Ocarina of Time Zora's Sapphire Ring

Greetings, internet traveler!  Nearly a year after I started and then shelved the concept, I have now finished creating my own engagement ring inspired by the Zora's Sapphire from the Legend of Zelda: Ocarina of Time.  This project took a lot of power, strength, and courage!  I've never made any sort of jewelry before, let alone something requiring gem settings.  To be honest, I feel like I really lucked out on this one.

This ring was designed in the 3D modeling program Maya, printed in wax then cast into gold by Shapeways, and then fitted with gems by a local jeweler.

Ugh, pretend my hands don't look so bad

Step One: Modeling

As usual, this project started with a model.  Since I've never modeled a ring before, most of this process was trial, error, and correction.  Over and over.  Seriously, I'm pretty sure I started over at least twice, not including the times I scrapped the band.

My first two "final" models

Last model before printing the end result

A ring capable of holding gems is a complicated thing, so there were many factors I had to consider when making the model.  For readability's sake, I'll make a list.

• Design: perhaps the hardest part of all was simply choosing how to make the darn thing look.  I wanted to stay as true to the game's design as possible, but translating a low-poly 3D mesh that floats in the air into a physical thing meant making some changes.  The most obvious is probably the gems-- pentagons aren't really a normal gem cut, so I chose to arrange the prongs themselves in a pentagon and just use standard diamond-cut gems.  I wasn't sure it would work, but I'm happy with the end result.
• Band size: not every ring is going to fit every finger, so getting the size right (or at least reasonably close) is important.  My US ring size is about a 5.5.  Conversion charts tell me that a 5.5 band has a 16.1 mm inner diameter.  Luckily Maya has separate options for setting overall radius and section radius, so getting the rough sizing going was pretty easy.  Do keep in mind that Maya isn't exact, though-- you might need to measure in a separate program like NetFabb and tweak it a bit.  I went with a 1.3 mm thickness and then flattened out the inner edges, which fit my finger pretty snugly.
• Prong thickness: this is one I really struggled to find information on.  As I have no formal training as a jeweler, I kinda had to make an educated guess and used Shapeways' metals printing guide to choose a thickness for the prongs.  Minimum recommended thickness of unsupported wires is 1.0 mm?  Works for me!  I left some extra length on them, as well, so that the professional jeweler would have plenty to work with.
• Overall gem seating thickness: again, I defaulted to the Shapeways guide because I didn't know what else to do.  Just about every part of this ring is at least 1.0 mm thick, minus the tips of the wings coming from the base triangle.
• The gems themselves: as you can see from the render, I went ahead and modeled in a 5 mm diamond-cut gem to use as a reference while arranging the prongs.  This part of the design made me the most nervous.  Traditional gem setting has the jeweler cutting a "seat" into the prongs to properly fit the gem, while I modeled the prongs to already fit around the gem.  Unfortunately, I modeled the gems themselves wrong, so this seat looks a little wonky in person.  The jeweler was able to work around it, though.  I'll revise this for model 2.0.

There was one thing I didn't think about, however-- how the ring would actually sit once it was put on a finger.  The final design has it standing up higher than I had imagined, meaning it's prone to catching on and knocking against just about everything it comes into contact with.  This is mostly an issue caused by my final band design, which I tried to make a little more fancy when I shouldn't have.  I will want to correct this for version 2.0, of course, but it'll do for now.

Step Two: Printing

Thanks to Shapeways, printing was the easiest part.

Test prints

I went through three different test prints.  The first one in plastic, seen on the left, was a little too fragile and didn't quite make it through the printing process.  For the second one, center, I scaled up the base triangle a bit to give it more bulk.  I also went with a thicker band, which ended up not quite fitting with the style I was trying to create.  The last one was my metal test.  I had this printed in bronze in the hopes that a jeweler could use it to test the gem settings.  Unfortunately, this led me to finding out that most jewelers won't touch bronze because it's so brittle.  We had to try several stores before someone would agree to even take a look.  Once we finally got an OK from someone who knew what they were doing, I took a deep breath and ordered the gold ring.

Step Three: Finishing

I was on pins and needles for a couple weeks waiting for the gold ring to come.  All I could think about was the potential for things to go wrong.  I was not going to be happy if I had spent a bunch of cash on a hunk of broken ring.  Eventually it did arrive and appeared to have all the prongs, so we took it back to the jeweler to have the gems set and the polygons polished out.  The jeweler we went with offered birth stone jewelry, so they picked us out some Swiss blue topazes to use instead of the super cheap cubic zircon I had bought as test gems.  A week or two later, the ring was done.

Do do do dooo!

<3

During this time, my boyfriend and I decided it was time to settle down.  We didn't exactly know how long this process was going to take, so we kinda jumped the gun and had the proposal before we got the gems set.  Oops!


I'll eventually get around to making corrections to this model and then releasing it for sale on Shapeways.  For now, I kinda want to just enjoy knowing that the ring on my finger is completely one-of-a-kind and represents me and my fiance better than a store-bought ring ever could.  End of sappy post.

Ocarina of Time Coasters: Part Five

Hello!  Hey, listen!  Welcome to the fifth and final part of my Ocarina of Time coaster project!  In this part, I'll finally get around to finishing these things so I can move on with my life.  For the previous post, see Part Four: Casting.

Part Five: Finishing

After giving my coasters a day or so to fully cure, I peeled them out of their molds to see how they turned out.  I'll be honest, they're not great.  There are still some surface bubbles and rough edges in the molds themselves, which transferred directly onto the casts.  I cleaned up what I could with my X-acto knife and sanded down any jagged edges I created in the process.

Since my originals didn't get a super smooth finish, the casts came out a little dull.  I tried to remedy this by giving them a quick spray with some glossy sealant I had on hand.  For whatever reason, this ended up being a terrible idea.  The sealant ended up pooling up on the surface in some spots, causing ugly little bumps I can't even reach with my sandpaper.  It looks like I'll need better molds to get a shiny surface on these coasters!

Despite this setback, I decided it'd be best to just continue on now that I'd gotten so close to being done with this set of coasters.  I bought a roll of adhesive-backed cork-- originally meant to be used as a shelf liner-- to use as a backing for the coasters to prevent them from sliding around and scratching up the tables they'll be used on.  Cutting them with my X-acto knife was a little messy, but it let me hold them down and cut an outline into the backing before I peeled off the paper.

My poor table gets cut a lot...

At long last, I've gathered the six medallions!  This project ended up being a lot more complicated and time-consuming than I expected, and I'm glad to finally call it done.

Woo!

I had originally thought to sell these, but considering how wonky they look and how many man-hours they require, I'm not confident that's a route I want to take right now.  Either way, I've learned a lot from this project, and I hope you have gotten something of value out of reading these posts.  Happy crafting, dear reader!

Ocarina of Time Coasters: Part Four

Click here for Part Three: Making the Molds

After an eternity of sanding, molding, and cursing, I finally finished all of my molds and could start the casting process!  The end of this project is finally in sight-- yay!

Part Four: Casting

Compared to all the other parts, casting was probably the most frustrating to figure out but the most relaxing to actually do.  Once I got in the groove, settling down to cast all six coasters in one go was kinda pleasant.  However, getting into that groove really tried my patience, and I nearly gave up on this project while trying to get a good cast.

The first issue I ran into was bubbling.  So much bubbling!  This was really discouraging-- after all, if you can't get rid of the bubbles, how on earth are you supposed to get a good cast?  The cause turned out to be three different issues that combined together to make a mess.  First, I was trying to use a clear-cast but fast-curing resin, which are very much prone to bubbles if you don't have a $200+ vacuum chamber sitting around.  Switching to an opaque resin helped a lot.  Second, I was trying to cast in 110-degree weather, meaning my pot life of 3 minutes was turning into a pot life of like 50 seconds.  Luckily, the fall came in while I was trying to solve my coloring problem that I'll cover later.  Third, I kept trying to use every sort of dye except for the expensive stuff meant specifically for use in resin.  This was adding moisture to the resin, which in turn made more bubbles.  I'd imagine a slow-cure resin or vacuum chamber could help with this, as well, but it was much more cost-effective to just buy actual resin pigments.

Some of my ill-fated early attempts

The second issue I had was deciding on how much of those fancy pigments to use, and then how to consistently measure that amount each time I mix up some resin.  Switching to opaque resin meant that I had to add enough dye to overcome its natural white cure color and not accidentally create pastel colors.  And since I have six different colors going on, mixing up a large batch for each color (as is frequently recommended) didn't seem very cost-effective.  So after many failed attempts and some bad advice off the resin casting forums I visit, my dear future brother-in-law recommended industrial syringes.  It feels a little creepy to have a bunch of syringes lying about, but I can get (mostly) exact measurements now, up to 0.01 mL accuracy.  After some trial and error, I decided on about 0.1 mL per cup of resin, which I'm using approximately 50 mL of total (25 mL of part A and of part B).  That's well under the recommended max of 3% dye per total volume that Smooth-On lists on their site, so it shouldn't be enough to interfere with curing.  Hopefully.

Fifth time's the charm, I guess?

The last major issue was simply that the molds ended up kinda crappy, partly because of the issues I listed in the previous section and partly because my originals just weren't that great.  Unfortunately, this isn't something I can just fix-- I'd have to completely start over again to get better casts.  Considering I still haven't finished version 1 and recently lost all my 3D models in a computer crash, this isn't exactly something I'm willing to do right now.  That being said, the pulls I'm getting could certainly look a lot worse, so I'm going to focus on doing the best I can with them until the molds fall apart and need to be remade.

Issues aside, I'm super excited to go demold my casts and see how they turned out.  It looks like I could revise the dye measurements slightly on some of the pieces-- especially the red of the Fire Medallion, since it turned into hot pink instead-- but otherwise this round of casts look like they just need to be finished off and then tested.  And by tested I mean scattered around my apartment as decorative coasters!

My phone isn't great with colors, but here they are in their molds

Hopefully this has been educational, if not kinda interesting.  My next post will cover finishing off the pieces to make them into actual coasters, as well as do a recap on what I did and maybe what tools I used.  May the super glue be with you, not on you!

Ocarina of Time Coasters: Part Three

Click here for Part Two: Sanding and Finishing

Welcome back!  In this post, I'll cover the process I'm using to create the molds for my coasters.  So far, creating the molds has been the easiest part of this entire process, as the issues I've had mostly stem from me doing a bad job of finishing the originals (see Part Two for details).

Part Three: Making the Molds

Since these coasters are flat and really don't have anything complicated going on, I'm making simple one-part molds.  I start by building a little cardboard box to contain the liquid silicone, trying to leave about a half-inch around the edge of my coaster to give the silicone some strength.  Some people use acrylic sheets to build their mold boxes, but I'm getting so many things shipped to my home lately that I'm just about swimming in cardboard, so it's the cheapest option I had.  If you follow in my footsteps, just keep in mind that cardboard is going to absorb moisture, which will make it a little tough to remove the box once the mold sets up.  To seal the box, I used a gratuitous amount of hot glue around the inner and outer edges.  I also used a ring of hot glue to attach the coaster to the bottom of the box, but I'm sloppy with hot glue, so I still had some silicone to trim off the bottom when I was done.

Derpy little cardboard box

The next part was super easy, despite my fear that I was going to completely screw it up.  I took my two-part silicone (Smooth-On Mold Star 15 Slow, as you can see from the picture), mixed equal parts A and B, and slowly poured it into a corner until it covered the model with about a half-inch thickness.  I kinda guesstimated the amount of silicone to use by measuring the box itself, getting its volume, and subtracting an approximate volume of the coaster to get 200 mL of silicone needed.  After pouring, I tapped on the sides of the box for a couple minutes in the hopes it'd help release any bubbles trapped on the surface of the coaster, and then let it cure until the instructions that came with the silicone said it was safe to touch.

Apparently I can't decide on a camera orientation when I take these pictures...

After it cures, you can just rip the cardboard box apart to free your mold.  I had to gently trim the edges of the mold away from the cardboard to loosen it, as it soaked into the cardboard along the top edge.  Once the mold was free, I gently pried the coaster out and admired my sloppy handiwork.

Yay!

So far I've had two main issues with my molds-- using a rough original, and not using a flat surface when I pour the mold.  The former is obvious, but the latter I didn't even think about.  Since my crafting table dips inwards in the center and my porch seems to be at a slight slant, both test molds have been very crooked.  The severity of this issue didn't become obvious until I had a cast spill over one edge but not fill the other side of the mold.  I'm not quite sure what to do about this at the moment, other than perhaps take note of the alignment of the mold relative to both surfaces and be sure to keep it in the exact same orientation when I cast, which sounds pretty error-prone.  Trying to get my table perfectly level might actually be easier.  I'll have to do some more tests to figure this issue out.  Until then, may the super glue be with you, not on you!


UPDATE 8 November:  Today I learned that I've been butchering my molds.  Trimming the edges around the original is a terribly dangerous thing to do, and can leave you with rips, dents, and other oddities that sure as hell will show up in your casts.  Since hot gluing the original to the bottom of my mold box has been the culprit, I'll be using some spare (sulphur-free) clay as an adhesive for my next mold.  Hopefully that will make it easier to get a nice clean edge.  Replacing my X-Acto knife blade ought to help, too, now that I think about it...

Ocarina of Time Coasters: Part Two

Click here for Part One: Modeling

Welcome back!  I'm slowly but surely making progress on these damn coasters, and am now at the stage where I need to give the originals a good finish so they'll actually make good-looking molds.  This should be a shorter post than the previous one, as the process isn't super exciting... except for maybe the parts of it I messed up, of course.

Part Two: Finishing My Originals

Due to the wonky ways I've been printing my parts, they don't exactly come off the printer looking fabulous.  In fact, the edges that touched the supports can get pretty jagged.  This is where my dear friends sandpaper and Dremel tool come into play.  Pardon the lack of pictures-- the sanding process makes an absolute mess, and I don't want to get plastic dust all over my shiny new phone.

To give a rough overview, I've been using a sanding drum of unknown/ very low grit on the Dremel to hack off the more jagged pieces.  After that, I had no problem scrubbing away with a sheet of 220 grit sandpaper attached to a foam sanding handle, which it came with in a small kit from the hardware store.  Once I can't feel any jagged pieces with my fingers, I've started giving it one more go-over with a wet/dry sanding block that I bought years ago to use on clay sculptures.  I have no idea what grit it is, but it's pretty fine even compared to the 220.  Iterations Three and Four have benefited the most from this mystery sanding block, as I can actually reach the inner surface of the coaster and sand it until it's nice and smooth.

Oh, something important I learned!  Don't rely on your eyes when trying to sand 3D prints, as the thin walls will let light through in some places but cast shadows in others.  This'll cause a lot of weird discrepancies between what the surface looks like and what the surface feels like.  I thought I had a big gouge in the middle of my Forest Medallion print, but it turned out to just be a thin part on the backside of the wall.

Now, as to the matter of the prints I ruined...

Ugh, this again?

So while I was getting started in home printing, I came across a neat product by Smooth-On that's meant specifically for smoothing out 3D prints.  It's called XTC-3D and is a gooey resin that's supposed to self-level after you brush it onto your prints.  Although it's a really nifty product that can turn a crummy print into a smooth, shiny print, I've had mixed results in my attempt to use it.  Most of my applications have at best still been kinda bumpy, at worst... well, again, the lumpy mess you see above.  Smooth-On recommends working at room temperature, which I'm willing to bet is causing my issues-- I'm working outside in the Arizona desert, where it's still getting up to 100F during the day while most people are enjoying cool autumn weather right now.  Since exposing a heat-curing plastic to heat makes it cure faster, my working time decreases faster than I can work when I try to finish a large batch of prints.  I'm pretty sure that's where I slipped up with Iteration Two.  I went back to do a second application and smooth things out a bit more after coating all six coasters, and by then the resin was starting to set up.  If I had left them alone, they probably would have just been a little bumpy like the time before.

On a related note, if you decide to use this product on your own prints, resist the urge to use cheap chip brushes.  Seriously, they've caused me so much grief that I kick myself every time I use one instead of throwing them out and getting foam brushes instead.  XTC-3D is a little tacky, and gets increasingly more tacky as it starts to cure.  Tacky plus cheap brush equals so many bristles getting pulled out and stuck to the resin!  Ahem.  Naturally, this causes you to either end up with bristles stuck in your print, or take the time to pull each one out and redo the section of resin you just screwed up by digging it out.  In my case, you're likely to have both occur and get really mad at yourself for being dumb.

At this point, I'm hoping sanding alone will suffice for Iteration Four.  However, gluing the two pieces together has left a small gap around the inner edge of the Forest Medallion prototype, which showed up as a thin rim sticking out of my new test mold.  I trimmed it as best I could, but my X-Acto knife didn't exactly slice the rubber too well, so it ended up looking sloppy.  If it shows up in the final cast and can't just be filled in with some sealant spray, I'll have to give XTC-3D another shot.  I'll be sure to post an update if that happens.  Until then, may the super glue be with you, not on you!


UPDATE 10 October: After the above-mentioned mold was used, I've discovered that whatever grit my wet/dry sanding block just doesn't cut it for mold-making-- I was left with obvious layer lines in the mold and test cast even though my fingertip couldn't feel them.  After a bit of Googling, I've found people recommend getting up to 1500+ grit, so I bought a variety pack of wet/dry sandpaper sheets.  My current plan is to do a coat of XTC-3D and follow it up with a good sanding from each grit of sandpaper I got-- if I still can't get a good mold, I'll be sure to come back and update this page again.

Ocarina of Time Coasters: Part One

Hello!  Hey, listen!

So I seem to have an odd obsession with coasters.  I'm not quite sure where it stemmed from, as I've never really made my own coasters before despite thinking about it since I moved into my first roommate-free apartment.  I had actually forgotten about my plans until I was browsing Thingiverse for gift ideas and came across these bad boys.  Yes, 3D printed video game themed coasters.  The nerd in me felt like she had just seen the face of God when I beheld their 8-bit glory.  However, they were not quite good enough for me.  To be fair, nothing really is-- mayhaps that's why I feel inclined to make things.  But I digress.  After starting to play Ocarina of Time again, I could not resist the thought of turning the Sage Medallions into coasters.  After a quick Google search to ensure no one else had already made them how I planned to make them, I dove right in.

This project is special to me for more than just my coaster craze.  It's combining a few different processes I've been wanting to try for years-- 3D printing, molding, and casting.  I've never been able to afford trying the latter two, as getting started requires a lot of materials and gear, so I'm super excited and terrified at the same time.  On top of that, I credit Ocarina of Time as the game that got me into the world of video games, so this should be fun!

Part One: Modeling

The first step, naturally, was to create the models my Printrbot Play would turn into piles of plastic.  After trying to get back into Blender and failing for the billionth time, I bit the proverbial bullet and shelled out some cash for a license to Maya LT.  Some people feel you can't beat free, but those people obviously weren't forced to get intimate with Maya very quickly for a college modeling course and found themselves unable to learn any other program afterwards.  You could make a Stockholm Syndrome joke if you'd like-- I won't stop you.

I decided my coasters should be approximately 3.5 inches wide and 0.5 inches tall, with the inner surface closer to 0.25 inches tall.  Maya, like almost all other civilized creatures, prefers metric units, so I worked on a scale of 25.4 mm equals 1 inch.  I essentially modeled it pretending the unit grid was in inches and then scaled the whole thing by 25.4 before exporting it to an .obj file for printing.  I included a picture of the actual medallion from the game as an Image Plane as a reference while working, and decided to keep it all low-poly to better suit the spirit of the N64 era.  (I might at some point make a video outlining the process I used for modeling these coasters, as it isn't trivial if you've never done something like this before-- if you'd like to see that, please let me know)

The first model took me ages-- I'm still pretty rusty!

Unfortunately, I only saved a couple work-in-progress shots, one of which is of the first iteration of the Forest Medallion.  After printing the above model out, I realized the inside was way too small to hold a normal-sized cup because I didn't bother measuring the inside.  I guess I'm really getting into the "rapid prototyping" spirit with this project, eh?  Iteration one also taught me the value of positioning your models properly to better suit the capabilities of your printer.  For example, my Play is great at Z-resolution, but is kinda sloppy on surfaces at a small angle from the print bed.  Since I printed this one laying flat, the precision of the Z-axis was wasted on the outer edge, leaving the inner detail rough and ugly due to imprecise surface printing.

So many splotches... why are they splotchy?

Iteration Two-- where I had actually printed out all six coasters-- had an unfortunate mishap that I will go into more detail about in the next post.  Well, it had several unfortunate mishaps in my ill-fated attempt to sand and smooth, but one of them was enough to actually ruin the coasters.  Hopefully you can see the issue in the above picture-- splotchy bits across every surface of every coaster!  Before I ruined them, these had a thinner outer ring than Iteration One, making them much better at actually holding cups.  Printing them standing up improved the details, but left a few of the edges a little rough.  Sanding helped with the outside, but it was pretty much impossible for me to reach any part of the inside.  This leads us to...

The inner part of the Water Medallion in center, back-up pieces to the side

Iteration Three, which is currently in the process of getting a much more thorough quality assurance pass than the previous two attempts.  For this round, I've decided to print each coaster in two parts-- one for the inner detail, another for the outer ring.  This means I can orient both to utilize the superior Z-resolution of my printer.  Above you can see the inner piece of the Water Medallion-- this piece will be stuffed into, and glued to, the outer ring.  I have again used the Forest Medallion as a test.  This time around I'm actually able to properly sand the inner piece.  Yay!  Combine that with a smaller print resolution, and I've already got a silky-smooth tester going.  There are a few small gaps between the two pieces, but I'm hoping a little super glue will fill them in instead of just sticking to my finger.  If not, I'm assuming it'll just turn into a small blemish in the mold that I can just trim out.  Hopefully.


Hopefully Iteration Three will be the last iteration, as I'm still trying to figure out what to do with all these failed attempts that are watching me with judgmental, non-existent eyes.  For the next post, I'll go over some of the post-processing techniques I've been trying out and how exactly I've messed up each one.  May the super glue be with you, not on you!

UPDATE 3 October: Iteration Three was not the final iteration, naturally.  I'm now up to Four, which is essentially the same method as Three except I've added a little "foot" to the back of the inner models so they won't topple over while printing if the supports start coming loose.  After the printing is done, I'm simply hacking off the little foot with a saw/ Dremel tool so I'll still have a flat surface for molding.  Yes, I came up with this after I checked on a print and found it getting dragged around  with the raft and supports still firmly attached to the print bed.  Live and learn, I guess.  I'm going to run out of filament soon!

UPDATE 10 October: Iteration Four was almost there.  The way I modeled the two pieces left a noticeable gap between them when attached, so Iteration Five added a little ledge to the underside of the outer ring on which the inner piece can rest, that way I can actually push the pieces together firmly when gluing to prevent gaps.  I'm going to try printing them lying down so I don't have to hack off that little foot anymore, which means it'll need more intensive finishing. We'll see how that goes.