Print with dozens of colors: Our new open-source ColorMix for PrusaSlicer
blog.prusa3d.com
blog.prusa3d.com
I'm hoping for the next innovation with mixed extrusion to reduce print times. We are lacking an automatic extrusion amount and nozzle size mixing within a "layer". Not just fine layers everywhere with mixed colors on the inside.
Goal: print the infill and inner perimeter from a larger nozzle and thick layer height. Use the fine nozzle and fancy layer-mixing only on the outside where needed. It is not going to be strict layers any more - I understand, this makes it difficult certainly. Then the Prusa printers could shine that exchange fully loaded and pre heated print heads quickly.
Until then, I'll happily wait for 2 days to get a spool of orange filament delivered.. Instead of waiting for a 20hour print job
The article is by Prusa Research and it's about a recent, novel 3D printing technique. This is very much an area of active research and also development (not to be confused with R&D).
The "FullSpectrum" thing is the name given to the project in which the developer, ratdoux, who is presumably an individual, demonstrated the technique. There's no Orcas in it, either.
The INDX has extremely low waste when you switch from one tool head to another. Just a little tiny nugget.
It also supports different sized nozzles in the different heads, like the Prusa XL.
So I suspect having far more people with access to that will help push better uses for that. PrusaSlicer 3 is coming soon with lots of improvements (according to them) but we don’t know what yet. I’m hoping that’s one of em.
Please accept 28 Advertisement cookies with 1+ year expiration to play one YouTube video...
It’s not like many companies in the world where the video IS the article and they just assume no one reads.
…
"… there’s a real path to a proper perceptual colorspace model in the spirit of Mixbox or Spectral.js – predictions that are correct by construction across the full ratio space, not patches on residuals. None of this is hard. The math exists. It’s a data problem."
…
"Not in some distant future. Now. The hardware exists. The slicers exist. The filaments exist."
(For reference: https://www.youtube.com/watch?v=ERgnSetWkEA)
Ffs, stop with the gatekeeping please. Yes, privacy matters to me. Nah, not doing all that. Instead, I reported them to uoou.gov.cz/en for this gross violation of GDPR.
Even better would be a mechanism like to Cerilica's Truism which would allow one to use arbitrary filaments and preview how they will blend when printed.
And you’re only limited in quantity by how many filaments you can load at once. A full INDX setup on a Core One is 8. I thought you could daisy chain Bambu AMS units on their printers, which would let you get your 16 maybe? I’m not very familiar with their offerings.
I’d hate to see the cost of all that though.
And it looks like the software support needs work too - the obvious way to do it is being done able to import a jpeg or png to project or wrap onto the surface, a bit like texture maps in video games.
There's an app called Hueforge that produces models that color mix to reproduce photos:
It's been around for years. There are databases of filaments with their TD values and color measurements to use. The blog really sells Prusa's attempts to do this with their own PLA, but there's a long history of color mixing in the community with extensive measurements of filaments that anyone interested should check out, too.
Will have to keep it in mind for when I get a printer which does nozzle/toolhead switching.
The print file will have pauses inserted at the layer swaps. You manually swap the filaments when it pauses. There are only a few layers and colors so it's not too hard.
primed3d can do photos onto models. still limited by print resolution, but very cool concept.
* the simplest is just mixing filaments, like one mixes paint. The article doesn't spell out the reason it doesn't work, I am curious as to why.
* together with alternating layers, colors could be alternated in the same layer. Some purging may be necessary, but I think you could either: accept some mixing (compute its impact to compensate) / take into account the volume in the nozzle (extrusion "latency") / discard the unwanted part in the infill (at the cost of less smooth edges)
Of course, the hard work with any approach, including their current work, is calibration, as the article highlights. I wonder if off-the-shelf monitor calibration sensors could help with measuring the filament you have at hand.
This requires you to control both filaments independently directly at the extruder. Dual direct drive for a single nozzle sounds like an engineering nightmare. The extruder head is going to be huge.
There is also the obvious problem of how to stirr the filament. Printing temperatures aren't hot enough to turn the plastic liquid, they just make it soft enough to drip out the nozzle. This means you can't just feed the filaments at continuos rates, you will have to use a PWM scheme where you extrude the first filament and then the second filament in extremely small discrete increments. That switching will give you the necessary agitation without building a throwaway nozzle that can't be cleaned after a clog.
All of this sounds like it would take at least a year for a well equipped research department to figure out. It's definitely not the simplest solution.
EDIT: https://www.reddit.com/r/ender3v2/comments/ssuw3i/my_crazy_p...
Just the hot end of this extruder costs $70 alone. This is definitely not going to be cheap to do.
You would need special nozzle geometry that encourages mixing, and likely much higher temperatures. And any such mixing geometry will trap some filament. Switching from dark to light colors might require purging with meters of light filament to get all traces of black filament out
I wish I had more free time, a gearbox or a 4-way differential can probably be used on the extruder head to control the mixing ratios.
Regarding the necessary research, I can picture it being cracked either by hobbyists, or industrial-scale (patent-heavy) players. I imagine that consumer-facing players such as Prusa do not have the means to heavily invest in R&D.
Plastic flow is laminar, where colour mixing requires turbulence. If you make a turbulent nozzle, it's basically impossible to print reliably with it (the pressure used to push filament out of the nozzle is mostly absorbed/redirect into turbulence).
That is actually the hardest way to do it, because that's not at all how 3d printing works. 3d printers take strands of plastic (aka "filament"), soften them up to being melty but not melted, and then "extrude" them, like cake frosting onto a surface. As with cake frosting, in order to mix colors, you have to do so before the extrusion step, so you would have to make your own (filament), and the machinery to do so is not cheap.
The thing about first order thinking is that it is very rarely useful, because the actual experts in the field have almost certainly thought of all the things that first order thinkers come up with, and deemed those ideas unworkable for various reasons.
Sure, but a useful article would focus on explaining the consideration and rejection of those obvious ideas, and what actually had to be done to implement something similar — rather than focusing on the even-more-obvious background material (how 2d printers have worked for decades) motivating the obvious ideas.
If you want a discussion about the other ideas and why this specific method was chosen, go to the original tool where the creator discusses those things.
They're just putting the technique into their branded slicer which should make it easier to access for people who don't like using OrcaSlicer.
No this isn't rocket science, and there's definitely a vibrant FOSS community actually pioneering this and that is probably the best place to be on the true frontier, but there is productization effort here. Considering people always advocate for Bambu for "making it easy to buy", Prusa also deserves credit when they try. They certainly get knocked when they don't.
As someone deeply embedded into the FOSS community myself, it's sometimes really annoying when we sabotage the better players. It only helps the worse ones.
PrusaSlicer is used as a base by some others, they’ll get this.
Version 3 is coming soon, they’ve promised good things. I’m curious what shows up.
They also open sourced their color mixing model so if people think it’s better they can switch. And they’re using and adding their stuff to the open print tag database they’ve already cooperated with others on.
This seems like all upside to the community to me.
For what it's worth, CMY filament bundles have been available forever and they're well characterized for use with HueForge there are open databases with measured color and TD values. It's great that Prusa is launching their own bundle with their own measurements. I'm just trying to point out that this all exists and has existed for a long time, and there are multiple resources available for it.
> As someone deeply embedded into the FOSS community myself, it's sometimes really annoying when we sabotage the better players
Not trying to sabotage anyone, just trying to help the community with some more information.
INDX already has fast color changes and produces far FAR less waste than an AMS. And that’s what sold me.
Then the coloring mixing stuff started coming from the community. Now you don’t need to buy 30 colors of filament for many uses. Thats a serious upside. And it really benefits from multiple toolheads.
It’s a great confluence of events if you’re in the Prusa ecosystem or just don’t want a Bambu or U1.
I do agree though that direct feeding each tool head offers the best experience vs the AMS approach.
I’m glad to see Prius’s catching up to Bambu on the color mixing front, Bambu has had CMYK filaments for a long time and has supported color mixing in their slicer for at least a month.
INDX no longer needs a tower. They say there is 13 milligrams of waste (which they call less than a grain of rice) on each filament change. So a print with 1,000 changes wastes 13 grams of filament. Details:
https://blog.prusa3d.com/prusa-core-one-indx-orders-now-open...
So... honestly it's kinda silly IMO. If I wanted to mix colors in a hacky way I'd just... wait, hold my beer.
I'm excited for INDX too, but like you said it's not a Prusa exclusive system. I think this is great news for people who like to play entirely within the Prusa ecosystem, but I also think it's good to let people know that there are a lot of options outside of that ecosystem.
The Snapmaker U1 is looking good at $899 shipped for a 4-color printer with no waste https://us.snapmaker.com/products/snapmaker-u1-3d-printer
I guess I've been in 3D printing long enough to not consider myself in anyone's ecosystem. I'm jumped between printers and had multiple printers at home for years and never felt tied to any one ecosystem.
I started with a MakerBot Thing-O-Matic when it was new. After leaving the hobby for a while I got a MK3(S?) because it seemed the most reliable to me at the time.
I’ve stuck with Prusa since then since I’ve been happy and I’m used to their stuff. I had a MK4S because I could upgrade my existing printer to it. The next time I was interested I didn’t see anything I was more interested in so I stuck with what I liked. There is also a “known quantity” factor in the company since I’ve dealt with them a few times now.
I would leave them if I saw something I liked better, but haven’t at a time I was buying. I do like the look of Bambu printers much more. I looked at their printers for a minute when I wanted to replace my MK4S but didn’t like it enough to bother. I didn’t get a Core One until after the INDX was announced, and that was a big part of why I decided to upgrade.
If I was starting over the U1 would be a VERY big draw for me. Price/capability is very impressive. Prusa’s low end is quite lacking as far as I know.
I understand OSS people don't like Bambu, but as pure end user, they are great and well put together.
This might seems like a lot but this is the reality of the system.
And it also takes time. The difference can be a few hours vs a day of print time. Plenty of videos online that show case examples.
It can be multiple times the weight of the actual thing you’re printing. Exactly how bad depends on the model. Here’s an extreme version:
https://www.reddit.com/r/BambuLab/comments/1jluvde/35_g_mode...
It makes one of those little color blobs every single time it changes colors.
I bet people somewhere find some very interesting special effects all this could bring with the right odd combinations.
When I first saw this pop-up in the community it was clear this was a fix. No it’s not as good as owning a roll of some specific color, but for a ton of use cases it’s absolutely good enough or maybe even perfect.
It made me want an INDX all over again, now thinking I should buy more heads. I knew they’d jump on, I’m glad it’s this early so it will be available by the time mine arrives.
I’m sure this is a huge boon for them, Snapmaker’s U1, and the new Bambu with more than 2 heads. HUGE value add just through software. Speed difference between those and MMU/AMS is now more important than ever if you want this.
I had written it off, because of how irritating Mosaic’s cloud slicer is. I’ve been pleasantly surprised how well it works with a fork of P2PP (vhspace/p2pp), a post processor for PrusaSlicer that is completely local. All it does is swap out the filament change commands with Palette splice instructions. I fixed a few gcode interpreter bugs that solved the issues I had with bed calibration and extrusion, and even splicing seems quite reliable. I’ve been using it for simple 2-color same-brand prints that would require a lot of manual changing, so not complex but it’s much nicer to use than an MMU2.
I’m excited to try this out for sure.
I’m sure the same kind of thing would be possible using Prusa’s documented methods with a little extra work.
Seems like the volumetric extension of 3mf files could support it. That would make cross slicer file mgmt easier.
... That doesn't sound like something that should have been difficult to achieve? Has it really not worked this way before in 3d printing?
The "Prusa ColorMix Cones" model is not what I'd recommend. I don't know why they made it like that for 3 and 4 colors other than to do something different than what the community was already doing. For 4 colors the PeggyPallette mini they used as inspiration is a much better model: https://makerworld.com/en/models/2519356-peggypalette-mini-3... You specifically want the dome shape to visualize how the layers blend at different angles. The fixed angle of the cones in the Prusa model misses the point and I don't know why they did that other than to be different.
The article goes on at length about their filament mixing model which sounds cool until you see the part that they only tested with Prusament PLA. Again, I think the open source community was already doing a good job with this.
There are several filament databases other than the one they're linking to that have TD values, which sprang out of the HueForge community. There are cheap tools from small makers to measure TD and color, too. One database: https://3dfilamentprofiles.com/
I'm glad they gave attribution to some of the sources of all of these ideas, but to be completely honest it's getting a little tiring to see everything the open source communities do get wrapped up, prefixed with a Prusa- brand prefix, and resold to us. Make sure you look beyond the Prusa official everything to get a sense of what the community is doing with all of this. I know I'm going to get downvoted for saying anything that isn't 100% pro-Prusa, but this is one topic where the open source community is quite a bit ahead and it's worth looking at what's out there.
Yeah they’re focusing on their filaments but that’s their product. It would be weird if they didn’t start there. Plus they’re working with the open filament database that as established to go with the open source NFC tags they cooperated with other companies on. That seems sand too.
Ok they’ve only done PLA so far, is that such a big deal? This is an announcement not a release. They’re still working on all of it.
And it seemed to me like they did a great job giving the community credit in the post. They made it clear it all came from the community, including the entire idea, and they’re building on that and giving back.
The problem with their models is that it's a cone shape. The angle is fixed.
The community models are domed so you can see the effect at different angles.
> Plus they’re working with the open filament database that as established to go with the open source NFC tags they cooperated with other companies on.
Right, but they're steering people back their sources when the community sites have more user-submitted coverage and it's what we've all been using successfully already.
I probably shouldn't have said anything given the topic and the audience. I know they gave some credit, but there's a long history in the 3D printing world of this stuff happening.
That’s just a guess. I’m sure there’s a reason.
It’s a reference to that.