Hello! I am the owner of MINI and I am waiting for POP3, scanning with revopoint opens up new possibilities for the 3d world, at the beginning of the journey you don’t need to expect instant success, it takes endurance and patience, with practice you will understand that the scanner has quite a lot of functionality, and with the advent of Revo Scan 5 it will work very simple! Yes, and you can always ask questions on the forum, do not be shy, they will not leave you without attention!
2 Likes
Hi @Singaii,
Yes, it is important to consider the scanning path before starting the scanning process. Additionally, performing the scanning movements correctly during the process is crucial for achieving accurate and high-quality results.
Thank you for sharing!
Hi @Cubic ,
Yes, using a suitable spray is also important for achieving excellent results. Thank you!
Hi @Alex666 ,
Yes, being patient, practicing more, and having the courage to ask questions or seek help on the forum are all important.
Here are few useful things i’ve found using the mini.
Don’t try to get the whole object in scan. Don’t overscan the object. Use partial scans and merge/ align them in RS.
If you are scanning smaller object, use perforated paper towel as the surface under the object. This will act as a feature fir tracking (you can always use small spacer to help with cleaning later. helps massively with tracking and picking up the frames.
If scanning objects that are larger or lack features: use gravel or little rocks as feature or modeling clay that you can shape into irregular objects that will help with tracking and aligning multiple scans.
If the object isn’t flat on the surface - use uhu patafix reusable sticking “clay”. It’s cheapish, can be shaped, it sticks on anything and it’s reusable.
All this tips help with tracking, and usually getting this right takes time. Once you master tracking and adjusting exposure, you’ll be quick and get great results.
Here’s patafix in action. Black surface with good exposure settings and no scanning spray.
2 Likes
Hi @Samster_3d ,
Nice suggestions. Thank you for sharing!
Could you show a picture of perforated paper towel? I think this is a great idea for new users. Thank you in advance.
1 Like
Sure.
Here are the pictures where i scanned strut brace mounts. The shape of the mounts doesn’t have any useful features for tracking. Paper towel has some kind of pattern and it made 3d scanning and tracking quite easy.
Here i also used spot light with rgb adjustment to see if any colour of the light would helps with scanning. Red seems to help just a tiny bit.
Hi, I like your fill light, can you send me a link to buy it? Thank you.
Hi everyone! I’d like to share my workflow considerations and tips. I have been scanning mostly paleontological/archaeological specimens using a Pop2 and Range. I have access to fossils, bones, casts, and imprints that are not very spray/scanner friendly so I have been learning the limitations and how to overcome them. Lately I’ve been working with taxidermy behind plexiglass and that has given me insights in the art of patience and trying new things.
Prep/some things to consider:
- Each subject you scan will have different ideal scan/processing settings to match its reflexive and size characteristics. The scanner has limitations in how effectively it can beam out a grid for tracing the object (like how your car’s headlights can’t reveal anything near horizontal, right in front of you, or far away), including how well it can see, how close/far it can be, and how fast it can retrieve the points.
- The scanner is more accurate than it is precise, that is to say it can capture a narrow group of points around the target but it lacks repeatability if you try to refine the cluster by scanning the same part in the same configuration over and over. Consider focusing your attention on having multiple scans in various angles so the group of points can average out.
- Signal is useful data, and noise is extraneous junk that can be from the low level of useful reflection that the sensor is detecting (if the object is black, reflective, moving), in addition to electronic temperature or even gamma rays from stars exploding. When there is not enough signal the scanner will start hallucinating floating blobs and sponges.
- The scanner has a narrow range where it can capture cloud points in the ideal position with high signal. This is the source of your frustration, but the key to overcoming them requires that you cultivate a zen-like approach to experimenting: only methodically changing one thing at a time and keeping it simple. Adding noisy scans will only create problems.
- You need to consider how to balance exposure (frame rate) with the rate of acquiring new data. The scanner is more like a timelapse camera than a live feed, so focusing on one region a while will introduce artifacts and moving too fast will make them smudge out, move but not too fast! The accuracy settings changes the frame rate. Bigger objects require quicker scans and darker objects require a longer exposure for the camera to capture them.
- Objects with complex geometries caving in where the scanner’s beam cannot reach will require some ingenuity (usually positioning the camera in different angles, making the surface more visible, or leaving it for post-processing)
- Consider the environmental conditions you are working with and how much control you have over the subject such as lighting and positioning. Solar UV and other lights will interfere with the mesh and color data. Do you have a fixture/mount that can secure the subject? Alternatives for turntables can be the dinner lazy susan or an office chair. When securing the subject it should be slightly raised above (so you can cut around it later). I use wood panels covered with office marker dots.
- You can artificially increase the signal-to-noise ratio by providing sources of high-signal that the scanner can readily pick up, crumped up pieces of tissue or unpainted wood pieces are really great. Sticky office dots placed on the subject can work to dull and bring out shiny/dark areas. Spread them randomly and make them apparent in the scan.
Scanning:
- First position the scanner to the darkest and closest area, having the histogram be mostly green at the excellent end and start the scan rotated from there. This will prevent you from clipping the most readily visible area. Make sure your scanning distance setting is appropriate, you might want to bump down the max just where your tray holding the object is visible so background things don’t mess it up. Next use the left/right arrow keys to increase exposure until you see some red highlights, and then step it back a few steps back now trying to avoid the blue as much as possible (over/under exposure indicators respectively). You might not get it perfect with dark objects but first you should make sure the camera can see the subject or add artificial matte surfaces like stickers. Then you might want to do a few practice pans to see how the distances and motions will affect the capture. The rotation will always throw your histogram off, just keep an eye on it and return to the areas that are deemed too far later. Consider moving in one axis only, whether rotating or going sideways.
- There is a magic number of frames for your object, depending on your scanner, how large it is and other conditions. Experiment to find out what this may be for you and have your movement be planned to reach around this amount for the next scans. The point of diminishing returns has been reached when you cannot really capture detail with more frames or by stacking from multiple angles. There will be errors due to spread in the cluster of point clouds, so aim for a reasonable amount and don’t stress yourself!
- It is best to complete full 360 degree rotations for each scan even if the other areas are not so perfect. Take scans from multiple angles, each one facing a different area: Tilted slightly downwards, at the same level, and raised upwards.
- I alternate between capturing/stacking and continually iterate upon the previous work. This allows the camera to cool down while I run through a preliminary cloudpoint merge and can see the areas I missed.
Processing (not a one step process, redo this multiple times!):
- Remove any unwanted background away from the subject using lasso tool drawing a C wrapping the subject and deleting the other stuff so you don’t waste computing power later (this has been moved to the right and done after Fusion on RevoScan 5)
- Create a mesh and clean the model with fusion, isolation and overlap detection. A lower point distance for fusion will preserve more detail and noise. The isolation tool is meant to remove extraneous clusters away from the subject and the overlap detection will smooth out the bumps. The tools will remove some points and lower the quality, so start from a low number and work your way up; usually automatic settings work but you have to mess around with the sliders and apply it multiple times at different rates until you have removed noise but not so much that you lose detail.
- Remove small floating parts or fill missing holes, but be careful this might introduce flat and curves where there shouldn’t be.
- When you have processed multiple scans you are ready to merge them, I find it best to merge them all at a time so the scans are all averaged out evenly
Once you have merged the files and happy, you can proceed with smooth/simplify. This process will blur out the surface and decimates the vertices but can be done with other software,
- Post-Processing for FDM/resin/CNC in order of difficulty:
- Repair/conversion: PrusaSlicer, 3D builder, old Revopoint studio
- Smoothing and sculpting: Meshmixer, Blender
- CloudCompare - advanced smoothing and cleaning for getting rid of junk
- Fusion 360 - parametric environment for reverse engineering, if you are lazy and the part is not complex you can maybe use Tinkercad and fix up the model that has been filled in
6 Likes