I have a fairly new desktop PC (Asrock Deskmini, AMD 4300G, 16GB DDR4, 4K60hz 3840x2160, Win10 Pro 21H2) All the Revoscan 5 releases have worked well on my dekstop.
But for scanning away from my desk I want to use my old laptop Dell Precision M3800 (Core i7 4702HQ, 16GB, Nvidia Quadro K1100 2GB, 4K60hz 3200x1800, Win 10 Pro 22H2).
I have only managed to get Revoscan 5.01 to work on my laptop. Revoscan 5.02 and 5.06 crashed.
I have a Revopoint Mini which I bought in November 2022.
I would really like that “Home” button moved, I can not keep count how many times I’ve hit that instead of the “File” button.
Also I would like the ability to transfer multiple scans in the “Import From” option connected to my phone.
On the Android app Revoscan V5.0.4 I find the camera adjustment sliders difficult to use, I cant but think it would be a benefit to make these slightly larger . Phone is HUAWEI P30 Pro.
Overall though I am finding it more user friendly than the previous separate apps.
There seems to be a bug (or feature) that once you mesh point cloud and then return to the point cloud work flow - you cannot delete points using selection tools. This would be nice to make work flows between mesh/pointcloud independent. IE if I mesh and want to delete something - I can do it in the point cloud. As merging many scans its cumbersome to have to basically do a fake “Merge” so that the point cloud can be edited again.
Better tool-tips/descriptors for each filter/setting (ie for overlap - describing what the mm setting means and if a smaller number is selected the tool will be more selective or vice-versa). Also - with tools some tools are redundant between pointcloud/mesh workflows. I am not sure if better to do in Mesh or PC setting or if it matters.
when filling in holes manually - if we could have an option to delete neighbor points/triangles like Gom Inspect that would greatly improve filling holes due to using markers and reduce artifacts caused by markers (currently we can do manually but would be more streamlined if it could be done automatically as well.
Spacemouse/3D mouse support
improve memory consumption - perhaps look into different storage containers like Voxels. As project grows, the software gets slow quickly.
Adding a Coordinate system/alignment tool.
Adding performance options/settings (IE use hardware acceleration, increase/decrease graphic render performance), etc.
Otherwise - for a free software program - Revo Scan 5 is amazing. Once I learned how to use it - I feel less like I need other softwares to do what I want.
Hi @jheins3 I do not experienced that bug yet , you can move back and forth between Fuse and Mesh and use the tools to remove partial points .
It is independent, Fuse files are a different than Mesh , both of them are saved independently while processing .
That would make them both depends on each other and they are independent separate files you editing direct from your project folder .
After you create a Mesh from a point cloud , it is total new object file and structure . It do not have all the data anymore that the fused object do , not even the same amount of points , for that reason impossible idea .
You need to edit each point cloud scan before you merge it into full object , after merging you need to clean it again to make sure there are not overlaps left from the merging .
And even if you Mesh the merged object, you can still go back to the fused object and edit it and mesh again reflecting the new changes .
I believe you go back to the Raw files ( Frame cell Points ) that can’t be edited since you need to create first a point cloud to be editable . The software saving automatic the fused point cloud to your project folder and can be accessed and edited regardless of meshing or merging .
I’m at work currently so I cannot screenshot, but if you scan an item and mesh it, and you go back to point cloud operations, then on the left side you use box or lasso selection, once a segment of scan is selected, the delete button remains greyed out and not selectable.
Edit: basically what I experience is once a mesh is created, you cannot go backwards and make changes even though the data is still there.
Remember switch to Merge tab and back to Edit Tab if any of the edit function do not response , it has to do sometimes with the OpenGL that do not update correctly , old bug that is not completely fixed yet probably.
Any chance the turn table usage could be more “user friendly?”… the revo scan 4 wasn’t the best either but at least I had the option to set multiple passes in multiple angles as a preset and not do every run manually and its annoying as hell to have to pause a scan and switch to another window to pause the turn table to… just add the option to set the number of passes like when ppl want 4 passes they type 4 to passes and below 4 lines could appear each with select-able angle if they type 10 then 10 lines etc… also you could set a box for an option to stop turntable after finishing and reset or or not… the current system is a chore and annoying… sorry but it is worst than before… with the above options you could have the old and the new… and everyone could find what they prefer… Ah also in meshing the highest is 6 but in studio i had a 7 and it looked better with the scans I made.
Revo Scan version: RevoScan_V5.0.5
System : Windows11
Scanner: POP2
My feedback and suggestion:
Direct export stl and remember the last time saved directory.
Support customizable workflow, after scanning is completed, automatically delete isolated items and export stl file, at the same time support specify the file path and name.
Support specify the scanning time, click Start to finish it automatically at the specified time. ( The turntable completes one revolution )
Friendly to developers, providing api for personal development and calling.
Support correcting stl model by minimum bounding box algorithm.
Improve scanning stability, scanner sometimes disconnected, reconnection time is slow, and the software sometimes get stuck.
Revo Scan version: RevoScan5-5.0.6.256-windows System : Windows 10 (e.g. Samsung Galaxy S10, Samsung Exynos 9820, 6GB ram, Android 12) Scanner: Range & POP2 (e.g.MINI) Your feedback & suggestion: I think it would be very useful if it’s possible to add additional selection options to clean the pointcloud as well as the mesh.
One will be “brush” so we can almost click on individual points or triangles, and the other one will be “flood” when you touch one point and everything connected to that one, is selected.
It would be cool if a new mode can get programmed into the range where I can spot points manually similar to this machine (not a 3D scanner) on Kickstarter. It seems like it might have all the hardware to do such a task.
The Kickstarter project looks a lot like a cross between an old “David” 3D scanner and a “Proliner”,
Both are really old technology.
I am not sure why you want to use either when you have a Mini / Pop/ Range Scanner ?
The earliest digital 3D scanners in the 80’s used contact probes that physically touched an object thousands of times until the device had enough points to create a 3D model. Very time consuming , that was short before the 3D scanning revolution …
In the late 90’s I still used contact probe to “scan” gypsum casts of human faces for digitalizing . First you cast the face , after you actually paint on the cast the quad topology and then touch probe moving with the topology lines to complete the full face scan …
So yes, based on a very old technology and nothing new.
I was working on the scan furniture challenge (I didn’t submit because of the aforementioned issue).
After merging many scans and developing a large point cloud, I would notice the point cloud density would effectively halve or be 1/4 of original scan resolution. I NEVER simplified the point cloud, or otherwise edited the point cloud (only occasionally meshed to check my scan data). Usually after meshing the point cloud would shrink. This caused scans to be very pixelated and lack detail it once had.
I was using a Lenovo P1 gen 2 with 16gb ram and a Nvidia T1000 integrated GPU. It is possible that my hard drive was running out of room - but there was no notification of such in Revoscan.
If I was running out of storage, I think it would have been better to notify me so I could make room instead of deleting data.
If you run out of space , you would probably crashed in the process .
I never experienced that , remember that the point cloud will have overlapped points that need to be removed , so it can decrease to even 50% of the original size by removing junk data , it can’t shrink unless it was meshed on lower level or you used Auto option for meshing what is a blob compared to manual meshing .
The level 6 is not optimal in some cases especially when the scanned model is huge around 2 meters , but with 16GB of RAM you would never mesh a point cloud at 50% as it would require 32GB of RAM to be safe and higher level of meshing .
I made huge scans and they never shrink or lost details, volume or accuracy after merging or meshing .
You can send the point cloud to putvmail@gmail.com I will mesh it and let you know what maximum level you can get .
I think that is the issue , you meshed the point cloud back and forth … the meshing algorithms removing overlapped points and junk data from the point cloud .
You only can do that if the point cloud has a full project intact including Raw data , if you imported the point cloud to your project and meshed it back and forth , you may lose details in the process .
