Another great video! Thanks to the first video in this series I was finally able to get clean results from scans for my complex part but I suffered through trying to process the scans on the software. (Mainly because I had no references to what the options meant so it was a lot of trial and error). After watching part 2 I feel as if I’ll have something of much better quality ready in time for the last video because in trying to import the final file to F360, I couldn’t for the life of me get it to become a nice solid file. I could only get it to become tens of thousands of triangles with no way to manipulate them similar to my native fusion designs. Part 3 can’t come soon enough!
Awesome! Glad to hear it. I love the Miraco! We will be doing another video soon looking deeper at the Miraco and Mini2 together. Scanning larger parts that are a bit tough for the mini2 just to see how to handle them.
as an ex-motocross journalist and photographer (meaning I think about light and how to use it, quite a lot) who just started 3d scanning I want to say that I think it is not better with darker environment or less light but I think it is about how the light is transfered (if one can say that this way). I realized when I scanned in the morning and I had really much light because of the position of the windows the scans were so much better because scanning needs light with these low budget scanners. but exposure to artificial light that is too direct to the surfaces causes problems (like direct sun light) , it produces hard shadows and reflections and the quality and the feasability will go down the drain. so I think/ or guess a good solution would be the same as the one that photographers use and get indirect light (like with a softbox or so). this means checking your "light-setup" in your garage and changing it a bit to produce indirect and soft light could potentially be a very rewarding thing. these are my thoughts, correct me if I'm wrong !
Ex motocross journalist eh! :) sounds like a fun job. Would have been fun to be in England this past weekend for MXoN. So on the scanners there are 3 main types that we deal with. IR Structured light, Blue Structured light, and Laser. There are other variations but most of the lower priced consumer scanners are IR Structured light, or with like the Mini2 from revopoint a blue light scanner. The reason that is important is because they do different things with the surfaces. Scan spray works well because it gives the parts a consistent surface finish that does a bit of absorption. So not as much worry for things like ambient occlusion on the part(shadows will kill your scans). The scanners all project a varying pattern of lines or grids onto the object and the depth cameras will record what they see and use that to capture the point clouds in 3d. If the object is over exposed (red marbling seen in the scan software preview) then it scatters too much and the data you get, if any, will be junk. If its under exposed then if you get anything back it will be noisy. When i say over/under exposed what I am actually talking about here is the brightness of the IR Leds that are used to illuminate the object. If there are light panels (not soft boxes) with direct light on the objects this can cause an over exposure for the depth cameras. I would imagine using a soft box could be a good solution. In my garage i have 5600k hex lights on the ceiling and if/when i use LED light panels i try to make sure they are indirect. Most of these scanners can't be used outside. Some exceptions to the rule of course but 30k lux is about the limit of anything I have played with. So cloudy day basically. But with if you scan in the dark or a darker room it really means that you need to crank up the IR Leds. You will have shadowing issues and there is a balance. If you can get a nice well lit room that is diffused I imagine that would create an ideal situation. I have found my best results come from scan spray (for most scanners) and the 5600k ceiling lights seem to do alright. But every scanner is a little different and finding what works for your scanner is key. For example a Creality scan ferret did well outside on shiny objects while other scanners throw up their hands and walk off.
Is there a discord for this channel? Would be cool to connect with other CAD users, new and experienced, to learn and compare problem solving techniques.
Most scan software is paired with the scanners and can only be used for that. In some cases you can import meshes but i think working with raw point cloud data without a revopoint scanner in revoscan wouldn't work unfortunately. You can bring raw point data into MeshLab which is free, but the user interface and workflow takes a bit to get used to
I love the series! Great job! I'm trying to reverse engineer my own scan (with mini 2) and I am following your tutorial. I have a question about smoothing. I'm having difficulty understanding the purpose of this process. I get the impression that after smoothing, it will be harder for me to follow the lines of my scan in Fusion 360. I understand that when the scan goes directly to printing, it's useful, but when we're creating our own model based on the scan, does smoothing have any positive effect, or can I just skip this step? I have one more question regarding the merge. I've done several scans as you recommended at different angles. The object I'm scanning has many gaps visible from the top, which are clearly visible in the top scan, but then the sides of the object are not visible. Therefore, I decided that merging two scans would be the best solution. However, the problem is that the scan taken at a lower angle so that the sides of the object are well visible produces a lot of noise in the gaps visible from the top. Unfortunately, this noise contaminates the internal part of the scan. As a result, I came up with the idea of cutting out the entire interior of the scan taken at a lower angle and leaving only the sides because only they are not visible in the first top scan. Unfortunately, according to the Revo Scan, it seems I have deleted too much because when I try to merge, I get the message: "the overlapping area is less than 10%" (even though I aligned it before cutting out the center in the second scan). Do you have any ideas on how to solve this problem? Once again, big thanks for sharing your knowledge!
Hey There. Ill try to answer what I can but know that it is somewhat subjective. Smoothing - When you scan something the points created in 3d can vary based on several factors. In a perfect lab we would hit the accuracy and precision claims of the scanner. +- .02mm. So this means the points can be high or low. Smoothing gives Revoscan the ability to "normalize" that a bit. So lets say you have points that are high and low. it can move them in 3d space (based on the Strength you set) to try and get closer to the real part. The curve ball here is that density, or the number of points captured, has an influence here. In the videos the part had a lot of points meaning there was next to nothing that smoothing did. If you simplify the points first then smooth you will notice a more drastic change on things like sharp edges. So can you skip it? Yes. Generally i skip smoothing at the points but might use smoothing at the mesh level. They behave a little different. At the point level its looking at moving points in space. At the mesh level its looking a bit more at triangle angles to adjacent triangles. Again density has an affect here. The more dense your mesh is, the less influence or change you will see. You kind of have to play around with your scans and see what works best. If you have a really noisy scan you might find you need to do some smoothing at the point level. I would say as a general rule (rules are meant to be broken), lean toward smoothing at the mesh level over the point level. With the merging, I haven't run into that one before, but what I would say is to leave more overlap if possible then use the Overlap detection on the merged points. As long as the edge of the scan isn't super noisy you should be able to merge with a good bit of overlap. One thing you might try is to align them, merge and then go back and edit the original. I am pretty sure that the originals move to the new locations (well the 1st selection stays). then when you make some more edits and go back to merge it might work better. You can also export the points from Revoscan to Meshlab which is free. You can align and process in meshlab and bring the points back aligned and merged. I plan to do a video on this in the future but its lower on the list.
Hmm- I went ahead and bought the Einstar mainly based on your review and videos on it- so disappointed to see you have now jumped scanners again? I will try and follow the high level processes and apply them in the Einstar software as otherwise this is great tuition and content!
Sorry to hear that Paul. I have never reviewed an Einstar though. I did use the ProHD scanner from Shining, but that and the Einstar are like $15000 apart.. The Einstar is a great scanner but its resolution (.2mm) wouldn't be good for small detailed parts like this. I still have a relationship with Shining 3d and will be covering more of their scanners in the future as well as others from Creality and hopefully other MFGs as well. While we are an affiliate for Revopoint (basically just means if you use the link to buy from them we get a small amount), but before signing on for that we were very clear that we will be covering other brands/scanners on the channel and that had to factor in.
I can't miss the chance to say hello to the man who taught me many of my skills. Excellent work Mr. Perez
My pleasure
A++++ you are the best teacher out there 🙂
Wow thank you!
Another great video! Thanks to the first video in this series I was finally able to get clean results from scans for my complex part but I suffered through trying to process the scans on the software. (Mainly because I had no references to what the options meant so it was a lot of trial and error).
After watching part 2 I feel as if I’ll have something of much better quality ready in time for the last video because in trying to import the final file to F360, I couldn’t for the life of me get it to become a nice solid file. I could only get it to become tens of thousands of triangles with no way to manipulate them similar to my native fusion designs.
Part 3 can’t come soon enough!
Awesome! hopefully in the next day or two it will be live. just needs final editing.
Killer video! Excellent explanations of Revoscan! My Miraco has been an absolute game changer for design work
Awesome! Glad to hear it. I love the Miraco! We will be doing another video soon looking deeper at the Miraco and Mini2 together. Scanning larger parts that are a bit tough for the mini2 just to see how to handle them.
as an ex-motocross journalist and photographer (meaning I think about light and how to use it, quite a lot) who just started 3d scanning I want to say that I think it is not better with darker environment or less light but I think it is about how the light is transfered (if one can say that this way).
I realized when I scanned in the morning and I had really much light because of the position of the windows the scans were so much better because scanning needs light with these low budget scanners. but exposure to artificial light that is too direct to the surfaces causes problems (like direct sun light) , it produces hard shadows and reflections and the quality and the feasability will go down the drain.
so I think/ or guess a good solution would be the same as the one that photographers use and get indirect light (like with a softbox or so). this means checking your "light-setup" in your garage and changing it a bit to produce indirect and soft light could potentially be a very rewarding thing.
these are my thoughts, correct me if I'm wrong !
Ex motocross journalist eh! :) sounds like a fun job. Would have been fun to be in England this past weekend for MXoN.
So on the scanners there are 3 main types that we deal with. IR Structured light, Blue Structured light, and Laser. There are other variations but most of the lower priced consumer scanners are IR Structured light, or with like the Mini2 from revopoint a blue light scanner. The reason that is important is because they do different things with the surfaces.
Scan spray works well because it gives the parts a consistent surface finish that does a bit of absorption. So not as much worry for things like ambient occlusion on the part(shadows will kill your scans). The scanners all project a varying pattern of lines or grids onto the object and the depth cameras will record what they see and use that to capture the point clouds in 3d. If the object is over exposed (red marbling seen in the scan software preview) then it scatters too much and the data you get, if any, will be junk. If its under exposed then if you get anything back it will be noisy. When i say over/under exposed what I am actually talking about here is the brightness of the IR Leds that are used to illuminate the object. If there are light panels (not soft boxes) with direct light on the objects this can cause an over exposure for the depth cameras. I would imagine using a soft box could be a good solution. In my garage i have 5600k hex lights on the ceiling and if/when i use LED light panels i try to make sure they are indirect. Most of these scanners can't be used outside. Some exceptions to the rule of course but 30k lux is about the limit of anything I have played with. So cloudy day basically. But with if you scan in the dark or a darker room it really means that you need to crank up the IR Leds. You will have shadowing issues and there is a balance.
If you can get a nice well lit room that is diffused I imagine that would create an ideal situation. I have found my best results come from scan spray (for most scanners) and the 5600k ceiling lights seem to do alright. But every scanner is a little different and finding what works for your scanner is key. For example a Creality scan ferret did well outside on shiny objects while other scanners throw up their hands and walk off.
Is there a discord for this channel? Would be cool to connect with other CAD users, new and experienced, to learn and compare problem solving techniques.
Sure is. Drop me an email support@caducator.com and ill give you an invite link.
I have a moose scanner. Can I use this software with the moose. This software looks better than JMStudio. This video is awesome. Thx
Most scan software is paired with the scanners and can only be used for that. In some cases you can import meshes but i think working with raw point cloud data without a revopoint scanner in revoscan wouldn't work unfortunately. You can bring raw point data into MeshLab which is free, but the user interface and workflow takes a bit to get used to
Love how informational your videos are. Is there still a discord channel?
Thanks! and yeah just email me support@caducator.com and ill share a link.
I love the series! Great job!
I'm trying to reverse engineer my own scan (with mini 2) and I am following your tutorial.
I have a question about smoothing. I'm having difficulty understanding the purpose of this process. I get the impression that after smoothing, it will be harder for me to follow the lines of my scan in Fusion 360.
I understand that when the scan goes directly to printing, it's useful, but when we're creating our own model based on the scan, does smoothing have any positive effect, or can I just skip this step?
I have one more question regarding the merge.
I've done several scans as you recommended at different angles. The object I'm scanning has many gaps visible from the top, which are clearly visible in the top scan, but then the sides of the object are not visible.
Therefore, I decided that merging two scans would be the best solution. However, the problem is that the scan taken at a lower angle so that the sides of the object are well visible produces a lot of noise in the gaps visible from the top.
Unfortunately, this noise contaminates the internal part of the scan.
As a result, I came up with the idea of cutting out the entire interior of the scan taken at a lower angle and leaving only the sides because only they are not visible in the first top scan.
Unfortunately, according to the Revo Scan, it seems I have deleted too much because when I try to merge, I get the message: "the overlapping area is less than 10%" (even though I aligned it before cutting out the center in the second scan).
Do you have any ideas on how to solve this problem?
Once again, big thanks for sharing your knowledge!
Hey There. Ill try to answer what I can but know that it is somewhat subjective.
Smoothing - When you scan something the points created in 3d can vary based on several factors. In a perfect lab we would hit the accuracy and precision claims of the scanner. +- .02mm. So this means the points can be high or low. Smoothing gives Revoscan the ability to "normalize" that a bit. So lets say you have points that are high and low. it can move them in 3d space (based on the Strength you set) to try and get closer to the real part. The curve ball here is that density, or the number of points captured, has an influence here. In the videos the part had a lot of points meaning there was next to nothing that smoothing did. If you simplify the points first then smooth you will notice a more drastic change on things like sharp edges.
So can you skip it? Yes. Generally i skip smoothing at the points but might use smoothing at the mesh level. They behave a little different. At the point level its looking at moving points in space. At the mesh level its looking a bit more at triangle angles to adjacent triangles. Again density has an affect here. The more dense your mesh is, the less influence or change you will see.
You kind of have to play around with your scans and see what works best. If you have a really noisy scan you might find you need to do some smoothing at the point level. I would say as a general rule (rules are meant to be broken), lean toward smoothing at the mesh level over the point level.
With the merging, I haven't run into that one before, but what I would say is to leave more overlap if possible then use the Overlap detection on the merged points. As long as the edge of the scan isn't super noisy you should be able to merge with a good bit of overlap. One thing you might try is to align them, merge and then go back and edit the original. I am pretty sure that the originals move to the new locations (well the 1st selection stays). then when you make some more edits and go back to merge it might work better.
You can also export the points from Revoscan to Meshlab which is free. You can align and process in meshlab and bring the points back aligned and merged. I plan to do a video on this in the future but its lower on the list.
Hmm- I went ahead and bought the Einstar mainly based on your review and videos on it- so disappointed to see you have now jumped scanners again? I will try and follow the high level processes and apply them in the Einstar software as otherwise this is great tuition and content!
Sorry to hear that Paul. I have never reviewed an Einstar though. I did use the ProHD scanner from Shining, but that and the Einstar are like $15000 apart.. The Einstar is a great scanner but its resolution (.2mm) wouldn't be good for small detailed parts like this. I still have a relationship with Shining 3d and will be covering more of their scanners in the future as well as others from Creality and hopefully other MFGs as well. While we are an affiliate for Revopoint (basically just means if you use the link to buy from them we get a small amount), but before signing on for that we were very clear that we will be covering other brands/scanners on the channel and that had to factor in.
The revo scan software looks much more powerful than the creality one
I would agree. The creality software holds them back.
@@LearnEverythingAboutDesign hopefully they are working on improving it. The hardware seems really solid