I love revopoint and own the pop2, pop3, and range -
Two questions for the community:
Recently I grabbed the Miraco Plus with zoom - as stated in some posts, the zoom honestly doesn’t seem to give more detail. I’ve tested it a lot.
Revopoint documentation says it focuses the light to reduce noise or problems in detail areas? But to be honest, I notice no differences whether on near or far mode.
I’m trying to scan full size human body - and not in A- pose (much easier and recommended), but I intentionally need to scan in T-pose.
I read PUTV’s posts here, bought a turntable motor to rotate people as they pose, which improved the results some. However there’s a real problem: when I do far mode Miraco on high accuracy, and only scan someone’s face, the detail is good, and the point distance can go down to 0.15. When I do a full body scan, again on high accuracy, the lowest point distance on my computer is usually around 0.57. I have 32 GB ram, but I’m guessing the problem is the ram is not enough.
If I get more RAM, will revoscan allow me to do smaller point distances on the same models?
If I don’t have more RAM, should I divide the models up, fuse them separately, and then merge them instead to get around having less RAM?
Are there 3rd party programs, such as CC, that would allow me to force the computer to go to a smaller point difference during fuse, but just take longer to process?
My goal is full body scans with point distance 0.2 at least.
Thank you guys for any thoughts on these questions.
I’m assuming you mean the point distance limitation using Advanced fusing. That limitation is based on the volume of the scan not available resources. Standard fusing doesn’t have the same limit. Give it a try and see if you get any benefit from it. The point clouds are not as processed and clean but the meshes aren’t as bad as the point cloud would suggest.
You may find more details, but you may just find more noise. Sweet spot is probably something less than the limit on Advanced, but not too much less. As someone with dubious skills who scans a lot of larger things, I find Advanced can gloss over some errors in the frame alignments that are revealed in Standard, which can be good or bad depending on your circumstances.
Hmm ok so if advanced fusing is based on dimension - what about - is there a way for me to slice the cloud into smaller sets, fuse them together so I can use smaller distances with advanced fusing on each, before re-combining them?
I searched the forum and didn’t find that, but I’m guessing I could just duplicate a point cloud, then using a box section to selectively delete pieces of each point cloud, fuse them separately on advanced then merge them?
I appreciated the suggestion to use standard fusing, that definitely allows it to go lower, but then it requires much more clean up.
I’m in support of going lower fuse distance in some cases, where you might want to pick up on those surface details (even accepting the noise and longer processing times), especially if you’re going to re-sculpt by hand in Zbrush anyway and using it as a rough starting point and not trying to create a finished product from a scan - the noise or scan lines are not something hard to sculpt away, but the lack of detail is a loss of rich organic information, especially if you have practiced anatomy, etc. (because you can kind of see through the noise). Do you know what I mean? I don’t know enough about how the advanced mode is programmed to know if this would matter, but it seems like allowing us to bypass the safety and go lower on advanced mode could reduce some of the work of cleaning up the standard fuse.
Unfortunately I can’t figure out how to slice the point cloud before fusing, so I can’t accomplish the advanced fusion of each section. Other than scanning each cloud separately to start with, is there a way to divide point clouds after scanning before fusing?
Not saying I agree with the somewhat arbitrary floor on point distance with Advanced fusing. Early versions with Advanced didn’t have that restriction (way early, so don’t try to find them as none of the current scanners will function correctly).
You could duplicate the project and use Frame Edit Mode to segment the scan. Even if that works well for the individual parts, there’s a bug (or feature that shouldn’t be) where merging applies smoothing to the result.
One more thing.. the total volume includes extraneous points as well. So if you have extra points of any sort that aren’t part of your subject (even just some noise from who knows where) they will contribute to the volume.
Here’s an extreme example of capturing on a turntable with some junk just beyond the turntable. Before removing the extra points, Revoscan suggested .8mm point distance, and after removing them in Frame Edit Mode the suggested point distance dropped to .4mm. Note that you’ll likely have to reload the project to get the new volume calculation.
Oh nice! I did wonder that, as in, was there any volume or it’s pure bounding box, and it looks like bounding box from your experiment. Thanks for sharing!
Early versions with Advanced didn’t have that restriction (way early, so don’t try to find them as none of the current scanners will function correctly).
