Upgrading from Cricut
This is more aimed at professional folks I think, or small business from hobby printers.
My wife is hitting the limits and frustrations of using Cricut Explore 3 to cut her heat vinyl projects. Suddenly, they've released a new machine type and adjusted the software they use which makes her machine cut faster (yay) but now it's more sloppy and not cutting as well as it did (it's not the blade, I repeat, IT IS NOT THE BLADE!)
Her frustrations are becoming mine due to the earache. Does anyone on Tildes do heat transfer and normal vinyl work? She's also looking at printing to vinyl roll and cutting; her aim is mostly tees and teddies. Due to this, loading a full roll and having a lot of waste isn't ideal.
We're after advice and what her upgrade steps are. We looked at printing to film (DTF) and garment (DTF) options, but we are definitely going to stay with vinyl for now.
Thoughts, advice, options, anything of value to say is all welcome. Not to self promote or anything, but to give you ideas of what she makes Https://thunderlizard.co.uk/shop and look at near any item. You can see it is like 2x4", 8x8" and max usually 12x12" sizes and a mix of vinyl types from plain through to glitter. Hopefully seeing that should allow for better and more accurate advice.
Thanks all!
The bit about the machine being faster but sloppier makes me think they just adjusted the speeds to the point you get lossy movements. I have a 3D printer and laser cutter and I can boost the speeds up to the point there's very jerky turns and wobbliness - I bet that's what happened.
Looking a bit online, it seems like cricut is very closed off - can't use external software, it's all online only, etc. An alternative mentioned on the Lightburn forums (Lightburn is used for Laser Cutters, it's an amazing piece of software) is Sillhouette Cameo (from https://www.silhouetteamerica.com, it looks like they have other newer models now). They also haven't sued other devs for daring to create software that works with their machine... but their own Sillhouette Studio software is free and decent.
As an alternative, (though I'm not sure how feasible it is with how closed off the cricut seems to be), you could try downgrading the software to the stable version, or see if there are 3rd party fixes for the issue... Can you even access the base machine settings somehow? setting your own speeds, etc?
Thank you for this comment. I think we've exhausted Cricut and it's just not up to standard. They change things whenever they want and the knock on effects everyone of their users.
I think what I want from this question is the best machines and software to move to next, away from Cricut.
I don’t have any comparative experience on other hardware, but I can vouch for Silhouette in terms of controllability, at least.
I came at this from pretty much the exact opposite end of the spectrum to your wife: I’m a software guy, needed something I could drive directly to do some small batches of precision cutting work, and ended up with a cheap second hand Silhouette Portrait 3 based on price and open source support.
Never even installed the first party software, just went straight to a python based CLI package and had full control of every setting right off the bat - it did take a bit of initial fiddling around (such is the nature of niche open source!), but once that was done it was just a matter of dialling in some speed and depth parameters for the material, feeding in an SVG, and watching it run. That’s exactly the kind of control and automation I’d be looking for if I needed to rely on one of their beefier machines for business use.
Not sure what settings Cricut does/doesn’t give you, but having granular inputs for the Silhouette was the difference between unusable and near-perfect for me - I was cutting vintage IBM punchcards, so a bit thicker and higher drag than vinyl, with tons of tiny sharp cornered cuts down to a millimetre or so from each other. First few runs either shredded the closer “punches” or left far too many hanging - which then shredded the ones next to them when I tried to pop them out with a pencil.
Ended up setting it up to run at half speed, fairly low pressure, and use a printed reference mark to position and re-run the same pattern twice over - the accuracy was high enough for it to happily go over the same lines twice when using reference marks (and it absolutely was not without them!), the final results were perfect, and the ability to automate and dial in all those weird little details is what made it work at all.
I would love to hear more about what you're up to with vintage IBM punchcards, if you don't mind sharing 👀
Very happy to, it's a fun little story, although probably a good bit sillier than you or @Nsutdwa were envisaging! Basically a passing idea for a gift that spiralled well past its logical conclusion.
A friend and I have been giving each other increasingly elaborate and inconvenient birthday presents for a while, to the point that the puzzle itself is really the gift nowadays, and one they gave me a year or two ago involved needing to track down a working floppy drive to access a key piece of information.
So, it's clearly a matter of honour there for me to escalate it to absurdity, and I figured it'd just be a matter of finding the right niche Etsy seller and a week later I'd have a punchcard with "The eagle flies at midnight" or whatever to use as a suitably opaque clue. Except there's nobody offering custom messages on a punchcard, anywhere on the internet, at least as far as I could find at the time - but there are a few people selling boxes of new old stock cards from the 70s for pretty reasonable prices (makes sense, there must have been literal billions of the things made), so no big deal, I'll get those and find a little rectangular single hole punch on Amazon and do it by hand. Bit time consuming, but it's only one card.
OK, now I'm doing it by hand so I need to make sure I understand what the character encoding on these things is - down that rabbit hole, lots of interesting reading on computer history, the various encoding standards (it turns out there were many), the binary versions, a whole stack of other research that I didn't really need to do but absolutely caught my interest! At this point I've also found a couple of web pages and github repos that do the text in -> punchcard out part, albeit only in digital image form, including one that happens to output SVGs.
I know what my punch pattern looks like, I have the physical media ready to go, just need to get one of those little handheld punches like train conductors used to use, but make sure it's got a rectangular hole of more or less the right size. Nothing, anywhere. Weird. I wonder if there are replica keypunches on AliExpress or something? Nope. Well, how much does an old one cost? Nope! I can't be the first person geeky enough to try DIY punchcards, there must be someone out there who's written a blog post about whatever tool they ended up hacking together for the job...
Well it turns out that most people with an interest in DIYing punchcards circa
current yeararen't doing so from a very, very out of hand rabbit hole around vintage computing, but as a way to run knitting machines, which apparently sometimes still use them to physically encode patterns. And while those use round holes, they use a lot of them, so doing it by hand is really not ideal - a few enterprising people in the overlap between crafting and coding seem to have had good luck with cutting plotters to automate it. Awesome!One trip to eBay later, with the extremely reassuring discovery that there are a ton of people who have old cutters that've sat in their craft cupboard for however long and are now getting cleared out for very little money, and there's finally a workable plan: I've got far more cards than I need (a box of 100 was only ~1.5x the price of an envelope with five of them), an SVG template, and now a way of quickly cutting as many as I like. And obviously, if I can do as many as I like then it makes it totally worth the whole up front process I've just gone through!
Rework the original idea a bit, back of the envelope calculation on how much data it'd take to encode a one-line script that'll display a QR code in the terminal, write some python to encode that, spit out the SVGs, and pipe them across as cutting paths - done!
So, that's where we're at now. My friend* has a small deck of punchcards and apparently still hasn't discovered that there's a library out there that'll decode them from a photo. I've got most of a box of the things remaining and I'm adding ideas around absurdist retrofuture home automation controls with an arduino-based optical reader to my ever growing list of projects that I'll absolutely definitely get to just as soon as I have a moment. All is as it should be with my small, very geeky corner of the world.
*They do work in research-level computer science, I'm not quite cruel enough to do that to a normal person
Absolutely seconded, feels like there is a good chance of an interesting hobbyhole there!
Just based on reviews from others in the laser community, I can recommend giving Sillhouette a good look - I've not used any myself, so I have no personal experience
I have a Silhouette Cameo in the makerspace I help run, and the thing has been absolutely bulletproof. Highly recommended.
I thought about getting a Cricut a while back but am glad I didn't because they've increasingly locked it down. Also I didn't really like the quality of the work it does. I tell the story of how a 3rd grade art class has maybe 2 kids in it that do really nice cutting and the rest of the kids are so-so, and the cricut cuts like one of the kids who isn't at the head of the class. It's sad that it's gotten even worse.
I've been super happy with my Silhouette Cameo 4! I never got a Cricut due to the whole fiasco with the software a few years back (you had to be online and couldnt really save locally). I've had zero issues that weren't purely user error from being a beginner. I still have it years later and it works good as new.