I’ll go find it. Thanks! It just didn’t click that it was for that use.
1 Like
Revo Scan 5 MetroX V5.6.1.82
Win11 (all updates applied)
PC: HP Victus - Intel I7-12700H, 16GB RAM, RTX 3060 8GB
Hi!
I’m new to 3D scanning and don’t have much experience yet, but since I received the Revopoint MetroX, I’ve been experimenting to better understand its potential.
That said, I’ve noticed a few areas for improvement in the Revoscan app:
-
When I manually edit the mesh to remove areas that “dirty” the model, those removed areas are restored automatically after performing a new operation.
It would be great if manual edits remained permanent until I decide to undo them.
(I don’t know if it is just a bug) -
It would be fantastic if the software could automatically recognize and remove the markers used during scanning, streamlining the workflow.
-
Adding the ability to set the orientation and origin position of the scanned model directly in the app, without needing external software, would be very useful. Moreover, a feature to make a flat cut at the base of the model and define it as the reference base would be perfect.
-
Lastly, having the option to fill or add thickness to surfaces would greatly simplify importing the model into other 3D modeling software and expand the scanner’s usability in different scenarios.
Thank you for your attention, and I hope these suggestions will be helpful for future updates!”
2 Likes
Revo Scan 5 MetroX Version: Revo Scan 5 MetroX V5.6.1.82
Operating System: Windows 11 Pro 64bit
PC Specs: Intel i9-13900HX, 32GB RAM, NVIDIA RTX 4070, SSD
Your Feedback & Suggestion:
- Marker detection is very poor and need to be improved. It’s impossible to do global marker scan of my scanning frame with about 200 markers. No problem at all using RaptorX.
- Scanning speed is terrible slow, despite frame rate is quite high with >40fps. Performance need to be improved. Scanning with RaptorX on same system is extremely smooth. Just to avoid people saying pc specs is not enough.
- Scan quality and resolution far behind RaptorX (parallel laser vs parallel laser). Quality to be improved on MetroX. Less smoothing, more details.
- No LED color indication on the MetroX scanner to show far/optimal/close distance.
- Auto-Turntable mode is very slow. Why 3x projection is needed. Should be improved to speed up the scanning process. All projection pattern should be done in a row like normal structured light scanner are doing. One position could be done in 2-3 seconds from my pov.
- Featue request: Auto-Turntable mode should also have option to calibrate the turntable axis. Once calibrated at the beginning the alignment should be done using the calculated position. Such featue are included in most structured light scanner with turntable. E.g. Shining3d Transcan C.
- GPU selection should be supported in Revo SW. I need to force use of main GPU in Nvidia settings. In Cr-Scan it’s also possible to select the GPU if more then one are available.
- Better support for AMD CPUs. I have high performance AMD desktop PC (second system) where Revo Scan SW shows not that pc power is not ok but not optimal. I can use both system (Intel i9 and AMD Desktop) without any performance issue using Cr-Scan. The only reason can only be SW supporting the HW.
- Better scanning noise auto removal by the system.
- Feature request: Merge by manual marker selection. Like Cr-Scan has.
- In general overall performance (scanning speed, lag of real movement vs shown position in SW, marker detection, global marker scan performance, pointcloud resolution and more details in final result) is far behind my expectations and your competition. SW updates to improve scanning performance urgently needed.
3 Likes
Is the RGB camera being used to help determine auto exposure settings? I’m finding that it does not auto expose very well for black or red objects but maybe if it could use the full color info from the RGB camera to analyze the object being scanned it could better anticipate the correct scan settings. On the same subject please allow even more range to set the exposure manually because many difficult to scan items need the exposure set to the max allowable but it is still not enough for a good scan.
- Revo Scan 5 MetroX Version: Revo Scan 5 MetroX V5.6.1.82
- Operating System: Windows 10 (desktop), Windows 11 (laptop)
- PC Specs (desktop): Ryzen 9 7950X, RAM: 64GB, GPU: NVIDIA RTX 3060 (6 GB)
- PC Specs (laptop): Intel i7 8850H, RAM: 32GB, GPU: Intel UHD graphics 630 (GPU0) + NVIDIA Quadro P3200 (GPU1)
Feedback & Suggestion:
Generally speaking, the MetroX is great scanner, however the SW package still needs to be worked on. These are known issues and bugs to me:
- All computation runs purely on CPU. No issue for desktop, but borderline for laptop.
- Full field mode lacks data quality indicator, so you either have to stick to laser mode or deal with very low quality scans (unless you are in turntable mode, which usually produces quite good scans for some reason).
- Feature scan is pretty much useless, it loses track all the time and creates nonsense caterpillars. Any hand shake will end up with a nonsense caterpillar.
- Marker scan usually works OK (not great, but well enough). Sometimes it just jumps and create whole new plane. Since there is no GO BACK button, you must cancel the scan and start all over. This is very annoying. This sometimes happens even in global marker mode, which makes no sense to me.
- While scanning in marker mode, I often end up with holes in place of the invisible markers (markers hidden under scanned thing). No idea, how this happens, if “Remove marker points” in Fusion is turned off.
- Merge function creates random holes on concave edges, even if the original scans don’t have them. WHY?!
- Revo Scan 5 randomly crashes. Always during scanning.
- Revo Scan 5 loses tracking of markers and doesn’t see most of them anymore. The only option is to delete the scan and start all over. Usually it works fine in a new scan (I have a nice video of this issue). When this happens, it may cause a crash mentioned above. Strangely enough, it neves shows the “not enough markers” warning.
- Even if I fill all holes, the final STL isn’t watertight and has open edges. How is it possible?
- Revo Scan 5 forgets scanner settings between scans. It’s super annoying, especially, if you need to use Global marker tracking mode and it switches back to Market tracking and adjusts Preset point distance to some default value. NO! Keep my settings from previous scan. Especially if I do more scans in one session, it makes no sense. I don’t care, if it goes to default setting when I close the app and open it fresh, but it shouldn’t change the settings during one session.
- Spacemouse doesn’t work in Merge mode. It works fine everywhere else.
Video of stuttering - https://www.youtube.com/watch?v=DUHL2ZZjKGQ
2 Likes
Revo Scan 5 MetroX Version:** Revo Scan 5 MetroX V5.6.1.82
Betriebssystem: Windows 11
MICROSOFT Surface Laptop Studio 2, Notebook, mit 14,4 Zoll Display Touchscreen, Intel® Core™ i7,i7-13700H Prozessor, 64 GB RAM, 1 TB SSD, NVIDIA GeForce RTX™ 4060, Platin.
A lot has already been said about what can or could be improved. Personally, I would like the following: a kind of user-defined one-click. Ideally, one could save different variants, e.g., point cloud, isolation, overlap, and finally the creation of the grid.
5 Likes
I have had this issue a lot.
How many frames do you have? I usually need to do around 10k frames for something palm sized. Bigger stuff needs more frames.
I didn’t realise how bored I’d get scanning with the MetroX. Generating 10000 frames feels a huge waste of time and leads to a horrible user experience for me.
1 Like
Same specs as mentioned previously
Scanning a dash with chrome and black trim next to each other is a very poor experience. It is a difficult environment but it is touted as working well with shiny surfaces so one would expect it to work well
It would be nice to have different, user customizable, presets for the one-click editing. Almost like choosing a print profile, if you are familiar with the 3D printing world. Basically make it easy to switch back and forth between one-click editing parameters. I would use it to have a “draft” setting that will quickly process the scanned data so I can see if I was able to capture what I wanted without having to wait a long time for processing, a normal setting for general scans and a highly detailed setting that may take longer to process but would preserve geometry for reverse engineering use.
I’d like to see the speed of switching modes from Marker Scan to Feature Scan improved. Currently switching modes is rather painfully long, but not sure if this is a hardware or software limitation. Would be nice.
1 Like
I would also like to see the M button configurable for different tasks in the ui. For example, if I could map the M button to undo, that would be amazing.
Revo Scan Metro X 5.6.3
Windows 10 (19405)
Processor Intel Xeon W-2255 CPU @ 3.70GHz, 3696 Mhz, 10 Core, 20 Logical Processors
128GB ram
Nvidia Quadro RTX 4000
Revoscan crashes almost every time, after finishing Global Marker Tracking scan, when I switch to cloud scan and change to Full Field scan mode.
I just tried it now, and switched to Full Field mode before the Global Marker scan, and then it was able to switch to cloud scan and stayed in full field mode, without crashing.
I used Revo Scan Metro X 5.6.4 several times yesterday, without this crash. scanning 330+ global markers and then switching to Full Field without crashing. Was able to scan 128Million points and edit the data down to 1.5M polys without issue.
Revo Scan Metro X 5.6.4
FW: 2.1.81.20250303
Windows 10 Pro
2x Intel Xeon E5-2665 @ 2.4GHz, 8 Cores, 16 Logical Procs (Total 16 Cores, 32 Logical Procs)
128GB ram
Nvidia Quadro RTX 4000
Hi folks,
My Quadro RTX 4000 GPU is not recognized, my 128GB RAM goes sometimes full and the resulting mesh is poor in comparison to photogrammetry.
I managed to get better results exporting the point cloud and meshing it with Metashape, so I think that revoscan does too much simplification.
I’m making photogrammetry with a basic reflex camera (Nikon D5200 and a Nikkor 60mm macro) and Metashape, and I have a very accurate and precise meshes. Scanner at left and photogrammetry at right:
The two reasons I bought a scanner is 1. to have already scaled and accurate measurements on the mesh (you have to make the scaling manually in photogrammetry), and 2. to reduce the processing time (but revoscan 5 makes it worse at the time writing this post).
As long as the photogrammetry results in a better mesh quality than the scanner does, I have the feeling to have lost 1000 bucks …
As the hardware itself is surely very good, I suppose that most improvements could be done to the software :
- be able to export the point cloud as taken by the scanner, without simplification (may be the firmware itself simplifies the data …)
- let the possibility for the user to mesh without any reduction or compression
- let CUDA compute all it can
- please recognize the Quadro RTX 4000
- improve the memory management
Some annoying details, but not really issues:
- changing scan mode is very slow
- having to set every time the exposition is tedious
And lastly, I’m a linux-only user since 20 years and I had to install a windows 10 in a dual boot just for scanning …
Many makers (somehow the guys who would use 3d scanners) are also linux freaks
- so please try to make an appimage for linux and I promise I’ll beta-test it for you
Thanks!