You are not logged in. Please login or register.
SoliForum - 3D Printing Community → ATLAS 3D & FreeLSS → Getting Good Results
5 scans merged together Meshlab then sculpted in Meshmixer. A little blender work as well.
Thanks Pirvan for the very useful tutorials. Learning the process is equally important to getting the scanner calibrated well.
Bob
The scan:
The object:
Very, very cool ! 
Nice work on the merge and post-scan cleanup.
What did you use Blender for? That particular program still eludes me.
I only used Blender to use its boolean difference function to clean up the holes on the sides of the neck. They were pretty much filled in after remeshing. I couldn't figure out how to remove some of the mesh to make a hole in either Meshlab or Meshmixer. Blender definitely isn't the most intuitive program but it is powerful. I'm still in the early learning stages using Blender.
Very cool!
Very good, my first scan looks deformed lol
My first 50 scans looked deformed. Keep at it.
Hmm I think i need to calibrate the scanner better :
it supposed to be a wooden model of a teddy LOL
also lighting. see the getting started instruction guide
my 1st scans are very poor compared to these. having hard time with laser/ambient light balance and stray spikes, rings and what looks like frost covered scanned object LMAO also getting turntable floor but camera focus is fuzzy on objects on turntable and crisp on background objects. not getting good test shots either only some laser lines and grey fog around them on object being scanned. mmore light washes it all black less light lowers laser spots hit til near gone. got a green and yellow plus sign in center seems just telling which laser is on. only red laser lines/dashes no yellow areas
I had the same focus issue with the camera. The lens is actually able to be turned and focused but it is glued in place. You have to break the glue loose in order to focus the camera. There is a thread in the forum about it. Once I focused the camera the picture became really clear.
All of my scans in this thread were done under normal fluorescent lights, although not ideal it worked ok. They were also multiple scans merged together. I also get some of the floor in my scans but it was easily removed in Meshmixer. If you didn't look at it already read Pirvan's tutorial. It saved me countless hours of researchng how to get a scan cleaned up. It is a sticky in the forums.
Bob
I need a tutorial on settings/setup b4 tweaking scans since I cant even get adjusted to see hardly anything in test views. this is brightest it gets.
cant get lighting any better anywhere from dark room to bright light, can make darker or worse laser lines but not see object.
here are a couple b4/after scans and stl of them. the Nscale truck2 has been talcum powdered to reduce shine/reflections but no help. they where done in high rez settings like Pirvan posted his in a topic with good calibration on camera and lasers also perfect alignment of distances in setup since each scanner can be a bit different I measured all points and set acordingly.
I have never used the STL file generated from the scanning software. It looked similar to yours when I viewed it once. I download the PLY file and follow the instructions at the below link. It pretty much lists step by step what you need to do to clean up your file. I had no experience with either program prior to following the tutorial.
http://www.soliforum.com/topic/12264/fr … -tutorial/
I did not adjust any advanced settings with the exception of the laser x and laser y values. The only time I got scans that looked as dark as yours were when I was playing around with different exposure settings. Make sure your exposure is set to auto.
All my scans are done using the high quality preset changed to use both lasers, 5mp and 1600 steps.
I need a tutorial on settings/setup b4 tweaking scans since I cant even get adjusted to see hardly anything in test views. this is brightest it gets.
You are looking at the calibration test image. This is normal. It only shows the lasers, nothing else, and it's intended for you to check that the laser exposure and alignment is correct. The 2 cross-hairs at the bottom are close, how are the cross-hairs at the top? The screenshot you posted is scrolled all the way down, so I can't see the top.
If they just as close, you're in pretty good shape.
Additionally, if you always have the turntable surface in the scan, that means either your camera is too low, or the settings in the Setup are not correct. Measure the actual height of your camera from the turntable surface to the center of the lens. Then enter that value in the settings page Camera Y.
On the Setup page there is a setting called Ground Plane Height. That should correspond to the surface of your turntable. If you still get a little bit of the turntable in your scans AFTER you entered the correct value in Camera Y, you can enter a value (positive number) here. That will tell the scanner to ignore anything below that height. For example. If you enter a value of .25", the scanner will only scan the object starting .25" ABOVE the turntable and ignore everything below.
So assuming the Camera Y value is accurate, Ground Plane Height should be 0", but if you still get just a tiny bit of it, then make Ground Plane Height something like 0.05" and it should disappear from your scans.
I have never used the STL file generated from the scanning software. It looked similar to yours when I viewed it once. I download the PLY file and follow the instructions at the below link. It pretty much lists step by step what you need to do to clean up your file. I had no experience with either program prior to following the tutorial.
http://www.soliforum.com/topic/12264/fr … -tutorial/
I did not adjust any advanced settings with the exception of the laser x and laser y values. The only time I got scans that looked as dark as yours were when I was playing around with different exposure settings. Make sure your exposure is set to auto.
All my scans are done using the high quality preset changed to use both lasers, 5mp and 1600 steps.
I made both Ply and STL and both looked the same. also the 1st scan was using auto, I tried all other exposure settings also laser settings and that was like I said the best result. also when we change any laser settings do we have to calibrate and save again or just save settings after change like I did?
I made both Ply and STL and both looked the same. also the 1st scan was using auto, I tried all other exposure settings also laser settings and that was like I said the best result. also when we change any laser settings do we have to calibrate and save again or just save settings after change like I did?
The PLY file will look the same, but unlike the STL, it will also contain the Point cloud, which is what you really want to work with mesh you see is not really usable. Open the PLY file in in MeshLab, then follow the clean up tutorial in the sticky.
n2ri wrote:I need a tutorial on settings/setup b4 tweaking scans since I cant even get adjusted to see hardly anything in test views. this is brightest it gets.
You are looking at the calibration test image. This is normal. It only shows the lasers, nothing else, and it's intended for you to check that the laser exposure and alignment is correct. The 2 cross-hairs at the bottom are close, how are the cross-hairs at the top? The screenshot you posted is scrolled all the way down, so I can't see the top.
If they just as close, you're in pretty good shape.
Additionally, if you always have the turntable surface in the scan, that means either your camera is too low, or the settings in the Setup are not correct.
yes the color plus signs are very close together top and bottom. I changed my ground setting to zero on 2nd scan like in your settings but also since these I rechecked all measurements and found my camera setting (since its to be level with lasers) was too high and lowered the number to match what the Y setting for camera and lasers are 2.75" the left and right are also dif from my camera by a bit as is the camera Z so I corrected all these hoping to get better calculations on scans including turntable base as my turntable does seem perfectly level which was 1 nice surprize. cant tell its moving for several minutes while scanning like watching an hour hand on clock plus I painted entire scanner flat Black plus printed of Black ABS. I am mostly concerned about the spikes/frosty look of model lacking any detailed or smooth surface and even holes/gaps left in parts of the models (I understand this when laser is not seen by camera in spots but some are where laser was seen and center seems to have a spot that is missing and still has a dimple under turntable where camera cant even see some how.
when calibrating lasers right and left merge like 1 beam at center of turntable and I have a perfectly square item with a sand paper strip on side (like a large matchbox) and the lasers are true all the way up. yet there is a spot at center about 3/16" dia and 1/4" into turntable that is hollow and sticking out bottom of turntable in 3D view like STL or PLY. very odd. as it could not be scanned that way. must be put there in software or something. no settings change that either and model placement only changes what part of model is in this 'black hole'
as is the Camera out of focus and coarse width/blurred laser lines are not any where near good enough for any sort of High Resolution detail but more like low poly scans of basic shaped items that are at least 3" wide/tall with details equivalent of a dull Crayon or at least a 3D printer nozzle of 3mm plus. not a .35mm one. no way can it be used for tiny models like N scale vehicles nor HO scale Hotwheels if dulled down would look like the bodies where made of corrugated roofing or something.
part of the center hole issue I found was the tiny center hole in turntable is not truly centered when rotating so had to realign lasers a tiny bit. likely due to the way it mounts and screws pulling it over some on shaft. may want to make note to check this b4 calibrating
just started another scan with the new settings etc. still says a bit over 10mins left to finish and thats with same settings as Pirvan posted for hig res. so 3200 would be 2x as long likely but not worth it unless camera and maybe lasers where better focused at least.
update; well it scanned with little cleanup but only thing it made was a big hairy blob a bit bigger than model looked like straw pile with a large gap hole and nothing under that even resembled model nore color (it was a metal casting with powder dusted on to dull glare. this is ridicules for 2hrs scanning! gotta be more to it than whats been said unless its just hipe like Santa Clause its not ready to market to public ATM. need more chalk board time.
I had similar results and mostly got them fixed via calibration. The lasers need to be aligned and there is a calibrate button to align them via software (only in 5M scan mode). Also the object must be in the middle of the turn table to scan with no reflections coming from the model. Don't forget to remove the three little blobs of glue on the Pi camera lens housing and lens so you can focus it. So it pays off to get the scanner well calibrated. Some objects just don't scan well like a banana peel, it's just too complicated. Simple objects without holes do work out well.
done the calibration several times after realigning lasers etc. seems lighting is main issue and nothing seems to work for that. I also need to build my dark room box for scanner for not only light control but eye protection while its scanning so long.
A scan of my toy toll
and here is the STL file created via Meshlab (screen shot from Netfabb)
SoliForum - 3D Printing Community → ATLAS 3D & FreeLSS → Getting Good Results
Powered by PunBB, supported by Informer Technologies, Inc.
