I cannot get decent scans out of the mini. Every time I try to scan the bust I end up with strange artifacts. I built a box that is all black inside. I did an accuracy test in there and it was 0.06. Not the 0.02 I was promised. I tried to calibrate and it went from 0.06 to 0.08.
The board that came with the mini doesn’t match the one in the video. I think I have it orientated correctly but can’t say for sure.
I need help with:
Calibration so the accuracy in the black box is 0.02 as advertised.
I sent 4 pictures to Revopoint asking if these items could be scanned and was told yes they can. So far the 1st was a total failure so I need help with getting that scanned.
It is possible you are overscanning (too many scans of the same area). This is most commonly done when making successive scans with different orientations (but still many areas that overlap in each scan).
Alternatively, the object could be reflecting light sources that are interfering with the pattern. Recently, I had to wait from late afternoon to the middle of the night to make some scans because the sunlight was making it through the window slats.
0.0600 is not bad score at all .
You will be not able to check the accuracy at home , you need a special instrument for that ( very expensive) , and the accuracy is point accuracy from one frame , a distance between 2 single points.
So no way you can test it precisely.
Plus the best accuracy is only at 10cm distance between the scanner and the object .
You saying you getting artifacts
Post a screenshot of that and a picture of the object you want to scan .
It is hard to imagine what you trying to do as each object is unique and needs own approach.
Keep the scanner at the exact distance from the object , and if you use Dual Axis turntable use the manual settings , because the semi automatic don’t deliver the exact precision you are looking for because of the many overlap framed that altering the sharpness of the final scan.
Perfect scan need one 360 degrees rotation scan data of one angle , and as many angles as it needs to be merged later if you going for precision , if you trying to scan everything at once the chance is big that you would lose the sharpness and accuracy in the process .
And if you can’t scan the bust correctly, try to not use the Revo Scan open in full screen , try to resize it down slightly, I saw post about a bug that affects some users while the software window is in full screen . This bug is addressed already by the team .
I will have about 3 hours this weekend to work on this and we’ll post screenshots at that time.
I have the basic turntable. When it comes to the bus I’ve been setting it in the center of the turntable and scanning for one full rotation. Then I pause and lay the bust on its side or back and scan again and that takes care of getting most of it. When it’s compiled in the software is when the artifacts are shown.
I have another all white object slightly smaller than the bust that I have tried to scan and failed. Prior to purchasing the mini I sent a picture of this object along with Dimensions to Revel point with the question can I scan this with the mini? Their reply was yes. So far it does a terrible job when it is vertically positioned on the turntable and loses tracking as soon as I turn it on its side or back. All detail is lost. Again I will post pictures this weekend probably with a brand new post.
You need to start scanning first on the side , and then put it in the normal position , it will improve the tracking .
You need to fuse the points at least and mesh it for the final result .
Or use Revo Studio to clean manually the scan data after fusing it .
I scanned the bust so many times I can do it blindfolded without any artifacts in result.
So looking forward to see your screenshot and see what is going on .
Also include the photo of the small object, I scanned objects that are 0.76mm in diameter , and 6 x 6 x 6 mm in volume
So there is always a solution… waiting for your screenshots when you are ready .
I finally have some time to work on this.
I built a black box of felt and that seemed to help with the bust. It turned out alright.
I decided to move on to a 30mm^3 mech.
This didn’t work. It would scan it the 1st time in any position on the turntable, but when I changed it’s position the scanner lost track. It didn’t seem to matter if I started with the mech upright, on it’s back, on it’s side, on it’s face. No matter what I did, it would loos track.
As you can see in the picture, there is quite a bit of detail. I scanned the mech vertically with 1 full rotation. I then fused the points at 0.02. And then ran the mesh at qual 6 with denoise at 2. This is absolutely horrendous. I sent a picture of this exact mech to Revopoint prior to purchasing this mini and asked “will the mini scan this successfully?” I was told yes.
I was told in the facebook group I needed to use scanning spray. Here are the results of spraying a resin printed mini wih AB2G White scanning spray. Once again, the scanner looses track when I reposition.
Rat I was trying to scan. It doesn’t look overly detailed (like the mech) but at the same time it is more of ratish blob than a reproduction of the scan.
When you scan figurines, you need to start scanning from the side first and finish with the top position , less issues with tracking .
Distance between scanner and center of the turntable 10cm for best sharpness .
I can scan a needle with MINI without issues and finger prints , so it is definitely your device has issue … maybe the calibration went down while shipping when it was shaked back and forth .
I place the figures upright 1st. When laying them on the side OR moving the scanner tracking is lost. For example with the ghost. I scanned it with it upright on the turntable. I then adjusted the scanner to get a more top-down view to fill the hole on the very top. Tracking was lost.
I have adjusted the distance to the majority of the time it is at excellent with a few drops to good.
I did an accuracy test with the Revo Calibration software. The first time it was a 0.06. After that it never got below 0.08.
I tried recalibration and it never got better than 0.08.
EDIT
Apparently I can’t post multiple pictures or continue to answering questions until 8hrs from now.
In reply to the below comment.
The majority of the ghost was still in the frame. The adjustment was minor to get the small hole at the top.
I have tried getting within 1cm of the object. This required holding the scanner and did not work. I have also tried getting as close as turntable and stand will physically allow. Again, no difference.
At this point accuracy testing does not get below 0.08. Sometimes is above.
All revo programs suck massive amount of resources at full screen. I do not run any of their programs that way.
Of course you can’t scan a hole so it is obvious it lost tracking , but regular figurines with solid bottom need to be scanned on the side first so the tracking have captured some of the features first .
The distance excellent to good is above 20cm , you will not get very fine details at this distance , it need to be from too near to excellent exactly 10 cm from the scanner to the center of turntable for maximum accuracy and details . How far you go , how less details will be , and scanning so small objects request the closest distance , ignore the excellent label on top , measure the distance for best results .
It should get below 0.08 at least to 0.04
0.08 do not affect as much the meshing and it is still acceptable, but I am surprised it never went below that value.
Can you open your Revo Scan , but do not open it in full screen , just slightly smaller than full screen , and see if you got any difference, there was some bug in the software that affected some users that should be fixed already in the commons days …
Let’s see if this is the reason or something else , it would be easier for the team to determine what is going on with your device .
Hi,
Please change the setting as below:
Distance 10cm, point distance 0.02mm, meshing 6, denoise 3.
And scan the models again.
If the results are still the same, please export the model. And then check out the data file, share the point cloud data (named xcc_fuse) with us.
So that we can analyze what the problem is.
