The new 0.2.2.0 version of the software was released, and I decided to try a complete scan and print of a rock I've used for testing from time to time. I set up my scan workstation:
I made sure the LED lights were positioned so the turntable was illuminated, but the webcam on the scanner was in shadow. I ran calibration again with the new software and ligh positions. I put a pickle jar lid on the turntable with a blob of blue Play Doh on it, then smushed the bottom of the rock into the Play Doh to hold it stationary during the scan.
I selected the 3rd default scan option for wide objects with lots of details and let it run (it definitely took at least the estimated hour and a half, maybe closer to two hours). When it was finished, I saved the raw scan file, then manually cleaned (making sure to cut off all the visible blue Play Doh from the bottom of the scan) followed by an auto clean, and saved the cleaned file.
I then repeated the whole process again with the rock on its side in the Play Doh.
Having gotten two scans that covered all points on the rock, I ran combine, and converted the combined file to ziprock.zip (which is a zip file holding the compressed stl file):
As you can see at the top of this meshlab screenshot, there were a few little excess blobs in the result, but I cleaned them up with meshlab before trying to print. The scan is also craggy like a rock, but lots of fine detail and sharp edges have gone away (only to be expected).
I definitely think the 0.2.2.0 version of the software seemed to produce a better scan than I had previously seen for this rock.
Anyway, having gotten a pretty good scan, I decided I'd actually try to print a copy of the rock on my Solidoodle 2 3D printer:
Not too bad really. If I had an infinite amount of time, I'd scan the printed version and print it again and keep going to find out how many iterations it takes to melt into a plastic puddle, but I don't have that much patience :-).
Go back to my main scanner page.