Suggestions and Feedback for Revo Scan 5 MetroX

[Revopoint-Jane]

Hello everyone,

As you may know, we have recently launched specialized software, Revo Scan 5 MetroX.

Given that this software is in its initial release phase, some areas may require improvement. So, we’d love to hear your feedback and thoughts on how we could improve this software.

Please share your feedback in this thread in the following format:

• Revo Scan 5 MetroX Version: XXXX (e.g. Revo Scan 5 MetroX V5.6.1.82)
• Operating System: XXXX (e.g. Windows, macOS)
• PC Specs: XXX (e.g. CPU: Intel i9 12th Gen, RAM: 64GB GPU: NVIDIA RTX 4060 (8 GB))
• Your Feedback & Suggestion: XXXX

Thanks in advance for all the feedback and suggestions!

Best wishes,
Jane

[ttabbal]

Revo Scan 5 MetroX V5.6.1.82
Win11 (all updates applied)
PC: Lenovo Legion 5. i9-14900hx, 96GB RAM, 4060 8GB.

Stability. First release, so it is expected there will be bugs. But at the least, it would be nice if crashing were less frequent. I seem to get it less than others, but it does happen.

Better marker tracking. It’s not unusual to see them shown in the depth preview as overexposed, but it doesn’t track them. Sometimes it’s angle or distance, but it would be nice if it were a little more flexible. I get that it needs to be precise, so obviously that is the priority.

It would be nice if the lasers were more visible on the screen. When adjusting distance to get details that are slightly out of range, the lines help a lot to know that I’m getting the detail I need. It might be enough to let me pick a zoom level. I find it likes to show me more data than I care about most of the time while scanning.

Feature tracking could use improvement. I’m sure you are aware. Maybe add texture tracking.

GPU processing. Please get this working.

Better multi-threading. I know this is difficult, but it would help a LOT of people. And you likely need it for GPU anyway. Smaller units of work, spread across more cores. For an easy win, when using batch processing, run them in parallel. Right now, it’s only hitting 2 cores. I have 24. But at least get the performance cores going!

Merging can be difficult when thin objects are scanned. Think PCBs, SD cards, etc… Getting the other side can be difficult and merging often puts them back to back. Being able to rotate them to approximately the right orientation and then clicking points for manual alignment could help. Might be better options. Even when they do line up, I find I lose details sometimes. It’s like they are merged slightly off and it blurs the details. There might not be a ton to be done, but figured I’d mention it. Maybe if existing data is high quality, let me mark it to exclude merging into that area? I don’t want to remove parts, I might need them to align scans. Or a mode to fill holes only when merging?

It would be nice if mode switching were a little smoother. It basically hangs the program while it switches. It would be nice if that were a background task at least. Perhaps let the user pick a few options and THEN change the scanner mode? There are a lot of options. Kind of annoying when I want to set up a different scan and every click is a 20 second delay. I don’t mind the scanner needing a moment to adjust, it would just be nice if every UI click didn’t send a command needing that delay. It’s also started telling me there is a low frame rate when changing modes. Well, yeah, no data should be coming in…

It would be nice if when I select a mode other than auto turntable, it doesn’t run the turntable when I start the scan. Maybe a checkbox over near where the mode is set to enable/disable starting it up? Maybe some people want that. I just figure I’ll push the button on it if needed. Yes, I can disconnect it. If that is the preferred solution, please make it easier to disconnect it.

Ended up being a slightly bigger wall of text than expected… sorry about that.

[misterpepper]

Better compatibility with aftermarket marker dots. I have 3mm dot that scan fine in laser modes (they appear on screen as 6mm dots) but are not recognized at all in full field mode. I also have other dots that are 6mm that scan fine in full field mode but laser mode does not pick them up. They are reflective but do not reflect as brightly as the Revopoint markers.

It would be nice if there was a warning in full field, feature tracking mode that not enough features are in view before starting a scan, similar to the warning for not enough markers in view.

Please add some basic alignment and measurement tools so that we can align our meshes to a coordinate system before exporting or take measurements within RevoScan rather than needing to use another program.

Some people are reporting that when scanning in feature mode if there are any marker dots in frame it can affect the scan negatively. This should not be the case. RevoScan should be able to ignore marker dots in view if they are not being used for tracking.

[Elektro_freakz]

Revo Scan 5 MetroX V5.6.1
Win11 (all updates applied)
PC: Acer nitro N50-130 AMD, ryzen 7, 8 core processor, 64gb RAM, AMD Radeon RX 7600 8gb

I used the metrox a couple of thimes now. And i must say that when it works, it works good. The quality is fine and after scan processing works also nice.
But please fix the hanging in the program. It seems like it hangs at random moments, shuts down the scanner and shuts down the program.
At least the scan is saved at that moment. But its really annoying,.

Also, when connecting the scanner i somethimes have a lot of issues. The scanner is updated with the latest firmware and i have the latest software version. I connect the scanner, start revoscan (or the other way around) And it has a lot of difficulties to connect with the scanner. The scanner itself looks and sounds like its continiously rebooting, and trying to connect to the pc.
After a couple failed connections, it randmomly connects and everything is fine for the rest of the scan.

What would be nice to implement in the software you ask? I have my turntable set up with some marker blocks stuck to it and made a general marker map. So i can scan models easy without having to make a map of the markers. What would be a realy nice to have is an option to configure the 2 hardware buttons on the metrox ( + and - button) to control the turntable while scanning.
So lets say you have a tall object to scan. The turntable is on your desk and you have problems walking around the turntable. With the auto rotation function it somethimes goes to fast. You miss some features and have to wait for a full rotation to scan the missing feature. You miss another one and so it continues.

How cool would it be to just scan the project as far as you can, move the turntable with the buttons on the metroX, - is turn anti clockwise, + is turn clockwise at a set speed and you can scan the whole part around without the need to let go of the scanner or awkwardly bend around the object (with the risk of knocking it over)
I think this would add up to a lot of convenience.

[SphaeroX]

Revo Scan 5 MetroX Version: Revo Scan 5 MetroX V5.6.1
Operating System: Windows
PC Specs:
Intel Core i9 14900HX
64 GB RAM,
2000 GB SSD
GeForce RTX 4090 (16 GB)

I’ve been testing the new 3D scanner extensively, and I’d like to start by giving some well-deserved praise! The ability to capture fine details is truly impressive. For example, I scanned a necklace with a diameter of just 13 mm – the level of detail is simply astonishing.

Despite the positive experience, I have a few suggestions that could make the scanner even better:

1. E57 Export in Auto-Turntable Mode
   A user in the Discord community managed to export frames as an E57 file and import them into Reality Capture. The resulting rendered mesh looked incredible. It would be fantastic to have this E57 export feature directly available in Auto-Turntable mode (with Colors) – ideally including the color maps.

MetroX Scan processed in RealityCaputre:

2. LED Indicator in Single-Shot Mode
   I think it would be very helpful if the LEDs in Single-Shot mode turned red when a frame cannot be aligned and green when it aligns successfully. This would make it easier to walk around the object and take shots without having to constantly check the monitor.

3. Merging Models
   Currently, a meshing process must be completed before models can be merged, which seems unnecessary. The point clouds should be able to merge directly. I believe this might be a small software bug.

4. Overscanning Issue in Laser Scan Mode
   In Laser Scan mode, previously scanned areas often get overscanned if scanned repeatedly – especially with engravings or indentations. It would be helpful to implement a mechanism that prevents new points from being added if the area is already sufficiently scanned.

5. Easier Start for Global Markers
   At the moment, I have to click on the three dots and manually select the scanner to start a scan with global markers. It would be much more intuitive if the menu opened with a double-click on the marker.

6. Depth Camera Settings Not Saved
   One small annoyance is that the depth camera settings reset with every new scan. Even if I set it to automatic, the settings revert to default when I start a new scan. It would be great if the scanner could retain the last-used settings.

But I’m quite satisfied, and the software runs pretty fast, which surprises me a bit because I had heard otherwise. But since I have a current CPU and GPU, I guess that makes a difference. Still, I didn’t expect to be able to measure large things this quickly. That really surprised me – and of course, I’m happy about it.

[river65]

Revo Scan 5 MetroX V5.6.1.82
Windows 11 24H2 26100.2605
Laptop Intel i9-14900HX 2.20GHz / 64Gb ram / Nividia GeForce RTX4080 12Gb laptop gpu
Scanner firmware updated when it 1st connected

; back / undo button whilst scanning to clear errors

; align to planes for transfer to cad programs

; ability to move program smoothly to other screens / monitors and rescale to fit correctly, mine gets stuck between laptop and the 2 external monitors and have to close program from taskbar as top right “min/max/close” are not visable.

; better tracking in full field mode

; when clicking on “marker scan” or “feature scan” the options should drop down under them, not on the other side of the screen

[Erked]

Revo Scan 5 MetroX V5.6.1.82
Win11
PC: AMD Ryzen 9 9950X, RAM: 192GB, GPU: NVIDIA RTX 4090 24 GB

Please add 3mm marker recognition. Being able to scan 3mm markers and having them show up to size would be extremely helpful. I’m limited by not having that. I assumed that 3 & 6mm would be available from the start as my other scanner was capable of that. Bonus if you can include 10mm as well.

Adding a few more steps of adjustment for laser and camera would be nice for fine tuning the settings.

The waiting when switching modes is a little long. I feel that multicore utilization could be better.

EDIT:
I would like to have concise instructions for ideal scanning for best accuracy. I watched a recent video that says to keep the markers 6 to 7cm apart. then the next example says if the part is too small to put markers on the part use marker blocks placed around the small part. The marker blocks have markers that are 2cm apart from each other. The video then states “don’t use too many markers, but ensure that at least 5 markers are placed irregularly per scanned frame. Try to place more markers at the key locations” It shows an example in the bottom left of the screen of what appears to be 9 to 10 markers that would show up in the scan. So how many is too many? I think it was great that you made a video to help people but it left me with more questions than it answered. You guys have done a great job with the updates to make the previous scanners better. I’ll wait for the update before doing scans of my ball bars and ceramic gauge blocks to test the accuracy for my needs.

[misterpepper]

Please add some sort of countdown when using auto turntable mode so that we know how many scan positions have been scanned and how many are remaining.

When doing a new scan please give the option to keep the setting from the previous scan. Often we are just repositioning the same item to be scanned and want the scan settings to be consistent. In auto turntable, when doing a color scan the color/no color toggle turns off when starting a new scan, and if you have set the scan exposure manually it resets to “1” on the slider.

Is it possible to detect and give a waring if the markers in view are at too low of an angle to be useable? My understanding is that markers in view need to be within 45 degrees of perpendicular to the scanner to be recognized, even when it is obvious on screen that there are markers in view. Maybe color the markers differently if they are seen but not useful for tracking?

When processing multiple scans would it be possible to allow the one-click option to process in the background while you are working to manually clean up a different scan? Preferably if the software could batch-process multiple scans in the background and still allow you to use RevoScan for other things. It might be too processor intensive to allow batch processing while scanning but it should be okay to allow it while trimming an already processed scan.

[BenScan]

Revo Scan 5 MetroX V5.6.1
Win11
PC: Intel 13900HX, 4060 (8gb), 32gb DDr5

Honnestly spoken i had no problems with any kinds of crashes so far…on my laptop the software runs stable.

Here are my suggestions:

1. during post processing that would be nice if we had the chance to auto remove the markers and fill this holes automaticly.
2. Most of us using 3D printed marker pyramides or similar…what about to upload these 3D files in the software and the software would automaticly search for point clouds / meshes that look similar and than you have the chance to delete them. That would be very helpfull for post processing
3. Increase the FPS during FullField Scanning to optimized the tracking. For the point cloud itself you can only take 1/4 or 1/3 of the frames to avoid unnecessary data

[misterpepper]

Why does “one-click edit” require two clicks? Many times I have clicked the one-click button and forgotten to look to the opposite side of the screen and click “apply” after.

[Jayson]

Here are some ideas and suggestions. I also posted these on the forum.

1. Offloading to save RAM usage (proxy mode): When scanning offload point cloud data to a file after pausing data capture or after a certain amount of frames have been captured (user selectable), but still keep the marker data and general point cloud (a low quality proxy) on screen for reference. Then when the scanning is finished, have the software grab that saved data from drive when fusing the point cloud to save on RAM usage (could still be proxy till saved, “file option "Save not as proxy/Save as hi-quality”). Basically offloading the data to drive till it’s fully compiled. In video editing/rendering software this is a thing.

2. Skinnable/Color Scheme Interface: Have the interface able to be Skinnable or change color scheme.

3. Optimize for GPU Memory usage: Such as OpenCL or CUDA Support!?

4. Fix DPI issues in GUI

5. Create a smoother view port experience: When scanning by hiding a percentage of the cloud data so the GPU doesn’t have to work as hard.

6. Hide help info in the GUI: Like a mouse rollover and be able to turn on or off. This could save space to view more scan data

7. Temperature: Be able to see temperature data of the scanner or a warning when it gets to hot and have a heat protection mode

8. Undo button

9. Faster transition from different scan modes

10. Better algorithms for tracking

11. RGB scanning after laser scan: Be able to use market data to take pictures (photogrammetry) of objects for RGB texture capture for laser scans (could use global tracking). Then have it project on the meshed model to be saved as a texture.

12. Alignment tools

13. Measuring tools SAE and Metric: This is a metrology grade scanner should have those tools handy.

14. Re-Topology tools

Hope this helps!

[Nigelg]

Yes I absolutely agree that the turntable should not start automatically if you have not selected auto-turntable mode. If you want to start scanning in freehand full frame mode it stuffs up the scan right from the start. It’s easy enough to press the button when you are ready to rotate or as suggested by ttabbal, a tick box or start stop option within the software could be great also.

[stevehaycock]

AMD Ryzen 9 5900X
64GB RAM
Nvidia A5000 24GB
Revo Scan 5.6.1

This has been sent to support who have passed on to R&D to see if it is possible.

It’s only possible to capture RGB images using the auto-turntable mode to prevent motion blur. Here’s an idea for a potential workflow:

1. 3D Data Capture : Using global marker tracking, capture the 3D data in any mode.
2. Image Capture : After capturing the 3D data, and ensuring that neither the object nor the global markers have been moved, use the scanner’s RGB camera to capture the images. The global markers could then be used to align the images with the 3D data. To avoid motion blur, the scanner would need to remain static during image capture.

Potential Downsides :

• The workflow would require multiple passes over the object.
• Lighting differences during RGB capture could affect consistency, so the object would need to be evenly lit throughout the process.

To ensure alignment, the “Take Image” button in RevoScan could only be enabled if the minimum number of markers required for alignment are visible and locked.

[bakerboy]

Revo Scan 5 MetroX Version : Revo Scan 5 MetroX V5.6.1
Operating System : Windows
PC Specs :
Ryzen 7 7700x
64 GB RAM,
4TB M.2
GTX1080Ti

Allow us to turn the laser off if it’s just sitting there for a few minutes while we setup a scan.

Enable gpu acceleration

More reliable turn table connection, currently just connects and disconnects after working fine the previous day

Improve marker recognition so 3mm dots work

When looking back at frames to remove it should operate a bit like a 3d slicer where you start with nothing and then as you move along the frames the pointcloud builds up and you can slide the bar along until you’re happy and delete frames outside of this selection.

Remember the last damn settings we used, I don’t want to have to adjust the exposure every single time I start a scan, remember the last settings. Same goes with editing, always defaulting back. Stop it, we’ve made adjustments because we want those values, leave it until we change it.

[abiss221]

Along with what others have said. There are three main areas I have been struggling with RSMX.

Revo Scan 5 MetroX v5.6.1.139
OSX Sequoia 15.2
16" MBP (11/23), M3 Max, 48gb RAM

LASER SCAN QUALITY / RELIABILITY

• The first, and I don’t know why this is happening, is with nearly every scan I do, the software has trouble collecting data on seemingly random areas of models. Random in that there is nothing unique about the color, texture, or particular geometry in the difficult to scan areas of the physical model. I will go over the particular area over and over, but no data will get collected in the software and I end up with large holes in the point cloud. Example attached below.
  
  

DEPTH CAMERA PREVIEW ROTATION (i.e. landscape / portrait)

• Currently, you can only rotate the view to see it in 2 orientations. It would be extremely helpful to have the option to see preview in all four orientations so you can decide if the USB cable is hanging to the right or left.
• Can you please unlock the orientation setting during scanning so you can switch between landscape and portrait orientation during a scan?
• It would also be great to have a button on the scanner routed to the preview orientation.

DEPTH CAMERA EXPOSURE (IMPORTANT)

• It is currently impossible to judge if the exposure is correctly set, those controls are useless without a way to judge exposure.
• SOLUTION: Can you please add a “false color” overlay to the depth camera?
• i.e. competitor Scan has a “false color” preview: This adds a red overlay on the image areas that are overexposed (aka highlight clipping), and a blue overlay to clipped shadows. It would be extremely helpful to also have a green or pink overlay on properly exposed areas. Finally, another color could be added for detected markers.

MISC

• It would be really nice to add a way to complete the scan via physical button on the scanner. The competitor Raptor achieves this by having you hold the start button down for 3 or more seconds to complete the scan - extremely helpful!

[abiss221]

Also, please add an option to ignore the floor or flat base during laser scanning as well. competitor Scan has this I know you guys can figure this out. Thank you.

[MonsterMaxx]

Dell Precision laptop, i9-12950HX, 64GB, Quadro A4500, W11Pro, MetroX V5.6.1.
MetroX works fairly well for me in the full field mode. Feature tracking isn’t great even with very featured subjects, but I’ve been able to get some good scans with it. I have a lot of trouble with cross line mode. The laser scarfs out where it is and even with what is looking like a good scan (all green) fusing and isolation make a horrible mess. Maybe this is user error as I see others getting good crossline scans, but I cannot get shiny objects to scan nicely in cross line mode. (yet to move to black)

Anyway, good launch of a product that’s clearly still beta, it works surprisingly well for being beta.

Here’s my quick list of items I’d like to see happen to the software.
• 3mm markers working and displaying correctly.
• Better utilization of CPU and GPU. Currently it’s rare that the CPU is using more than 12% (single process) and looks like GPU is not being used at all. Leaving a lot of performance unused.
• Time to complete task on all task displays. Currently on a few, make this on all. Add a cancel button too to those that are missing it.
• Black and shiny metal modes in full field scans
• Fix whatever is causing the scarfed out scan in cross laser mode. I have scans that are totally green and when I fuse them they come out horribly scarfed out where the laser lines are. (images/files can be submitted to show this as requested)
• The ability to change to black or shiny metal during a scan
• Ability to rotate orientation during a scan. need the other 2 positions too.
• Better tracking, it frequently loses tracking even with plenty of markers in view.
• Hybrid tracking that does markers and features
• Merge needs help. Auto mode needs a user assisted mode, like getting the two scans in basic alignment before pressing go so it doesn’t come out flipped over (Einscan does this and it works well.)
• Ability to select and delete parts of the scan before fusing to reduce processing time.
• Clone a scan so one can make changes to it and still have the original. Better would be do a feature manager tree and allow edits of the individual scans once they have been merged – though that’s probably beyond the scope of fixing problems.
• Indicate what scans are in a merge
• Display the number of frames in the scan’s data field
• Crashing. Yes, I know, it’s beta software and you’ve done a good job, keep going.
• Add Base removal to cross line and single line scan modes.
• Speed up the switching from modes, or at least let the software continue on with it’s task and not have to wait on the scanner.
• Remember settings from mode to mode.
• Brightness and depth exposure settings in all modes
• Get the large turntable hooked up to the scan software
• Give us a menu command / button to get to global tracking, don’t make us dance thru modes to get there.
• More feedback on brightness and depth so we can make good setting choices (maybe some green coloring when it’s good?)

[TheBoatScans]

@ttabbal

It would be nice if when I select a mode other than auto turntable, it doesn’t run the turntable when I start the scan. Maybe a checkbox over near where the mode is set to enable/disable starting it up? Maybe some people want that. I just figure I’ll push the button on it if needed. Yes, I can disconnect it. If that is the preferred solution, please make it easier to disconnect it.

There’s already an option for this on the turntable settings panel. It’s called Autosync or something similar (not currently able to look unfortunately).

[GuyH77]

Lots of great suggestions in this thread.

I’ll try and add more as I get more time with my MetroX. One thing I had yesterday was when cleaning up a scan using the fence select, I zoomed in and expected the fence to adjust with the zoom. It didn’t. This seems the wrong behaviour to me.

[aluhao]

Would it be possible to have a selection option when doing point cloud isolation for clicking an area of the isolated points to keep? On larger scans even the lowest 0,5% will sometimes remove some wanted areas.

Also in another software for point cloud/mesh handling you can change from selection to view manipulation with Space key, this is very convenient when selecting areas of scans. I know this is something that i have been so used to now that the Space key reset gives me a lot of headache and is just perhaps something i need to work on but switching modes like that would give a lot more freedom for key shortcuts in each mode also.