Jump to content

Gallery of 3D scans with cool viewer


ozRocker
 Share

Recommended Posts

I've just added a gallery to my website where I've put some of my scans.  WARNING: Not safe for work!

If you click on one of the scans you will see the babylon.js viewer that I created to allow you to interact with the model.  I have panning working for mobile devices (use 2 fingers and drag) but its not perfect and sometimes conflicts with the pinching for zooming.  I still haven't figured out how to show quads in the wireframe which is a shame because triangles make the topology really confusing.  I wanted to use it as a selling point to show people how lean my topology is.

www.punkoffice.com/gallery

Link to comment
Share on other sites

4 hours ago, Pryme8 said:

Pretty cool but you you should have versions of them in tPose being displayed and maybe in a simple walk animation if possible.

Whats the point/poly count on these.

Yes, that is part of my plan.  Some of these people have been scanned in a-pose so I can turn them into avatars.  When I post them up the viewer will have buttons to allow you to animate them.

Poly count is between 6,500 and 8,000 quads.

Link to comment
Share on other sites

30 minutes ago, Pryme8 said:

You should be able to get them down to like 2.4 and then bake the normals from the 6k ones.  If you wanna Skype sometime we can go over retopo methods, I'd like to see what your doing I assume your using zbrush for the retopo?

Interesting.

I am actually using a normal map from a 200k model.  Funny thing is, I knew nothing about normal maps or 3D game dev until I joined this forum.  I showed everyone my 200k models and they were like "That's too big!  Professional games have high-def models and they don't even have 200k.  Use normal maps".  I didn't even know what a normal map was until then.

I really REALLY wish there was a way to show quads in the wireframe instead of tris because then you could actually see what my topology is and point out the bits that are unnecessary.  Wireframe is kinda useless in this instance :(  Yes I'm using Zbrush for the retopo.  I always believed I could do better than what the Zremesher gives me so I tried manual retopo tools but they took too long.  I wish there was something in-between, like automatic retopo that you can manually adjust afterwards to remove useless polys.

Add me on Skype marcus.milne1 (anyone in this forum, feel free to add me).  If I can get this down to 2.4 while still preserving important edges (nose, chin, elbows, lips) then I'd be stoked!

Link to comment
Share on other sites

Yes ill show you how if I get my system up and running again by tomorrow there are quite a few tools are your disposal!
If I had a rig like yours with the camera set up there would be nothing stopping me from pumping out a dumb amount of content for games.

Link to comment
Share on other sites

21 minutes ago, Pryme8 said:

Yes ill show you how if I get my system up and running again by tomorrow there are quite a few tools are your disposal!
If I had a rig like yours with the camera set up there would be nothing stopping me from pumping out a dumb amount of content for games.

ok, you can now press SHIFT-N to disable normal map in the 3D viewer.  If you click HIDE TEXTURE then turn off the normal map you'll see the topology more clearly.

I am scanning as many people as I can, however it all depends on who I can get to come in to get scanned.  I've run out of cool looking friends to scan and I don't have money to pay models so I've been going to Meetup events and modelling websites to network.  The models I've met have proven to be very helpful though and they're spreading the word to get their model friends to come in for a scan.  I've noticed that only the ones confident about their looks will volunteer to get scanned in.

If you want to scan inanimate objects you can just use one camera for that, no rig necessary.  I'm focusing on people (only have so much time in the day) but I can do object scanning if people ask for it.

Link to comment
Share on other sites

Here is a good tutorial for people who want to get into scanning https://blog.sketchfab.com/how-to-set-up-a-successful-photogrammetry-project/

All you need is one camera, even an iPhone camera can produce great results.  I've attached a photo of what a raw head statue scan looks like from the iPhone 6+ camera (my first attempt).  I'd be interested to see what other Babyloners do with scanned assets.

statue_scan.jpg

Link to comment
Share on other sites

7 hours ago, JCPalmer said:

Also, I think trying to display quads for wireframe is hopeless.  That is a webGL switch not something BJS does other than to switch on / off.  Whatever exporter you are using is converting quad & ngons to triangles, anyway.  Do not know how you would even put back to quads to begin with.

Thanks the info about the "texture" button.  That's important!  All fixed now.

I had hope that showing quad wireframes would be possible because Sketchfab does it straight from my .OBJ files.  They aren't using a special texture either because the wireframe stays the same thickness when you zoom in and out.  If I use that same .OBJ file in Babylon.js it will only show triangles

sketchfab_wireframe.jpg

Link to comment
Share on other sites

6 hours ago, Pryme8 said:

I do it with one camera all the time, I am very knowledgeable on the whole process, but nothing can replace a multi camera rig with a good reference grid in the background.

How come you can't put a reference grid in the background?  A single camera can do anything a multi camera rig can do except capturing moving objects.  Also, the process is a bit longer with a single camera.  But you can get better detail with one camera 'cos you can take as many photos as you need and you are free to be as close/far from object as you want.  If you have a multi-cam rig you are limited to the number of cameras in it.  Also it can be hard to re-configure if you want to move cameras to different positions and angles.  People with high-def multicam rigs will have them configured for a specific volume so they can get high accuracy.  The companies like Ten24 and Infinite Realities don't do animals or objects because their rig is configured for the dimensions of a person.

Have you posted your scanned assets anywhere?  I'd be interested to see how they turned out.  I'm teaching my web designer how to do that so she can handle object scanning if a client ever asks for it.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...