Видео

Hi, I am glad you find the video useful. In the cases where I don't have an occlusion map, I just set the colour to white. The difference is small enough to go unnoticed unless you're really looking for it.

I wish I had a ton of experience with Reality Capture to be able to answer this question more in debth! From the limited time I have spent with it, however, I would say Metashape does a better job at allining all images, even if they are taken from different sensors. Again, RC has improved over time and it might be just as good.

The pleasure is mine :)

Only if the view in plan is set to show parts. That’s the beauty of it. You can either shoe the Original (the wall as it always is), Show parts, or Show Both. The last one use useful if you need to see the parts but also tag the wall. Cheers

​@@danielkuzev4992Thank you for the clarification.

Glad to hear, brother!

I am glad you liked it :)

Hi, maybe not a video, but I can make a post on my website. Will make sure to reply here with a link once done.

The pleasure is mine!

Hi Michael! The short answer is yes! You can use it on any material. Just play around to your satisfaction.

Hey mate, It will probably take forever to explain in a commen here, so I made a post walking you trhough every step on how to scale a texture in Rhino and V-Ray on my website. Enjoy! www.danielkuzev.com/blog/resizing-textures-in-rhino

Thank you Sir! I appreciate it. With everything going on, it is difficult, but I have some ideas up my sleeve. And thank you for the topic recommendations!

Hey, you can absolutely use it, but the reults will vary! The best way is to try it out and ses.Depending on how grainy the surfae is, it might not look great ;( But hey, the more you try the more you will know.

Also at 12:12 the normal map needs the green channel to be flipped. You can see the light hitting the paving stones on the wrong side.

You’re quite possibly right, but then I look at the results and I’m pretty happy with it. And I am a big fan of not putting the AO map at all, especially on larger projects since you’re removing an entire map that needs to be loaded. As for the normal, I don’t know mate. Some companies give you the DX and the OpenGL normals and there you can see an inversion of the details… but here it looked pretty spot on.

@@L3nny666Also, genuinely, thank you for putting time stamps. That’s the kind of feedback that makes sense! I appreciate it.

@@danielkuzev4992 yeah the best way to do it is: make it look good, that's all that matters in the end. nobody is looking at a picture wondering in which slot you put which map.

Yeah the AO map should go in the Diffuse channel. Right Click diffuse channel -> Wrap In -> Mix Operator -> put AO in slot B -> choose Multiply option

Hey, here is the setup: CPU: 10850K GPU: (2) RTX 3080

Hi mate, I am verry happy to know I was able to help! Keep up the good work!

Hey Joe, I never import 3D files into Revit. In general, Revit isn`t that great with meshes imported from other softwares (except for Rhino nurbs surfaces... those it loves). The main difference between all the formats you mentioned and an .rcp file is that inside the .rcp you DON`T have a 3D model, but just a point cloud. And to be honest with you, if I need to model something in revit, I always go through a point cloud. In short; export as an .e57 format (make sure you use meters as your local system), start a new Recap project, import the .e57, wait for it it process, and you should be good to go.

@@danielkuzev4992 thanks for the feedback. I am just trying to think of ways we can use drones where I work (Arch/Engr design firm). For example, when we do an expansion project to an existing building, and there is no model (or sometimes even drawings) for that building, or architects are modeling that from scratch sometimes to show finished renderings to the client, so they can see what it will look like with existing and new construction when complete. I was wondering if some data captured with a drone (or any scanner) could at least cut down on that time instead of starting from scratch. I am not sure if they would be able to use a point cloud to start that process.... maybe so.

Ho Jose, The drone did not take a video. It took a series of photos. I set it to an interval shooting so it took a photo every 3 seconds. Otherwise, you can always put a vide inside Premiere Pro and export each frame. I am just not sure how many thousands of images you will end up with.

Haha, thank you! I am glad it came in handy!

Hi. If you are using photos taken with a phone or a camera (not a drone), this is completely normal. I would advise that you rotate your model in your 3D modeling software afterward. I usually use Rhino, so I would make a vertical, or a horizontal, line coming from one of the markers, then put myself in FRONT and LEFT views, rotate twice and I am good to go.

@@danielkuzev4992 But it is impractical and not very precise. I solve it by creating a control file with 3 points in the same plane (such as a table), which have a distance known to each other and that form a rectangle triangle. Each point assigns XYZ coordinates with respect to a 0.0.0 point and those points are assigned to the control points in the photographs before aligning the cameras. And it works, the model is created horizontal and at the correct scale.

Sorry mate, I can’t tell you how to get a cracked version.

Hey mate! I really appreciate it! I appreciate the detailed feedback too. It’s quite nice to see that it is being put to a good use! It has been on my mind to make more tutorials, especially about lighting, but it is on the back burner as I am running a scanning practice with my gf. Hopefully I can make some time for it soon.

Thanks mate! I am glad it came in handy!

Hey, apologies for the late response. The short answer is that you can make parametric tiles (and therefore bricks) using the tile texture in place of the Colour map. But from my experience it is t the most straight forward option. When I needed to do this, I usually tile the texture, say, 4 times in Photoshop, do the changes I need there and I’m good to go. Usually the only thing I need to change is: 1 - The colour map - changing shades 2 - The roughness map - usually making some parts more reflective by painting with dark grey. I know this isn’t the answer you were hoping for, but I hope it helps.

Thé pleasure is all mine! :)

That’s quite heartwarming. Thank you Jorge! And I’m happy it was useful!

Oh absolutely. Just make sure it is a closed mesh rather than an open one. You can do this either in Rhino or in Blender (if you’re looking for a free version)

@@danielkuzev4992 thank you very much for the reply and the tips!

M glad it was helpful, mate! Let us know how it goes and if you share the results somewhere, drop the link in here ;). Cheers

There are no waypoints on the Mini, unfortunately. So I just set the timer to 2sec and started flying around slowly to ensure enough overlap between the photos.

I am seeing this a bit late. But one thing you could try is to use the Nvidia denoiser. It works miracles. Another thing, use the same material in as many objects as possible. Also don’t use 4K textures unless you ABSOLUTELY need them. And finally, use blocks in rhino. So it loads the geometry once rather than tens or hundreds of times (ex. Having the same window all over the place)

um maybe u can duplicate model, select the roof to delete it, and u can decimate the others part. So basically u hv two parts, high poly roof and low poly body building. *optional: i think its easy if u already export to 3d software and decimate on it

I’m glad you found it useful. Cheers

Hey Riccardo. That's a simple one: nowhere :D . In reality you don't need an additional map to take up more memory for each material you create, and what you get from following my steps is as realistic as it gets. I don't know what your setup is, but if you want to cut down even more the maps to use, you can skip the AO map and set the slider to white. All of this is just optional of course. Have fun!

@@danielkuzev4992 Thanks for the reply ! I totally get what you mean but I was asking because the website I use for texture provides only diffuse, displacement, normal, AO and specular map and I was not sure where to put AO and specular. Usually I mix diffuse and AO and use specular as reflection glossiness / roughness. Let me know what you think !

Hi Edward. From what I have seen so far (always keeping the Mini in mind) that it can be pretty accurate. However, a report from Metashape tells me that the precision of the location of each photo can be anywhere from a foot to several meters. It is important to keep in mind that there is a difference between tolerance and what it actually is. Even if the tolerance is large, you could still have meaningful results. One thing you can do is to have a scale bar in your scan and put the distance between the two ends manually. Alternatively, you can use two points far away from each other and measure the distance with a laser and then force metashape to keep that distance.

That’s helpful to know! I’ve got to try this out for myself to learn more.

Thanks mate. It’s kinda impressive what can come out of it, but keep in mind that it can’t guarantee you anything resembling a survey precision (you’d need an RTK drone for this) I’m happy you found it interesting :)

My pleasure ☺️

Hey Joao. Yes you can but I can’t promise you the performance will remain the same. There are A LOT of triangles by the end! But you can export in .fbx format and include the texture ;)

Tbh, I haven’t tried the other ones. I’m getting an rtk drone this summer that can me controlled through drone deploy and I’m excited in trying it out. So far, Metashape has been a blast and they just added the ability to align drone scans with terrestrial LiDAR scans and I might just stay in that environment for a while more.