So, I’ve been busy. I finally got my scans of the horizontal mixer, and have been trying to get the final mesh to move over to F360. Keep running out of memory. Right now the combined files (5 scans) are 255gb. I did not thing that thing created so much data.
Anyhow, I will keep trying to reduce the triangles and anything else I can to reduce the size so the program will finish and get me a model.
I have an original POP that never worked great for my use cases. The MetroX hopefully resolves those problems with blue laser scanning. I backed it yesterday.
I started down the 3D scanning rabbit hole with some prosumer Revopoint3D scanners (mini and pop 3), and ultimately ended up buying a $3,500 used Faro Edge scanarm from a property disposition auction.
With handheld scanners, I got really tired of spraying and adding markers to parts, only for the scanner to constantly lose tracking.
With the faro scanarm, the contact probe and line scanner use the encoders on each axis to obtain the absolute position/orientation, solving the tracking problem entirely, and the laser line scanner is fairly agnostic to surface color/texture compared to the handheld structured light scanners.
The downsides are:
This type of setup is crazy expensive, a brand new setup might cost north of $50,000 and a used one like mine over $10k. Software can cost over $5k/year just to get data from the device.
They can be awkward to orient around a part, and they’re pretty cumbersome to move around.
What ultimately doesn’t change is the post-processing and reverse engineering workflow going from a mesh to a solid model, and even the expensive softwares require a lot of experience and time invested to get good results.
I used to use a Romer arm at work, very similar to the Faro. Agree with pretty much everything you had to say, but am hoping the blue laser scanning with the MetroX solved most of the problems scanning bare metal and black surfaces.
After watching the edited livestream for the MetroX, I’m debating whether to still fund it. It still seemed like a really slow process to scan, and still used a ton of reflective markers. The scans looked pretty good, even the shiny muffler, but I’m worried it’ll be just enough of a hassle that I’ll avoid integrating it into my work flow when reverse engineering things.
I’ve got some time to decide, but I may wait for the next iteration of blue laser scanners in the hopes that scanning speed goes up, and tracking marker use goes down.
The ‘boss’ actually approved me to fund it but I backed out when I saw that the software is $486 annual subscription. I don’t know if that software is required but I am assuming it is.
I do have other reasons to drag my feet on this. My computer is stretched pretty thin so it is doubtful that it has the horsepower for all the bytes needed.
I have a Lenovo laptop with Xeon W-10855M processor @2.80GHz and 128GB of RAM, 5TB of M.2 storage, running Windows 11 Pro for Workstations. I’m pretty sure mine is fine for scanning purposes from a raw horsepower standpoint.
I am however getting that twitch in my gut that wonders if I’d be better off waiting till the next product they design using the blue laser scanning so they have a chance to work out the bugs and get it working even more reliably and give the software another year or two to mature even more.
Any of you guys try “scanning” with Polycam to create models? Watching some videos, I’m impressed with how far photogrammetry has come. Hard to say if it’s worth paying for yet, I’ve got to find a couple things I’m itching to scan rather than throwing away my free scans on things I don’t care about just to try it.
Revopoint has a kind of annoying habit of burning early adopters by quickly iterating through versions of products, so I’d probably wait for the second generation.
My original Mini was frankly unusable given the state of software at the time, and I eventually bought a Pop 3 which was then replaced by the Pop 3 Plus less than 6mo later.