tomoe gozen (fgo) cosplay project
22 Jun 2026 - digitalily
Tomoe is a character whose design I saw and went 'you are going to be the favorite character,' and things sort of went from there. With her 2nd ascension in particular, I appreciated that in terms of design, while it was definitely embellished as all things in Fate especially are, it felt much more restrained. It invokes 'samurai' very well, without leaning too far into (or really at all) fanservice, and is also visually striking in terms of the colors. It wasn't until some friends were discussing a group Fate cosplay that wheels started turning about potentially cosplaying her (and woops I missed it for the year we actually 'did' that), but once I started thinking about it things started flowing.
However it has been a while, and at this point, this is a cosplay I've wanted to do for a long time... and have attempted to start multiple times. The first time, I modeled and printed a fairly (in my opinion) good reconstruction of her naginata's blade and the decorative bit below it (painted and all), but I stalled out after that because I had originally wanted to use a bit of plastic plumbing pipe for the haft. The second time, I fully modeled the whole naginata, as well as the shoulder armor pieces, but get stuck when trying to deal with the chest armor, as this was where I had to break away from my model & 3D print & finish each piece routine, and started working with EVA foam for the first time, as well as having to pattern something for the first time. I ended up with two plastic cemented pieces for the front and the back, as well as some design embellishment on the front when I hit the brick wall that time around. Mostly, I couldn't figure out how to satisfactorily connect the pieces, how to have the belt which sits on top of the lower part attach properly, and how I was going to be able to support many of the other pieces with something that flimsy. Additionally I had never painted this material before and I was quite afraid of making a huge mess out of the pieces I had already but quite a bit of time and money into. In the end, work stalled overall with a minor resurgence to model and test out how the skirt armor plates would work. But now, armed with a fancier printer and using ASA instead of PLA for the first time, I'm going back at it with renewed vigor. Part of the problem with 3D printing is the finishing- the amount of sanding and filler primer layers required is great, and even then results are far from guaranteed, but ASA is a material which is supposedly much easier to sand than PLA, and also has the option for acetone vapor smoothing (something I don't expect to try, but does exist). The current aim is to do some test work on a piece of the armor to see how smooth I can get it feeling & looking, and then doing work to finish more as I go along, which I'm hoping will provide the thrust of motivation I need to get through it this time around.
Beyond that, I figure writing about it will also help, so in this, the first post for this project (and first item in the projects page list) I intend to go back a little and talk about the initial research and horrendous napkin math that led me to my initial models and some of the trials and tribulations experienced along the way! So without further ado...
I am not this person
One of the first places my mind went when addressing the task of designing a cosplay that requires a lot of bespoke components is to the question of proportion. To begin with, how on earth am I to know how big these components are? Specifically in terms of armor and weapons. But related to this is the fact that, as the title of this section may imply, different heights are involved, and that impacts everything. Tomoe's in-game profile lists her height at 162cm or ~5.3ft, a number which un/fortunately doesn't match my own, but at least with both our heights, a ratio emerges. The problem then is, there isn't really any documentation on character costume related dimensions or anything like that, and even the art like above doesn't exactly give the whole picture because 1) do horns count? (I am almost certain the answer to this is no) and 2) the image isn't even a true full body shot. With this situation before me, I could only really turn to one thing- the relevant in-battle sprite, pictured above...
With the aforementioned ratio in mind, I could use some dubiously accurate pixel based distance math to begin to understand how large these pieces should be in real life. Extremely suspect in terms of accuracy but really the best I could do without straight eyeballing it, but even that said there was a little bit of necessary eyeballing, due to distance and angling and all that. The first and least risky place to apply this to was to me pretty clearly the naginata. Since a good 80% of it's length is 'line,' it's about as non-complicated as it can be for using this method of calculating length. Even when the line gets a bit spicy by the blade and curves, it's not too difficult to get more than good enough estimates. Where things got trickier was specifically on what I call the hilt/blade guard (which is probably not the right terminology but it works for me)- the golden piece with the curves sticking out right below the blade. While getting dimensions for it wasn't the worst, getting the model to look accurate enough while being large enough to support the weight of the lengthy blade atop it was a challenge. This model was made ages ago, and I think with what I have learned in the interim, I could redesign it well enough, but I like how it has turned out, so it'll stay for now. Fortunately because the whole thing needs to be split into printer-sized parts, I could always change that piece out until glue is firmly in place.
I do not know Fusion
I then went on to begin looking at the armor plates on the upper arm. This is where the pixel math got a bit more dubious, and I had to rely more on visual appearance by printing out prototypes and seeing how they looked. This was also the first time I had to really fight against Fusion- specifically with the top curved up piece. Making a half potato chip shape by itself isn't too bad, but what was a nightmare was getting the rounded border piece to look the way I wanted it to when swept around the edge. What was a super nightmare was getting the curved strips to even remotely work. It turns out that, especially with the level of knowledge and lack of general training I have with Fusion or CAD in general, this was an extremely difficult task. While I could get to a point where I could more or less get the shape to bend in the right direction, I was met with endless issues trying to get the bodies of the main part and each strip to combine- frequently seeing the dreaded boolean error. As you can see from the mesh view in the images, it ends up being incredibly funky. I had to manually delete some sections of the mesh and do some weird surface merging, but I think the end result ended up looking alright, and when attached, the pieces function.
And attaching was a question I had to grapple with. Fortunately, after looking up examples of real world samurai armor, I found that the artist for Tomoe did a fantastic job at designing relatively realistic armor. As the abundance of holes in each piece may indicate, it's intended for some sort of cords/fiber/what have you to loop through and crisscross to connect each piece in a way that is ultimately pretty darn sturdy. You can see this in the art at the top, though the glimpses we get of the back side both there and in the sprite (which you can see more in certain attack animations) doesn't really make sense with this in mind, since you can't really see the cords from the reverse side. Being able to look at historical examples really helped out, since people you could see every angle and get a better idea of how these things were held together. There were also several blogs I found of people trying to make their own using historical techniques, which was another great resource. With this figured out and a satisfying plan for the shoulder armor, the next part was the skirt armor, and boy was that one much less obvious to figure out.
The actual plates were much simpler to design, since there wasn't a piece that curved in an extra dimension in a strange way. The weird part however was what appeared to be the connecting cords for these pieces, made very little sense in the original art. You can see similar crossing x shapes as in the shoulder plates, but above it you have way too many bits of cord to make sense- not all would be able to have a destination hole to connect to. Further, as you go down, each plate gets more x's. So there are a lot of extra cords, but then some actually need to be there? How the plates overlapped was also less clear at first, but I was able to make sense of it after thinking it through and experimenting- ultimately these pieces are much more tightly bound than the shoulder pieces. As you can see by the image, I'm not quite finished with these. I only just figured out how to best handle the curved strips that go in the middle space and am currently doing a painting test with the top piece. I still have one major bottom piece to completely design (though it shouldn't be too difficult), and I have to experiment further with the crest on the central bottom piece. Very unfortunately, the bottom piece is just a bit too large for my printer to do in one go, and my previous printer couldn't even do any of them in one go. The ways I came up with to link the pieces together on the old printer were not great, so I'll need to figure out something better for these. Unlike the naginata, these are much thinner (as I need 8 columns of these in the end...) and the weight is distributed differently, so I'm not quite able to use the same dowel + connector tricks.
... and that's it for the first post. The only major piece I've yet to discuss is the chest armor I mentioned in the beginning, but I think that can wait until I look at that part a bit more closely. For now, it's a lot of waiting on the printer, sanding, and more sanding. Below will be a few pictures of pieces printed on the old printer in an unfinished state. With the post now made as well, I'm hoping to soon have the projects sub-area built out and a hub page for this project created, as well as links to this and all future posts, and some form of image gallery, so look forward to it!
That's all for now- 'til next time!
