: When you download and installed, Reboot F360 you will find the option under Tools next to sheet metal under Add-ins
When I save my 3D projects I save the blend files, the stl files, the gcode files and the 3mf. I never thought anyone would be interested in my 3MF file though. Especially since slicer settings seem to be so individual. Everyone has their slicer set up for their printer. Those settings can change depending on their printer, the filament they use and even the room temperature their printer is in. Before I used anyone else’s slicer settings I would want to print something small with it just to test it on my machine. I think going right into a 10 hour print with untried settings is pretty risky. And wouldn’t it be possible to stick some malicious gcode at the beginning or end of the print? I’d be willing to bet most people, me included, don’t know how to read gcode. And you pointed out one of the biggest stumbling blocks. 3MFs are slicer specific at the moment. Unless that changes I don’t think it’ll happen.
Harold D It seems the 3mf file benefits prusa because the printers are so great out of the box, very few people will modify them. Meaning the sliver settings should carry over between users very easily.
Makers Muse did some testing on 'malicious' gcode as I recall. For the most part, how malicious gcode can be is defined by the checks and balances of your slicer/machine (e.g. axis stop detection, temperature cutouts, etc.). I think some of the sites that accept gcode files don't allow files containing certain codes that could be used maliciously, but it's been a while since I watched Angus' vid, so, I could be misremembering that. Personally I find that my machine is in greatest danger (from itself/me) when I'm trying to callibrate it...
I will never share a 3mf file. It would assume the person on the other end is using the same printer/fillament/nozzle. Why would I do that to someone? Let's just say I got screwed by this from a file off prusaprinters and did not see the little pop up (which is way to small and should be right in your face) that says it was setup for a different printer. Maybe if Prusa Slicer made a setting that would ignore the settings in a 3mf that is dragged and dropped in. Till then I will stick to STL's and tools that pull STL's out of 3mf files.
Very interesting, thank you for sharing this. You certainly know far more than I and most people watching this because of your vast experience building and bringing 3D printing to the average person like myself so I look forward to where this new idea leads and I’m sure that you’ll be watching closely. Again, thank you for all that you’ve done for us in sharing your knowledge and all you’ve done to let us be able to 3D print things of all types at home at an affordable cost, you have done an amazing thing and I for one appreciate what you have done and thank you here publicly, well done and thank you. 😎👍🏻
Xml is great on this, storing structured data, and it support schmas. So it can automatically verify files and generate importer/exporter code automatically. And text are easy to compress efficiently, so no penalty there either. You can repair files with a text editor.
My guess is STL is still the predominant format is because there's really no mechanism for DRM. Anything other format would open the door to that and that's why we still have STL. But that's just my guess.
I don't see the point in .3mf. Saving slicer settings is pointless. If your reprinting something on the same printer, with the same filament, then just use the same gcode. If not, you don't want to have the same settings. Also, .stl files are already tiny. If you really want a better file type, use .amf for its curved triangles.
AMF (additive manufacturing format) triangles can describe curved surfaces and/or non-planar edges on 3D surfaces more accurately than .STL files. They are not limited to straight-edged, planar triangles. Better that 3MF
I think community has to develop that kind of file that shall be editable in or behave like STEP etc. Since the stl and 3mf both are same in terms of meshing..Because lots of people wants to modify premade stl files but they can't due to mesh. So editability can gives more powers to 3d printing sociaty . *If editability is impossible to develop for present CAD softwares then a perticular individual software has to be develop for only stl & 3mf files which can easily edit those files,like CAD stp files..*
There are SOOO many 3d file formats already - why reinvent the wheel? STL is NOT the problem, or rather it already fits the needs exactly. The problem are the people that think that a generic format like STL should check for things that cause problems for specific scenarios. You want models that are guaranteed to be printable? How about making sure the models is correct. That is not the job of the file-format.
stl might be ok to slice and print, but editing it is much harder than needed file sizes on complex geometry is also bad its resolution is finite, unlike on scalable vector geometry
I guess problem is simplicity. I'm not sure how complex file format like STEP is and what does it contain, but I can imagine some constructive solid geometry and complicated parametric curves and surfaces. Maybe something such as SVG for 2D vector graphics. Problem of formats like this is to write parser with all of features and some geometry processing library as well. Reading of something relatively simple such as screw modeled by cylinder+helix+hexagonal head with rounded edges and hole for screw driver can be extremely complex. Polygonal model can consist of 50K triangles and STL will occupy 2.5MB, but it's much easier to read.
@@skewty I always felt XML could be more readable than JSON depending on who designed the schema, but am sure TONS of ppl would disagree. JSON is clearly more efficient speed and space wise. I also personally dislike indents (yaml/python) being so important for structure since accidentally adding spaces while editing a file is so easy (I debug this kind of thing all the time). However, that's also probably just a product of what I am comfortable with.
@@skewty No. One would use json 95% of the time when dealing with portable data formats. But in complicated data structures and critical communication systems XML has the advantage by providing native data validation through XSD as well as namespaces and attributes. In this case json would probably have been fine.
People got so burned out by really badly written XML that they somehow that somehow they are blind to the fact that JSON is way harder to read and write
the software i use to model objects doesn't have a 3MF exporter, only a STL exporter. so im going to stick with STL till 3MF becomes more adopted by modeling software.
I'd really just be happy if slicers compiled STEP (or some other CAD interchange format)->Gcode. STL needs to go, but I'm not 100% sure another triangle format is the answer. Also, with a quick glance at the organization's web page, they apparently support DRM, so I'd say "burn it to the ground", regardless of any other concern.
Sharing hahaha.... sounds like something a shill would say. He's not sharing, he's a business, we're customers. That's like thanking Walmart for sharing.
Yes i do use 3MF, however I wasn't aware of food instance the slicer settings being saved. Well check wherever the format is clever enough to save Cura and Prussa slicer settings.
Hm, FreeCAD seems to support AMF but not 3MF. But in general, I tend to update the STL and let the Slicer apply the settings. Both Cura and Slic3r are better at preparing a print than any modeler/CAD I have seen.
Just yesterday I saved a project which used a support blocker mesh and two cutting meshes. I further tweaked the support settings for this specific part because it requires some really tricky support. I think Cura only saves the settings I changed, so when I open the project in a newer version I can still benefit from the better settings, but I also benefit from my project specific settings.
@@angelorf PrusaSlicer takes a slightly different approach: All the settings are stored, but when you load the 3MF, it will show you at the parameter pages, which parameters are different from the current system settings. This way you may reprint your old 3MF the same way as before.
It seems to me this is all about digital protection of a model someone has created. You emphasized the copyright and license terms. It seems as if he industry is trying to manage the community that has been working just fine with .STL models. As others have pointed out - "there are too many "standards"". It is best to have a RAW file that is "standard" and have others (like Prusa or Cura) to import the RAW model and enhance it. Just my never to be humbled opinion.
I have tried to use 3MF but unfortunately Fusion 360 does not support 3MF well enough to use for precision parts. You cannot specify fine enough surface deviation so that enough triangles are generated. This means that you get facets on the faces of round objects like cylinders. This is even a problem with regular STL files but tolerable. An example of this is that a 6 inch diameter cylinder (~150 mm) will have 480 facets on the outside of the cylinder in the STL when you specify the minimum surface deviation in Fusion 360. This is visible in the print on a well tuned machine. In 3MF there is currently no way to specify the surface deviation or even any level of refinement so you get a pretty coarse output file. I believe this same example had something like 100 facets on the face of the cylinder. This is visible on a poorly tuned machine. It produces ugly or in many cases unusable parts. If you are making engineering parts you need to wait a while for the software to catch up. I was hoping that a new format would not use triangles to represent curved surfaces but 3MF still uses triangles.
3d printers could make bridges at an angle that's not horizontal, and follow a curve with the z axis, will these features ever become available in the slicer? It would make more things possible to print
Manifoldness is when one edge is shared by more than two triangles. For example when one part drawn in CAD touches another part and shares edge or surface. I can't recall why it's inconvenient to have them in the model, but there can be some ambiguities, separate mechanical parts can't be separated and they likely create many problems with mesh operations such as simplification.
It wasnt clear from the video how 3MF fixes it. Seeing as you still need to export the STL from the modelling software, it sounds like until you fix it, it wont save to 3MF.
From the 3MF website about their "repair" tool: " 3MF Service allows everyone access to the Windows 10 3D printing APIs. Repair performs the following actions: Repairs the 3D mesh by closing holes and fixing geometry to create a water-tight 3D mesh suitable for 3D printing." In strict definition: {A manifold is a topological space that is locally Euclidean} - any 3 dimensional object that you would want to print... if it deviated from this mathematical property.... you can't print it. This property can be checked for all the points and volumes described in the 3MF file... so presumably the software performs that check. Also, take that to mean === Concerning "manifoldness" It looks for things like floating points, planes, or volumes that cannot be printed. It also looks for completeness in the volumetric shells (holes in the manifolds, providing for "watertightness). I think it also sums up the volumes of any distinct parts and ensure the sum of the part volumes sum up to the whole volume (otherwise you would have internal volumes whose walls protrude or overlap into other volumes in the space, making it have a higher internal volume sum than external collective volume sum... making it (potentially) unprintable.) It likely also looks for things that can be made in a drawing program that can't be made in real life... like walls or partitions with a thickness of zero or two objects meeting at a "point" of infinitesimal size.
The standard says that models must manifold. So it's basically up to modeling software to enforce it. I predict we'll start seeing "Are you sure?" save dialogs.
What I'd find more interesting would be a file standard which could be properly altered and worked on in CAD Programms like fusion etc... I'd love to slightly change around a few designs on thingiverse but that is an absolute pain in the bum with only an stl file which can't be imported into a cad Programm as a solid
That would be cool but CAD developers would have to adopt a standard (which has happened in the past e.g. DXF) but unfortunately i just don't see that happening. It's a great idea though.
Is there any print quality improvements? Are we going to have better prints as a result of this or are we basically adding another level of complexity on top of what's already commonly used?
I'm not sure you really understand STL .... Anyway size , scale , mesh orientation are saved and reflect only the work you did with your 3d modeling software. STL is good but do not save material/color information, but it is perfect to save shape just like an obj file ! This video is about promoting 3MF , maybe good for the new needs arising , but stl is also very good , it's simple and frugal. If you need a sculpt only , it is good enough
I save everything I have dialed in as 3MF, except once you do that, if I change the STL then the "reload from disk" no longer reloads the STL but reverts to the last 3MF save.
Simon Belanger In Prusa Slicer when I load several STL, and start working with the modifiers, I might realize I should change my design on one of the objects. So I will go to Fusion360 and fix it, export to STL (overwriting the original), and when I go to Prusa Slicer I right-click that object and reload from disk. It will update it and I can proceed. Once I save the project, it doesn’t work the same way anymore and will reload the whole 3MF file instead of just that one STL. I just expected it to work like it does before saving the project.
Well I would use .3mf but Fusion 360 doesn't support this file format, direct transfer to Cura which I use often, doesn't even allow you to change the file format...
On the STL's that I spend days on running and getting right (scaling, rotation, position, supports, etc.) Through multiple test prints... I absolutely save in .3MF. In fact, I need to just get in that habit now that I've reached the point of needing to reprint some objects (Such as a collision that ripped my fan shroud off, so I needed super glue and duct tape to print a new shroud).
I am a freelancer 3d modeler, and I have a bunch of clients who work with 3d printing, and sometimes they send me an stl to edit something, and most of the time I have to remodel it from scratch, since the stl has a crappy topology... This new format will save me in these situations
Probably a stupid question, but just verifying, when you export to Gcode will it still be a "normal" Gcode? So any printer will still read it as it does now, so like on my MK2s? I don't see it being an issue but just making sure
Well I design all of my stuf in BLender and export them as STL to load into Slic3R There's no reason I would need to export something from Slic3r OTHER than Gcode, right ?
If I load a model that is just faces (not solid anywhere) will it print nothing? I modeled a car a while ago for CG purposes, but want to print it... haven't been able to find a way to get it solid.
That's ok! Export STL - Import it into a PrusaSlicer and save 3MF with your slicing settings. The transition from CAD to slicer can stil be in STL, even though it's not optimal.
Looks cool, but my modeling apps don't support it without a separate outside conversion, and so far haven't done anything large enuff to have a need for it (yet).
I'm only sad that the ISO standard .AMF format is the one being ignored where the corporation-controlled .3MF is the one that is that is taking off. Still, market forces..
If 3MF is open-sourced, it's available to everyone for free. But, honestly, it's the first thing I heard about either. So I'm curious. What are the advantages of AMF?
When we introduced our new PrusaSlicer 2.0 and launched the new PrusaPrinters community hub, some people asked us why we prefer sharing the models in 3MF format over AMF. It’s simple: the 3MF specification is newer and unambiguous, and it’s available for free. Also, 3MF files are supported by Cura and Microsoft 3D builder (among other apps).On the other hand, AMF is ambiguous and the specification is to be paid for. AMF files were never properly supported by any software that we know of. At best, AMF files were used to the same extent as STLs, meaning that the AMF only carried a single mesh without any additional attributes. With that said, PrusaSlicer can still load AMF. ( - Mikolas)
Some CAD packages aren't well suited to 3MF output (yet); because we use both Ultimaker Cura and PrusaSlicer, asking one of the CAD designers to give me a 3mf wouldn't make much sense. That said, if I think I'll need to print a part again, you bet I'm saving that sucker as a 3MF file.
In my opinion it is a bad idea to store slicerinformation INSIDE the file. It may get outdated as soon as a new version of the slicer comes around. So you would have to update your shared models over and over again. And which versions slicer information will you store? the latest one? All? Will get pretty crowded soon.
Same, but it's not really a big problem. You use STL for the transition from Blender to slicer, but then you save 3MF in the slicer with all of the fine-tuned settings (supports, variable layer height etc.) and that's what you share.
Blender -> stl -> ps -> 3mf + gcode is missing the elegance for iterations. It would br awesome, if ps could generate a blender mesh based on the gcode info, so it can be considered when optimising and iterating on a design - especially when details and tolerances matter (changing nozzle size, width, heigth and even material changes dimensions of the final object)
@@Prusa3D Then why would this "kill" the STL format? That makes no sense because you are still using STL. Unless you completely and 100% eliminate STL files, then you are just treading water and splashing it at people.
I am a beginner in 3D printing. Could you please tell me which software to use to create a 3MF model? Ideally should work on linux. (Could be also paid software.) Thank you very much.
im not an xml fan. it turns perfectly readable raw data into gibberish. i usually just stick a bunch of key value pairs in a text file with some sort of separator. the complexity of the format should reflect the complexity of the data being stored. using xml to basically store what amounts to a list of triangle vertices is overkill.
I disagreed most what you said in the beginning. For example A model don't have to be self validating. You just need an external validation tool. And don't have to be readable, again you can use tools. See how flatbuffer is so popular
Hey guys coming back to this that has not taken a lot of love in printable sadly. It could be nice that when uploading in printables there is a pop-up that suggest the step and 3mf format. And that if the person uploads an stl offer to transform it to 3mf automatically even without the printing settings just the stl inside of it. This could be nice to add. As more people see the format around more people will use it. A good way to push the format. (eventhought step is still a superior format 😅)
Hmm.. the instantiation referencing in the language is definitely going to be a GoodThing(tm), as well as the ability to avoid manifold issues. The Zip'd XML file structure also sounds promising. However, the embedding of printer related settings.... that is going to be something that is hit/miss depending on the printer being used. Ie, for folks using a fully standardized printer like a Prusa, it will be awesome. For others, I'm guessing the settings will be a sort of starting point... I see this as a huge benefit overall over STL files though. Hopefully adoption of the format will be widespread... but I also hope that the file format will remain something that can be edited/imported/exported.
I do not dispute the technical argument. The user base is not the people that need convincing. You just need to persuade the ecosystem of CAD and slicer software manufacturers that it's worth switching and the user base will follow.
Eh. If it was not based on XML, but instead on something more sane like Protocol Buffers, I would be in. I will stick to STL, fast and super easy to write parsers and generators with zero external dependencies. Also using zip, another meh. There are better, faster and more open standards to use with less quirks. Adding instancing and multi mesh support to STL is super easy. No need for new format for that. Comparing size of 3MF with raw STL without compressing it same zip compression settings is just silly and missleading.
Yea I see the addition of identity into a file as a way for the gov to try and control additive manufacturing all they have to do is require identification info in all transmitted 3mf files so they can " track" where it came from
*puts on tinfoil hat* you realize if you wanted, you could put the metadata into an STL. I really don't see what's wrong with having a field for the designer's name.
I think 3MF is nice, and i used it something in the last months. But, where there www.prusa3d.de/prusa-i3-druckbare/ are no 3MF Files?! Only STL and some GCODE.. Can you provide 3MF for all?
I like STEP ap214, VRML 97, & OBJ most of the time. STEP ap214 is usually very small for sharing. The real issue with STL is how difficult it is to bring into parametric CAD software it's very lossy relative to NURBS file. Again, STEP is king, maybe even IGES. I print with color printers so i need more than stl file. The main benefit of this seems to be how it can't have non-manifold edges, but most of what i see is self intersections from poor design intent prior to export. If it is still mesh, then self intersections seems like it will always be an issue. Mesh formats will probably never replace SLC for certain applications. I love how many companies tried to fix this together but still seems lossy. Do you disagree?? Please share thoughts :)
STL stands for standard tessellation language. Nothing else. Its predecessor was PHIGS+, Programmer's Hierarchical Interactive Graphics System (PHIGS).
AFAIK the verticies should get doubled so that there are two planes each facing other direction. Which removes the ambiguity. It does this if an edge belongs to more than two faces. Sure, you'll still have an infinitely thin wall, but that's a separate problem. Manifold problems aren't usually thin walls but more of an incorrect topology that is ambiguous and each app can interpret it differently. - Mikolas
That would be awesome, afaik no slicer in the world can do that atm. It's something that would be very hard to implement, but super useful. (I'm keeping on my wishlist with no date estimates :) ) - Mikolas
I use 3mf, but it is a pitty that not all settings can be reloaded. When I make a projekt, I direct upload to octoprint. works great. But on a realod of the 3mf, there is no octoprint information and i have to export gcode manualy OR switch to the printer - but then prusaslicer will delete much settings while switching and my "relaod" is not the same anymore. that is really anoying and make me sad. please correct that. THEN 3mf will rock like it could. Thanks, and happy printing!
I use Creo at work for model creation. Unfortunately, when I looked the other day it doesn't support 3MF output. It's too bad because 3MF looks like the future of 3D printing files.
Most good tools have built in checking for manifolds. STL Check in 3ds Max was a requirement for stencil shadows, back in the day. For modern mesh formats, we have FBX (de facto, not open,) collada (supported everywhere but buggy) and opengex (high quality, not as wide support as collada.) Inventing another format is just the wheel turning over again, with the same bugs to come as the previous formats have already worked out.
You know what’s great about standards? ...
There are many to choose from!
Yeah, but some are more open than others...
@@hajosmulders I feel like you missed the joke.
the Joke binge that if you got lots to choose from... its not much of a standard you got, regardless of how open they are.
Good one :-D
relevant xkcd: xkcd.com/927/
xkcd.com/927/
JPEG 2000 will replace JPEG... 19 year later... Still using JPEG.
Good enough files formats tent to stick. 3MF sounds like a really good idea. The compression itself could be what is needed to make people using it. Especially for large models.
@@PerMejdal Jokes on you, I actually sometimes use JPEG2000. Or should I say JFIF? :D
not really. jpeg has been replaced by heif this year ( within ios11 and android 9). All iphones and samsung phones (maybe even more) by default take heif photos and hvec videos. Only when clicking "send" or "export" the phone converts it back into jpeg to prevent incompatibility issues.
Fun fact: if you watch a movie in a modern cinema, the frames are usually stored as JPEG 2000. It has found its niche, I guess.
@@jackpat6068 you are exactly wrong 😉
sooooooo when are we gonna see 3mf files for prusa parts. The download page is gcode or stl.
In Prusa repo there are also .step and .scad files to download, you can easily export .step to any format, but .scad is a bit more tricky (and problem is that some parts are available in .step only, and the rest is only in .scad)
"uploading STL's is not necessary" only uploads stl.....
Lol you got him.....
Schablam!
The print settings really shouldn't be saved in the file. That's like saving the print settings in a PDF. Makes no sense considering people have different printers and quality requirements.
Stuff like number of perimeters, infill, custom supports, etc. are pretty much independent of the printes, it's pretty nice to have those set-up for you by the model creator.
- Mikolas
@@Prusa3D I agree, as long as they can be overridden after loading the file. Some of us simply like to experiment with stuff after all :)
I'm pretty confident that how slicers implement this feature is going to determine the adoption of 3MF. As an example: let's say the only supports I'm comfortable cleaning post print are tree style. Hypothetically if I have to override that setting for *every* print, it'd be easier for me to just stick to an stl where my user settings are respected.
I'm not an expert on the current PDF specification (I get the feeling no one is, these days, considering even Adobe has trouble transferring PDFs between their applications :P), but I'd be very surprised if it doesn't let applications store print settings and other custom data.
@@Prusa3D They're independent of the prints, but not independent of the *printers* (either the people or the machines). Not saying it's a bad idea, but I'd definitely want Prusa Slic3r to have some sort of prompt for whether I want to use the included settings or not (in a: yes/no/always ask sort of format). Custom supports in particular might be very dependent on the capabilities of an individual's printer.
The 3MF Consortium library reference implementation is a disaster - unbuildable for Node, with next to no documentation and missing bindings file despite claims that it exists in the V2 release. It’s going to be a long time before this proposed ‘standard’ becomes even remotely reliable
How does it compare to AMF?
TROPtastic I went looking to review the AMF file format specs defined by ASTM International which is purportedly an ‘open’ standard Additive Manufacturing Format but when I attempted to review the spec I found it was only available if I purchased the PDF. This is not the open standard you’re looking for.
Thank you for your feedback. We actually have put a lot of energy into making proper Node bindings. Can you contact me directly to sort out your issues and fix them? It might be as simple as a version issue as discussed here: github.com/3MFConsortium/lib3mf/issues/181
Regarding documentation, Node has never been a core focus, but the C++ extension is pretty extensive (lib3mf.readthedocs.io/en/master/source/Cpp/LandingPage.html) and gives you the same class hierarchy.
Especially when STL is simple (LR/LL parse). If there's anything to learn about that being widely used.. Use simple formats and glue it together with some archive format. XML=autofail.
@@henrituhola no, it isn't, if there are proper schema for the XML files. Then there are tools to generate proper code.
XML without schema, is auto-fail though, like any file format.
3MF does not prevent errors. You can convert every STL file into a 3MF file. If the software makes mistakes they will end up in the saved file no matter what format you use.
All cool but there is a reason why STL is still around: It's well tested and does the job in 99.9% of the time. Adding complexity and specialization (like material, print settings etc) is honestly not the right way to go. Keep it simple.
yeah, but considering the backers... STL is going out of norm
@@cocotug0 "STL is dead" is typical unserious advertising gossip.The industry and many enterprises use STL.
I, for example, mill highly exact teeth of zircon, gold or steel. Companies wich have to earn money with it pay attention only to its well functions, companies change it only if the new format is better and it is industrial standard!
@@nothahn yes, I agree. but.. considering that SIEMENS, Dassault, autodesk, Stratasys and 3dsystems among others are interested in this. It means that it is going to become an industry standard. I mean... among those first 3 you have more than 50% of CAD marketshare...
@@cocotug0 Between "interested" and industry standard can be a long time...
@@cocotug0 interested/= will be industry standard
In Windows 10, 3MF files have an icon with an image of the model to be printed
Not by default. You have to have some software that is providing that for you. 3d Builder is one free example.
my pc shows stl thumbnails
nothing installed specifically to do that
@@marc_frank You probably have 3D Builder or other software installed for other reasons. The thumbnail is a side benefit of those.
My STL files have images of the files too
I can't view thumbnails on any 3mf files. Is there a way to fix this?
If fusion360 did had native 3mf export i would use it.
Yeah, wtf? Why don't they incorporate this???
It does! Fusion has 3MF export. You have to go through a quick setup though and it looks like it's not very well setup for 3D printers. There is also addon for easy 3MF export, but it has the scale off by a factor of 10 :/
Also, it really doesn't matter whether Fusion360 has it or not. The workflow should be CAD - Slicer - Sharing 3MF anyway, so the fact that we still use STL during the transition from CAD to Slicer is fine :)
@@my.luminaire apps.autodesk.com/FUSION/en/Detail/Index?id=9199586017740700165&appLang=en&os=Win64 hint: When you download and installed, Reboot F360 you will find the option under Tools next to sheet metal under Add-ins
@@TheMikolasZuza Not at all. I have never seen any option, plugin, or anything for Fusion360 that exports 3MF format. Link?
When I save my 3D projects I save the blend files, the stl files, the gcode files and the 3mf. I never thought anyone would be interested in my 3MF file though. Especially since slicer settings seem to be so individual.
Everyone has their slicer set up for their printer. Those settings can change depending on their printer, the filament they use and even the room temperature their printer is in. Before I used anyone else’s slicer settings I would want to print something small with it just to test it on my machine. I think going right into a 10 hour print with untried settings is pretty risky.
And wouldn’t it be possible to stick some malicious gcode at the beginning or end of the print? I’d be willing to bet most people, me included, don’t know how to read gcode.
And you pointed out one of the biggest stumbling blocks. 3MFs are slicer specific at the moment. Unless that changes I don’t think it’ll happen.
Harold D
It seems the 3mf file benefits prusa because the printers are so great out of the box, very few people will modify them. Meaning the sliver settings should carry over between users very easily.
Makers Muse did some testing on 'malicious' gcode as I recall. For the most part, how malicious gcode can be is defined by the checks and balances of your slicer/machine (e.g. axis stop detection, temperature cutouts, etc.). I think some of the sites that accept gcode files don't allow files containing certain codes that could be used maliciously, but it's been a while since I watched Angus' vid, so, I could be misremembering that. Personally I find that my machine is in greatest danger (from itself/me) when I'm trying to callibrate it...
3MF consortium member lists shows one thing: This is not a community format. I am aware of STL's limitations but 3MF may not be the direction to go.
Just came here to type something similar - I saw Microsoft, HP and one other (too lazy to check again) and simply said - NO
Thank you for adding 3MF support. This is excellent! I'm excited to see this becoming the way forward for us.
I will never share a 3mf file. It would assume the person on the other end is using the same printer/fillament/nozzle. Why would I do that to someone? Let's just say I got screwed by this from a file off prusaprinters and did not see the little pop up (which is way to small and should be right in your face) that says it was setup for a different printer. Maybe if Prusa Slicer made a setting that would ignore the settings in a 3mf that is dragged and dropped in. Till then I will stick to STL's and tools that pull STL's out of 3mf files.
While it really comes down to personal preference, I think it would be a safe bet to assume that this didn't age well...
Very interesting, thank you for sharing this. You certainly know far more than I and most people watching this because of your vast experience building and bringing 3D printing to the average person like myself so I look forward to where this new idea leads and I’m sure that you’ll be watching closely. Again, thank you for all that you’ve done for us in sharing your knowledge and all you’ve done to let us be able to 3D print things of all types at home at an affordable cost, you have done an amazing thing and I for one appreciate what you have done and thank you here publicly, well done and thank you. 😎👍🏻
Always excited to see progress in this area!
For anyone wondering, 3MF stands for 3D Manufacturing Format
"...the data are stored in a human readable, XML-based format..." : Hit stop. Inhale. Loud sigh...
Xml is great on this, storing structured data, and it support schmas. So it can automatically verify files and generate importer/exporter code automatically.
And text are easy to compress efficiently, so no penalty there either. You can repair files with a text editor.
... and stop watching any further.
JSON entered the chat
@@coced let me tell you about yaml...
"Human readable ... XML ....", When I heard that, I could not stop myself from laughing. Hail robots!
Waiting for my Prusa Mini to try it!
Soon!
Me too!
@@Prusa3D I'm looking forward to having mine!
Ditto!
Waiting for the XL :)
But i think he doesnt comes before fall / winter 2020
My guess is STL is still the predominant format is because there's really no mechanism for DRM. Anything other format would open the door to that and that's why we still have STL. But that's just my guess.
I don't see the point in .3mf. Saving slicer settings is pointless. If your reprinting something on the same printer, with the same filament, then just use the same gcode. If not, you don't want to have the same settings. Also, .stl files are already tiny. If you really want a better file type, use .amf for its curved triangles.
AMF (additive manufacturing format) triangles can describe curved surfaces and/or non-planar edges on 3D surfaces more accurately than .STL files. They are not limited to straight-edged, planar triangles. Better that 3MF
I think community has to develop that kind of file that shall be editable in or behave like STEP etc. Since the stl and 3mf both are same in terms of meshing..Because lots of people wants to modify premade stl files but they can't due to mesh. So editability can gives more powers to 3d printing sociaty . *If editability is impossible to develop for present CAD softwares then a perticular individual software has to be develop for only stl & 3mf files which can easily edit those files,like CAD stp files..*
Exactly, those meshes aren't really usable. STEP would be much better
There are SOOO many 3d file formats already - why reinvent the wheel?
STL is NOT the problem, or rather it already fits the needs exactly.
The problem are the people that think that a generic format like STL should check for things that cause problems for specific scenarios.
You want models that are guaranteed to be printable? How about making sure the models is correct. That is not the job of the file-format.
stl might be ok to slice and print, but editing it is much harder than needed
file sizes on complex geometry is also bad
its resolution is finite, unlike on scalable vector geometry
I guess problem is simplicity. I'm not sure how complex file format like STEP is and what does it contain, but I can imagine some constructive solid geometry and complicated parametric curves and surfaces. Maybe something such as SVG for 2D vector graphics. Problem of formats like this is to write parser with all of features and some geometry processing library as well. Reading of something relatively simple such as screw modeled by cylinder+helix+hexagonal head with rounded edges and hole for screw driver can be extremely complex. Polygonal model can consist of 50K triangles and STL will occupy 2.5MB, but it's much easier to read.
Printing STEPs directly would be awesome. Changing tolerances in slicer, etc.
(-Mikolas)
I bet you can still make non-manifold meshes. If not by other means, then just by directly editing the file.
„human readable ... XML“
XML was basically replaced by JSON which is sometimes replaced with more readable YAML / TOML
@@skewty I always felt XML could be more readable than JSON depending on who designed the schema, but am sure TONS of ppl would disagree. JSON is clearly more efficient speed and space wise. I also personally dislike indents (yaml/python) being so important for structure since accidentally adding spaces while editing a file is so easy (I debug this kind of thing all the time). However, that's also probably just a product of what I am comfortable with.
@@skewty No. One would use json 95% of the time when dealing with portable data formats. But in complicated data structures and critical communication systems XML has the advantage by providing native data validation through XSD as well as namespaces and attributes. In this case json would probably have been fine.
@@hajosmulders there are (non-native) schema validations for JSON, though.
People got so burned out by really badly written XML that they somehow that somehow they are blind to the fact that JSON is way harder to read and write
the software i use to model objects doesn't have a 3MF exporter, only a STL exporter. so im going to stick with STL till 3MF becomes more adopted by modeling software.
I'd really just be happy if slicers compiled STEP (or some other CAD interchange format)->Gcode. STL needs to go, but I'm not 100% sure another triangle format is the answer. Also, with a quick glance at the organization's web page, they apparently support DRM, so I'd say "burn it to the ground", regardless of any other concern.
Nice, I just learned something new :)
Thanks for sharing!
Sharing hahaha.... sounds like something a shill would say. He's not sharing, he's a business, we're customers. That's like thanking Walmart for sharing.
and the funny thing is, the Prusa Mini model file available on your website is still in STL format instead of 3MF, which comes out after this video.
Yes i do use 3MF, however I wasn't aware of food instance the slicer settings being saved. Well check wherever the format is clever enough to save Cura and Prussa slicer settings.
yeah, but when i want to use *MY* settings for a model then it takes alot more to switch everything
Hm, FreeCAD seems to support AMF but not 3MF. But in general, I tend to update the STL and let the Slicer apply the settings. Both Cura and Slic3r are better at preparing a print than any modeler/CAD I have seen.
Just yesterday I saved a project which used a support blocker mesh and two cutting meshes. I further tweaked the support settings for this specific part because it requires some really tricky support. I think Cura only saves the settings I changed, so when I open the project in a newer version I can still benefit from the better settings, but I also benefit from my project specific settings.
@@angelorf PrusaSlicer takes a slightly different approach: All the settings are stored, but when you load the 3MF, it will show you at the parameter pages, which parameters are different from the current system settings. This way you may reprint your old 3MF the same way as before.
It seems to me this is all about digital protection of a model someone has created. You emphasized the copyright and license terms. It seems as if he industry is trying to manage the community that has been working just fine with .STL models. As others have pointed out - "there are too many "standards"". It is best to have a RAW file that is "standard" and have others (like Prusa or Cura) to import the RAW model and enhance it. Just my never to be humbled opinion.
I have tried to use 3MF but unfortunately Fusion 360 does not support 3MF well enough to use for precision parts. You cannot specify fine enough surface deviation so that enough triangles are generated. This means that you get facets on the faces of round objects like cylinders. This is even a problem with regular STL files but tolerable. An example of this is that a 6 inch diameter cylinder (~150 mm) will have 480 facets on the outside of the cylinder in the STL when you specify the minimum surface deviation in Fusion 360. This is visible in the print on a well tuned machine. In 3MF there is currently no way to specify the surface deviation or even any level of refinement so you get a pretty coarse output file. I believe this same example had something like 100 facets on the face of the cylinder. This is visible on a poorly tuned machine. It produces ugly or in many cases unusable parts. If you are making engineering parts you need to wait a while for the software to catch up. I was hoping that a new format would not use triangles to represent curved surfaces but 3MF still uses triangles.
3d printers could make bridges at an angle that's not horizontal, and follow a curve with the z axis, will these features ever become available in the slicer? It would make more things possible to print
What is Manifoldness and how does 3MF help? I read the article, and I still don't get it.
Manifoldness is when one edge is shared by more than two triangles. For example when one part drawn in CAD touches another part and shares edge or surface. I can't recall why it's inconvenient to have them in the model, but there can be some ambiguities, separate mechanical parts can't be separated and they likely create many problems with mesh operations such as simplification.
It wasnt clear from the video how 3MF fixes it. Seeing as you still need to export the STL from the modelling software, it sounds like until you fix it, it wont save to 3MF.
From the 3MF website about their "repair" tool:
" 3MF Service allows everyone access to the Windows 10 3D printing APIs.
Repair performs the following actions:
Repairs the 3D mesh by closing holes and fixing geometry to create a water-tight 3D mesh suitable for 3D printing."
In strict definition: {A manifold is a topological space that is locally Euclidean}
- any 3 dimensional object that you would want to print... if it deviated from this mathematical property.... you can't print it. This property can be checked for all the points and volumes described in the 3MF file... so presumably the software performs that check.
Also, take that to mean ===
Concerning "manifoldness" It looks for things like floating points, planes, or volumes that cannot be printed. It also looks for completeness in the volumetric shells (holes in the manifolds, providing for "watertightness). I think it also sums up the volumes of any distinct parts and ensure the sum of the part volumes sum up to the whole volume (otherwise you would have internal volumes whose walls protrude or overlap into other volumes in the space, making it have a higher internal volume sum than external collective volume sum... making it (potentially) unprintable.) It likely also looks for things that can be made in a drawing program that can't be made in real life... like walls or partitions with a thickness of zero or two objects meeting at a "point" of infinitesimal size.
The standard says that models must manifold. So it's basically up to modeling software to enforce it. I predict we'll start seeing "Are you sure?" save dialogs.
@@diggis79 you dont have to export stl, you are just able to if you want to (which seems silly)
What I'd find more interesting would be a file standard which could be properly altered and worked on in CAD Programms like fusion etc... I'd love to slightly change around a few designs on thingiverse but that is an absolute pain in the bum with only an stl file which can't be imported into a cad Programm as a solid
That would be cool but CAD developers would have to adopt a standard (which has happened in the past e.g. DXF) but unfortunately i just don't see that happening. It's a great idea though.
Is there any print quality improvements? Are we going to have better prints as a result of this or are we basically adding another level of complexity on top of what's already commonly used?
I'm not sure you really understand STL .... Anyway size , scale , mesh orientation are saved and reflect only the work you did with your 3d modeling software. STL is good but do not save material/color information, but it is perfect to save shape just like an obj file ! This video is about promoting 3MF , maybe good for the new needs arising , but stl is also very good , it's simple and frugal. If you need a sculpt only , it is good enough
Can it be printed with different colors without making the extra steps?
I save everything I have dialed in as 3MF, except once you do that, if I change the STL then the "reload from disk" no longer reloads the STL but reverts to the last 3MF save.
That... sounds like what it *should* be doing, though...
Simon Belanger In Prusa Slicer when I load several STL, and start working with the modifiers, I might realize I should change my design on one of the objects. So I will go to Fusion360 and fix it, export to STL (overwriting the original), and when I go to Prusa Slicer I right-click that object and reload from disk. It will update it and I can proceed. Once I save the project, it doesn’t work the same way anymore and will reload the whole 3MF file instead of just that one STL. I just expected it to work like it does before saving the project.
PrusaSlicer suffers from many other similar inconsistencies.
Well I would use .3mf but Fusion 360 doesn't support this file format, direct transfer to Cura which I use often, doesn't even allow you to change the file format...
I have found .stl does keep scale info. And all the other features don't matter since it isn't edited in .stl. that is the final format.
Fusion 360 Personal Edition appears to Export but not OPEN 3MF files. Am I missing something?
On the STL's that I spend days on running and getting right (scaling, rotation, position, supports, etc.) Through multiple test prints... I absolutely save in .3MF. In fact, I need to just get in that habit now that I've reached the point of needing to reprint some objects (Such as a collision that ripped my fan shroud off, so I needed super glue and duct tape to print a new shroud).
I am a freelancer 3d modeler, and I have a bunch of clients who work with 3d printing, and sometimes they send me an stl to edit something, and most of the time I have to remodel it from scratch, since the stl has a crappy topology... This new format will save me in these situations
Wow nice and clean explanation I will defiantly use and move to 3MF got me wondering for a while, Thanks
As I don't get my printers for a few months I will try both on both FDM and SLA printers in the future. Thanks for the info.
There is always AMF developed by ASTM and ISO, which was released before 3MF?!
Probably a stupid question, but just verifying, when you export to Gcode will it still be a "normal" Gcode? So any printer will still read it as it does now, so like on my MK2s? I don't see it being an issue but just making sure
Well I design all of my stuf in BLender and export them as STL to load into Slic3R
There's no reason I would need to export something from Slic3r OTHER than Gcode, right ?
I'll use 3mf when I can import and export in design programs, like blender and fusion 360.
Is there a support for storing UV coordinates (per triangle)?
If I load a model that is just faces (not solid anywhere) will it print nothing? I modeled a car a while ago for CG purposes, but want to print it... haven't been able to find a way to get it solid.
The issue is that many CAD programs don’t have exporting to this file type yet. It’ll be a while.
That's ok! Export STL - Import it into a PrusaSlicer and save 3MF with your slicing settings. The transition from CAD to slicer can stil be in STL, even though it's not optimal.
@@Prusa3D what if my additive manufacturing method is not fdm?...
Actually many of them do today: www.3mf.io/adoption/
Looks cool, but my modeling apps don't support it without a separate outside conversion, and so far haven't done anything large enuff to have a need for it (yet).
Windows 10 shipped with a stl to 3mf converter :P I am not sure why it hasn't taken off yet.
Near oct 2021- still using STL
I am interested to find a format where we can share it to the community without worrying that it would be modified or stolen.
I'm only sad that the ISO standard .AMF format is the one being ignored where the corporation-controlled .3MF is the one that is that is taking off.
Still, market forces..
If 3MF is open-sourced, it's available to everyone for free. But, honestly, it's the first thing I heard about either.
So I'm curious. What are the advantages of AMF?
When we introduced our new PrusaSlicer 2.0 and launched the new PrusaPrinters community hub, some people asked us why we prefer sharing the models in 3MF format over AMF. It’s simple: the 3MF specification is newer and unambiguous, and it’s available for free. Also, 3MF files are supported by Cura and Microsoft 3D builder (among other apps).On the other hand, AMF is ambiguous and the specification is to be paid for. AMF files were never properly supported by any software that we know of. At best, AMF files were used to the same extent as STLs, meaning that the AMF only carried a single mesh without any additional attributes. With that said, PrusaSlicer can still load AMF.
( - Mikolas)
What if I want to have a mesh with a non-manifold edge?
I always export to 3mf inm solidworks, but it's not like that helps because there are no slicersettings added to the file yet..
what is this number format at 1:34 for file sizes...
Europe uses commas ad decimals e.g 1,00 seconds etc
Some CAD packages aren't well suited to 3MF output (yet); because we use both Ultimaker Cura and PrusaSlicer, asking one of the CAD designers to give me a 3mf wouldn't make much sense. That said, if I think I'll need to print a part again, you bet I'm saving that sucker as a 3MF file.
In my opinion it is a bad idea to store slicerinformation INSIDE the file. It may get outdated as soon as a new version of the slicer comes around. So you would have to update your shared models over and over again. And which versions slicer information will you store? the latest one? All? Will get pretty crowded soon.
It would be cool if you could set author/url/description metadata from within prusaslicer. would greatly increase the visibility of this format
Blender 3d doesn't support 3MF, which is where I do the bulk of my designs
Same, but it's not really a big problem. You use STL for the transition from Blender to slicer, but then you save 3MF in the slicer with all of the fine-tuned settings (supports, variable layer height etc.) and that's what you share.
Blender -> stl -> ps -> 3mf + gcode is missing the elegance for iterations. It would br awesome, if ps could generate a blender mesh based on the gcode info, so it can be considered when optimising and iterating on a design - especially when details and tolerances matter (changing nozzle size, width, heigth and even material changes dimensions of the final object)
But it is an unnecessary hassle, and an annoyance when you design something with tight tolerances, and mesh is scaled down or up by a factor of 10
@@grogyan unless the tolerances are parametric and part of the model
@@Prusa3D Then why would this "kill" the STL format? That makes no sense because you are still using STL. Unless you completely and 100% eliminate STL files, then you are just treading water and splashing it at people.
Generally when someone says something is dead to create a clickbait title... it isn't.
I am a beginner in 3D printing. Could you please tell me which software to use to create a 3MF model? Ideally should work on linux. (Could be also paid software.) Thank you very much.
XML :(. Why ... Just why ... XML has so much useless information overhead.
im not an xml fan. it turns perfectly readable raw data into gibberish. i usually just stick a bunch of key value pairs in a text file with some sort of separator. the complexity of the format should reflect the complexity of the data being stored. using xml to basically store what amounts to a list of triangle vertices is overkill.
Josef, when do you expect to ship out the minis once available?
Null Null Feb 2020 for new orders, see help.prusa3d.com/article/6OdtbKF1bw-shippingtable
I disagreed most what you said in the beginning.
For example
A model don't have to be self validating. You just need an external validation tool. And don't have to be readable, again you can use tools.
See how flatbuffer is so popular
Solidworks saves the current view as the preview picture when exporting in the 3mf-fomat. Sweet! Thanks for sharing, will use from now on.
Is there a good video learning Prusa Slicer? I never used the features shown in this video but they can be very useful.
Hey guys coming back to this that has not taken a lot of love in printable sadly. It could be nice that when uploading in printables there is a pop-up that suggest the step and 3mf format. And that if the person uploads an stl offer to transform it to 3mf automatically even without the printing settings just the stl inside of it. This could be nice to add. As more people see the format around more people will use it. A good way to push the format. (eventhought step is still a superior format 😅)
I've been using 3mf ever since I was designing and printing!
Thank you again for sharing your video!
Hmm.. the instantiation referencing in the language is definitely going to be a GoodThing(tm), as well as the ability to avoid manifold issues. The Zip'd XML file structure also sounds promising.
However, the embedding of printer related settings.... that is going to be something that is hit/miss depending on the printer being used. Ie, for folks using a fully standardized printer like a Prusa, it will be awesome. For others, I'm guessing the settings will be a sort of starting point... I see this as a huge benefit overall over STL files though. Hopefully adoption of the format will be widespread... but I also hope that the file format will remain something that can be edited/imported/exported.
I do not dispute the technical argument. The user base is not the people that need convincing. You just need to persuade the ecosystem of CAD and slicer software manufacturers that it's worth switching and the user base will follow.
Eh. If it was not based on XML, but instead on something more sane like Protocol Buffers, I would be in. I will stick to STL, fast and super easy to write parsers and generators with zero external dependencies. Also using zip, another meh. There are better, faster and more open standards to use with less quirks. Adding instancing and multi mesh support to STL is super easy. No need for new format for that. Comparing size of 3MF with raw STL without compressing it same zip compression settings is just silly and missleading.
What if I dont want tooo use Prusa slicer or Prusa printer... ?!
Then don't! :) 3MF is supported by almost every slicer on the market.
- Mikolas
@@Prusa3D o.k Tyou 👌
where do i export 3mf if i use tinkercad??
Yea I see the addition of identity into a file as a way for the gov to try and control additive manufacturing all they have to do is require identification info in all transmitted 3mf files so they can " track" where it came from
*puts on tinfoil hat* you realize if you wanted, you could put the metadata into an STL. I really don't see what's wrong with having a field for the designer's name.
I think 3MF is nice, and i used it something in the last months.
But, where there www.prusa3d.de/prusa-i3-druckbare/ are no 3MF Files?! Only STL and some GCODE..
Can you provide 3MF for all?
Does m3f improve the print quality?
XML... Really? And you say it is human readable somehow.
Stl preview in explo from windows plz? for stl i know allready marlin tool
I will try for sure! Thanks, Subscribed!!
Xml and human readable on the same sentence? What a joke
I agree, but It's more readable than the stl binary file...
@@noahchristensen7342 Probably, but then why not use an even better format? I doubt any format will be a problem to parse for any machine
It's a interesting idea, I'm definitely trying it out soon!
I like STEP ap214, VRML 97, & OBJ most of the time. STEP ap214 is usually very small for sharing. The real issue with STL is how difficult it is to bring into parametric CAD software it's very lossy relative to NURBS file. Again, STEP is king, maybe even IGES. I print with color printers so i need more than stl file. The main benefit of this seems to be how it can't have non-manifold edges, but most of what i see is self intersections from poor design intent prior to export. If it is still mesh, then self intersections seems like it will always be an issue. Mesh formats will probably never replace SLC for certain applications. I love how many companies tried to fix this together but still seems lossy. Do you disagree?? Please share thoughts :)
Will you have some sort of black friday or cyber mondey on your products???
Thats awesome!! Is there any better use also for cnc router?
STL stands for standard tessellation language. Nothing else. Its predecessor was PHIGS+, Programmer's Hierarchical Interactive Graphics System (PHIGS).
So what happens if somebody designs non manifold edges (e.g. A house with walls of 0 thicknessin rhino) and tries to export as .3mf?
AFAIK the verticies should get doubled so that there are two planes each facing other direction. Which removes the ambiguity. It does this if an edge belongs to more than two faces. Sure, you'll still have an infinitely thin wall, but that's a separate problem. Manifold problems aren't usually thin walls but more of an incorrect topology that is ambiguous and each app can interpret it differently.
- Mikolas
Why use a mesh file at all? Why not just take a CAD file (.IPT, .IGES, .STEP, .PRT, etc) and load it directly?
That would be awesome, afaik no slicer in the world can do that atm. It's something that would be very hard to implement, but super useful. (I'm keeping on my wishlist with no date estimates :) )
- Mikolas
My main limitation to use .3MF is if Sketchup or Fusion 360 can produce the file. So far .STL is the one readily available.
I use 3mf, but it is a pitty that not all settings can be reloaded. When I make a projekt, I direct upload to octoprint. works great. But on a realod of the 3mf, there is no octoprint information and i have to export gcode manualy OR switch to the printer - but then prusaslicer will delete much settings while switching and my "relaod" is not the same anymore. that is really anoying and make me sad.
please correct that. THEN 3mf will rock like it could.
Thanks, and happy printing!
I use Creo at work for model creation. Unfortunately, when I looked the other day it doesn't support 3MF output. It's too bad because 3MF looks like the future of 3D printing files.
Hi there, Creo supports 3MF (support.ptc.com/help/creo/creo_pma/usascii/index.html#page/data_exchange/interface/About_the_3MF_format.html)
Most good tools have built in checking for manifolds. STL Check in 3ds Max was a requirement for stencil shadows, back in the day.
For modern mesh formats, we have FBX (de facto, not open,) collada (supported everywhere but buggy) and opengex (high quality, not as wide support as collada.)
Inventing another format is just the wheel turning over again, with the same bugs to come as the previous formats have already worked out.