It turns out it’s very hard to render a prism! This might be partially due to my geometry being off. Something else is going on. I have rendered this in a few different applications. This go around it’s Cycles (of course) and D5.
I have a feeling the internal angles are off and that is affecting the location of the reflected lines, but that doesn’t cover everything that’s missing. In addition to the angles, when I modeled this I made the internal mirrors copper colored. The exterior glass should actually have the copper color (it’s a coating). This alone might be half the issue. It seems like light just isn’t getting inside the prism and bouncing around like it ought to. I believe the reflection of the lines comes off the inside of the glass (checked the normals and they look okay… but might be wrong there too).
Here’s what the prism looks like irl. I also have one here at home and would snap pictures if there’s enough interest:
I pulled the mirrors closer towards the center. It looks like they have to come even closer. I’m so curious I almost want to crack open the one I have at home, but that would be an expensive curiosity. A more economic and less mentally lazy way would be to do a bit of research + trig.
Thank you for that!! Not sure if you played with the plastic material but it looks perfect! The updated file is more true-to-life in that the mirrors are actually mirrors and the copper color is moved to the exterior glass. I’m guessing on how to handle some of the materials: The lines/crosshairs have emission (maybe unnecessary as I haven’t checked since fixing the geometry). The exterior glass is partially metallic. I think this is a rare case where that’s actually appropriate given the copper coating.
just came back from watching ironman event and saw this post thought id end the night having a crack at this
used keyshot if i can know what materials is inside i can do it more accurate i think cant find much info, i tried copying what materials you had in rhino and giving it my touch but this is the result
Presumably, one must also be careful of the other settings… environment, lighting, and the camera lens will influence the effect. I was suprised at the change in going between lens lengths.
I reconstructed it to be as reasonably physical as I think I can get it (in 15 minutes before my bed time!). This is a bella bevel material on the glass, which applies at the point of render, and I think is my standard 0.25- to 0.5-mm corner radius/chamfer (which seems normal for CNC stuff, so I have always used it, on everything. )
Looks amazing!! And ya, both my model and my materials aren’t the easiest to work with. You’ve taken this quite far.
If I hold the actual prism at the same angle it looks virtually identical. You also fixed the internal geometry without having the actual prism for a reference (I think!). My brain has only just figured it out and I have the prism right next to me (more or less).
Figuring this thing out has kind of helped me. I’m trying to determine exactly what’s in these things. There’s really not much. Lots of math and brain power goes into the shape but after that the materials are pretty basic. These cost upwards of $1,000 USD just for a single prism. I’ve seen some listed for over $2,000. Chinese companies manufacture both cheap ‘knock-offs’ and legitimate budget replacements for them. It’s hard to tell which is which, but the cheaper replacements are a viable alternative for all but the most precise and sensitive work. The key is to just make sure you don’t end up with a cheap knockoff.
I think it is just 6 (3x3) rotated retroreflectors. They are chopped up corners of cubes, which are silvered, tessalated into eachother, and have some form of other bronzed dielectric triangular face on them. This is “fixed” in the geometry I guess, because as soon as you remove the misleading “prism” term, you know it must have purely orthogonal surfaces.
For example, the above render merely uses a bella silver IOR, and the faces are a bella quick material with a custom colour.
The mystery lies in the geometry of the corners. This could be something like a half silvering effect, but it seems a bit drastic to do this. It is this mystery which provides the glowing corner features. In reality, they are not glowing. I didn’t have time to do any further testing on the edges.
I really don’t have good render skills and don’t understand camera functionality fully, it’s sort of knowing what it does instead
But Keyshot has become so fast with GPU rendering in real time it feels like I’m painting on a canvas and giving everything my touch everything is so accessible and tweakable I know where everything is
So sometimes I get good shots on some of my projects sometimes I dont i can spend a good hour or 2 micro changing stuff and just step away from the monitor like im watching some art museum exhibition from afar just to get another perspective and views
3rd time is a charm lol! David’s model is still better - it has the center ‘dot’ true to the genuine Leica prism this is based off. Mine is actually a Chinese one (a good one I tested it) but lacks the little center mark. The reflections now do exactly what they’re supposed to:
When I first modeled this I was at a waaaaaaaaaaaay lower skill level in Rhino. I remember I didn’t know how to copy stuff and literally re-modeled things multiple times (each prism face was different too because my geometry was pooched. I figure it was all easy stuff… so maybe 5-15 minutes back then… now it just takes seconds. Something I haven’t seen many people do is using the gumball to mirror (-1 scale) - only learnt that a few months ago (huge time saver!!!):
When I have time I want to try another D5 render. Maybe even do an exploded animation. The geometry was quite a bit different than I originally though but makes complete sense now. Basically lots of 90 degree corners and equal length edges. Maybe I was let astray by an image with an incorrect aspect ration who knows!! I should have taken advantage of having the reference object close by.
I think a while back I mentioned that your shaded mode looks better than my renders lol!! The way you present both in Shaded mode and Rendered are both very visually appealing, especially for industrial design.
… and sorry one more edit : It’s still not modeled “true to life” as each reflector assembly would be a chopped up cube like David mentioned. One face (in this example) would have the copper coating and the other faces would have sliver/mirror coating of some kind. To be honest I don’t know the most effective way to model this, but just for visual purposes it works as-is. I could get rid of the overlapping faces I guess but I also have real work to get to finally hehe…
Here’s another one in case anyone wants to render it. Search “ATP1 Prism” for reference pics. Modelling this one KILLED me I’m still stunned it was so hard. Even the fairly basic stuff kicked my butt. And I had to make a lot of compromises.
Due to the shape of this one there are some neat reflections. The real one has some stuff going on that I couldn’t figure out. Due to challenge level and time constraints I had to simply mine.
I believe at some point I actually had the geometry much closer. This was racking my brain so bad I had to bail and just get it done. It’s amazing how much different the internal geometry appears to be just based on the reflections.
Hopefully I’ll have time to render this properly (and maybe make a quick animation) in the near future. I haven’t even had a go at the previous one yet since I fixed it! As always I love to see what you guys can achieve!!
From the Leica white paper on these, I believe the copper coating is on the large internal face and the three external faces have an unspecified anti-reflexive coating designed to minimise direct reflections from those faces (to avoid slightly off distance measurements). No silver involved. Copper used because it suits the wavelength of the laser beam.
I’m just off on vacation, but I will try and remember to share a ray plotter I knocked up in grasshopper when I get back. Fascinating how light bounces around inside these prisms.
Regards
Jeremy
EDIT: [Written in haste, repenting at leisure…] The large face is, of course, actually the external face and has the anti-reflexive coating, the three faces which are actually internal have the copper coating (and possibly also the fourth internal face created by truncating one point of the prism, but I doubt it). Sorry for the misdirection.
Originally I had the interior with the ‘copper’ color as it gave me the impression that the color was coming from inside. I read (misread) something that made me think it’s on the outside so I changed it. So the ‘fixed up’ version is wrong… but luckily the rendering looks almost the same.
By “Silver” I just meant mirror. I think I’m just using a ‘silver’ color but ya it’s just a mirror (I think!!). It’s interesting to dissect what goes into these given the huge price variance between the “genuine” ones and the knock-offs ($100-ish versus $1,000+). So far the biggest difference I can tell is just better gluing.
Early on I was adding a bit of “metallic” to the outside glass as one guess as to why my reflections were off/missing was that the glass was reflecting internal geometry. After some though I realize that would conflict with the functionality of these quite a lot - it has to bounce the laser back to the survey instrument after all - one big advantage of using these is that you can shoot at them from way further away compared to other survey targets.
Modeling these exposed a lot of flaws in my Rhino modeling ability! But also helped me learn and understand them better.
I think while the prism inside looks a littler dull, I suspect that this may actually be okay. My reasoning is that these objects are often imaged in very bright photo studios. However, in the real world, they are probably mounted at about 1.5 meters high, with the reflections coming from the floor and duller surorounding area, from the perspective of a person anyway. Don’t really know.
And you’re correct: It’s usually attached to something. In my case usually at eye level (170cm). The top and bottom is meant to have 5/8” (M16) female threads each end - I think it’s aluminum. These can be mounted on a survey pole or mounted to structures and used for construction site control/reference points.
This requires Anemone to be installed in Grasshopper.
The tool is specific to the associated prism found in the testbed file - you can’t go designing a telescope with this! Use the 2D slider to shift the source of the ray in front of the prism. Rays that bounce back parallel to the source are blue. Rays that shoot off at an angle are red. Rays that miss the prism stay blue (a bug I couldn’t be bothered to fix, given the limited application of this simulation).
This is illustrative only: the ray paths will be affected by the IOR of the glass used (I chose 1.66 which is in the ballpark for optical glass). On Leitz prisms, at least, there is also a surface coating which will have an effect, but for which technical details are not available.
The “cube corner design” seen here is widely used in metal radar reflectors and the ray returning parallel to its incoming direction well know. The solid glass prism disrupts that, to some extent, because returning rays can hit the outer face at an angle to the surface normal that exceeds the critical angle for the glass to air interface and are therefore reflected inside the glass instead of emerging.
)
: It’s still not modeled “true to life” as each reflector assembly would be a chopped up cube like David mentioned. One face (in this example) would have the copper coating and the other faces would have sliver/mirror coating of some kind. To be honest I don’t know the most effective way to model this, but just for visual purposes it works as-is. I could get rid of the overlapping faces I guess but I also have real work to get to finally hehe…
