3DShipWright

Masting the Confederacy Hey everyone, I’m happy to be back and I’m now working on the masts for Confederacy. I’ve a lot to talk about and solicit feedback on here, so I think it cleanest if I break this post into two parts: · The first part will cover methodology and some underlying decisions I’m making regarding this pivotal part of the build. · The second part will be a detail comparison between the two ships I’ve done and I hope to showcase the improvements I’m making over my 3D Brig Rose, since Rose also used masts taken directly from AOS Pandora (read on for clarification). Part 1: Methodology After extensive research and deliberation, I believe I’ve come up with a good pipeline that will yield realistic results on modelling Confederacy’s masting, yards, and later its rigging. My chosen approach will involve deliberately redoing some things later (a benefit of 3D builds), so in order to not confuse anyone reading this, please allow me to establish a few thoughts and assumptions upfront: 1. AVAILABILITY OF DOCUMENTS: Little (if any) documentation exists on confederacy's masting and rigging. (If you do have or know of any relevant masting plans, I’d be greatly obliged for links/access to them) 2. SCOPE: It is a given that many dependencies exist between the overall setup of the masts, yards, and rigging, but to draw a line in the sand here, let me say upfront that I am only concerned with masting at this point. 3. SOURCE MAT. AND RATIONAL: I want to develop each component from sources that are as complete and comprehensive as possible. Thus, I’ve chosen the AOS Pandora. While I’d arguably be better served were I to begin with plans from more size-comparable frigate, such as the AOS Essex or even online documentation from the Constitution, the precise dimensions of each component has proven very difficult to find. For example: a. Given diameters from step to the partners, then each subsequent qtr. segment diameter. b. Details down to 1/16 of an inch for the tressel trees, cross trees, chocks, hounds, bibs, bolsters… you get the idea 😊 c. The positioning between components, like where the foot/heeling from the topmast segments hang in relation to the tressel trees and caps of the main segments. 4. SOURCE MAT. AND RATIONAL (continued): The AOS pandora does a great job of everything listed in point #3 above. By comparison, the AOS Essex merely provides the maximum diameter and overall length of each segment in a data table in the beginning of the book, then only a scale drawing - with no dimensions - later on in the actual masts and spars section. 5. ADAPTATION: Having said that, do plan on adapting Pandora’s masts and spars to Confederacy in the following ways: a. ORIGIN AND RAKE: The origin of the steps and approximate rake of the masts is depicted on Confederacy’s framing and deck plans. To make some of the math easier I’m rounding to the nearest ¼ of a degree. b. PROPORTIONAL SCALING: The factor that most contributes to ‘realism’ is the overall proportions. Thus, if I can later establish any actual dimensions of Confederacy’s masts/spars, I should be able to extrapolate precise dimensions based on the proportions given to me by Pandora’s documentation… within reason, of course. (I won’t be making Confederacy’s masts 100 ft taller than Pandora’s just because it is 33% longer ship, lol) c. HARDWARE ADAPTATION: Finally, I will add, subtract, or swap out various hardware based upon belaying points and other differences in rigging configurations. Examples include: # of hoops and wooldings based upon total extrapolated length, angle (spread) of the bibs after the platform is tilted forward to compensate for the true rake of each mast, # number of thimbles to support the correct number of bowline blocks require for the adjusted sail sizing. Would a turn of the century ship still have used crows’ feet? Would the adjusted platform size require an extended mast cap and an additional support column… again, you get the idea. Smooth sailing, -Nate

Haha yeah, no worries... And to clarify something I said earlier, I'm currently using two textures: a bronze patina and a copper patina. Yes, I highl doubt that they were using pure bronze sculptures (not just cost, but also weight would become an issue). Now what we call bronze is an alloy, which I'm sure you know . The two most common alloys of the time were bronze and steel (original steel, not modern HCS). Steel has under-pinnings of iron; bronze has a copper foundation. That said, bronze is usually mixed with zinc, lead, or tin, all of which were beyond abundant in the Appalachians. And, as you alluded to earlier, a by product of this naturally occurring metallurgy is arsenic - highly toxic indeed! Anyway, while all that science/history is really cool - the actual modelling decisions I need to make are: Can I justify the pale green/ turquoise look of the sculptures? Can I justify the patinaed look on the sculptures? If not, what are my options? (remember this was an American ship, built at a time when we were definitely trying to distance ourselves from our English cousins...) Interested to know your thoughts. Thx as always, -Nate

Interesting... I'm curious, when Hunt writes "... (nothing)" does he mean he had no information to give Weir, or that there was no figurehead on the original HMS Surprise (HMS ROSE)? It could be interpreted either way. Aside - I'm going to stop referencing HMS Rose, and assume its well established that anytime we reference the Surprise, we mean the Canadian reconstruction of Rose, purchased by WB and renamed Surprise for the film.

Good to know Mark, thanks! I was also wondering about use of paint to simulate bronze patina? A notable example of this would be the HMS Surprise (HMS Rose)... The figurehead here is really what I'm shooting for, and there's a scene in Master and Commander where the crew is patching it up... in that scene, the underlying substance is clearly wood, so I'm wondering how they achieve the metallic look, complete with oxidation streaks... I would be curious to know your thoughts? Thanks!

@Martes - Hard to see from one screenshot, but if you were to look really close at the texture, I even went so far as to try to mimic a thin layer of bronze/copper atop a wood carving. I do believe the gilding - likely bronze due to copper shortages in the Americas during the revolution - to be historically accurate. The 'patina' material refers to the oxidation of these metals and occurs faster on thin layers. To accomplish this in 3d, I mixed the normal maps of the wood grain smoothly with that of the bronze sculpture material at about a 30-70 percent ratio. I then sharply painted the corrosion (patina) to the cervices (I tried using geometry nodes to calculate where the actual cervices are procedurally, but that proved to be serious overkill). The albedo/diffuse maps only mix the bronze and patina materials, understandably. As to the question of budget, yes there were funding issues during Confederacy's construction - lots of them! Yet Confederacy was built to impress; to be the unofficial 'flagship' of the American Navy at the time, and ironically, historical letters of correspondence will show that money was in fact wasted on her external beauty while neglecting important things like copper sheathing or even, you know, cannons. SCREENSHOT REFERENCE KEY: 1. Wood grain normal/bump map visible 2. Bronze normal/bump map visible (Note the chipping and denting effects) 3. Hand-painted patina effect. As an added bit of realism, the patina mapping also controls the metallic shader (Bronze is a ferrous metal, whereas rust and other environmental corrosions diminish these light-based properties)

Quick update: Starting to add ornamentation to the stern gallery, quarter galleries, and deck houses Color theme changed to a more 'classic' black and white theme, with patinaed (both copper and bronze) sculptures Rudder end cap completed Bolts finished on the gudgeon and pintle straps + much more not shown in this particular screenshot... addl updates to follow soon!

@Martes - moving our conversation to the build log as others may benefit from your advice as well. Okay, so I've done a quick and dirty fix to the 1st problem area you pointed out. A couple things: I do think for this to work, I will have to do the section immediately aft as well. also, I did my best to move the camera angle so the sun illuminates the top of each strake so we can see what's happening from top to bottom as you look down the hull (not sure it'll come out in the compressed photos though). Here's the before and after shots... 1. Before 2. After I am getting a slight 'warp' in the wale because, technically speaking, the additional smoothing would have to be perpendicular to the normals (i.e. normal tangent) and not just global XYZ to be correct. Still, it's encouraging that this may be easier to do than I initially thought. Thx, -Nate

Under the hood shot of the deck beam and framing. As it seems to be working well for this project, I once again modelled general shapes first, then gradually upped the detail and sophistication in subsequent passes and/or working sessions. About 50% done with the deck framing, at which point I'll move on to partitions, hanging knees (knee braces), and then the 'fixed' fittings.

This is the first of about a dozen relief sculptures that outline Confederacy's stern. I will use two organic characters I modelled from scratch for a 3D anatomy class I'm taking online - one male, one female - to get all the sculptures added to the stern. I will then use their armatures to not only pose them properly, but to change their facial features and body types subtly, so that hopefully in the final product they will look as if each one was carved individually, and so they don't look like clones of one another (even if in fact they are). Her hair is a different story. This will be the only time in this project that I use 3rd party models. I've decided to spend the $5-$10 bucks and buy a real-time hair pack from an online marketplace. Note: This one was very hastily posed, at this point I'm simply getting a feel for approx size as well as experimenting with how to get the cloth simulation to drape her skirt in a way that it looks like wood. Ironic, isn't it... cloth made to look like would that was sculpted to look like cloth? Anyway, here she is...

@Bruma - Thanks! Yeah, for a 'labor of love', this project certainly feels a lot more like a labor at times. And I'm always excited to meet fellow blender users

A Proper Rudder!

And we have a winner! Bear in mind this is a preview of the color theme only... The actual texturing will take months and will look much, much better To give you all an idea - I haven't even UV unwrapped everything yet. I just wanted a break from modelling. Still, not in bad shape for the 3 month mark.

Option B: A slightly more muted color scheme, visually similar to to that of the Continental Army's officer's uniform.

A preliminary color concept for confederacy.... A bit 'out there' but it conforms to all historical correspondence as to her appearance.

Good news is this is an easy issue to solve... Your problem is that your frames are way too thin. Even when using a uniform thickness due to the solidify modifier, the thickness of the frames should be, I'd guess, no less than 8-10 inches. Start with that and go from there. Best, -Nate

Good point! Along time ago I simply got in the habit of ensuring any and all mirror modifiers I use go to the very bottom of the stack... that way, you can mitigate potential issues.

Well done Sir! She's a beauty.

Keep in mind the boolean does leave behind a messy topology, but you can quickly clean it up by selecting all horizontal edges by selecting a couple from each part, using the select similar -> direction to get all 'intentional' horizontal edges, then press numpad 1 to select the vertices that comprise those edges, then ctrl + I to select any messy verts left over from the boolean, then finally delete -> dissolve vertices Hope that helps, -Nate

For the cant frames (or all frames for that matter), use a single solidified hull (modifier already applied), then use other objects as 'frame' cards and the boolean modifier (set to intersect) to get perfectly cut frames like so:

Excellent start... and yes, I extensively used all three fixes you listed above, I just didn't list them because the tutorial I made was already hard enough to follow, and I assumed anyone that could follow it would also be able to figure out the 'change direction' hack on the curve segments and how to enable the modifier views correctly in edit mode. On a side note, its truly irritating that Blender can't figure out that a series of curves that all start from z=0 on the world plane should be pointing the same direction, but oh well... I too own the AOS Pandora book, and in that book, note the complexity of the outer planking. in fact for my Confederacy build, I mimicked Pandora's hook-and-butt joints along the main wale, as well as how certain planks widen. to meet the gun ports in key spots. Anyway, great job - excited to follow your progress. -Nate

@Martes - Yep, agree with everything you just said, a couple things: The height of the copper plating isn't really something I can fix at this point sadly. I mean - I could of course, but it's unnecessary work given my final renders will have the ship in water. I hate to use that as an excuse, but technically speaking, Rose is already rife with problems, so I want to spend more of the remaining time on this project honing my lighting, compositing, and visual effects abilities. I have, however, sunk her down to the appropriate level in final renders. 2. The rings will be removed. When I previously said I saw these on historical plans, I now believe I was confusing modern-era dry-dock mounts with 18th century deck scuppers, which wouldn't be at that level on a small brig, if at all. 3. Finally, I will be replacing the union jack with a real Royal Navy flag from the time. The challenge is finding or color grading a texture to match the sail cloth material. Again, I can do this, but I want to finish up the rigging 1st. -Nate

Hey Everyone, I will be posting a major update to Rose shortly, hopefully one of the 2 or 3 final updates before I can stamp the word 'FINISHED' on the thread. That said, as I can't put Rose on a shelf in my bedroom or home office, I'd envisioned doing a series of epic renders with her, rife with special effects and postprocessing... which brings me to my question for the community: Would any 3D Artist out there be interested in doing a compilation collaboration - gotta love the alliteration - with me on a naval battle, perhaps; effectively a modern take on those classical maritime paintings I'd venture to say we all love? It will be a while, granted, but I wanted to put a feeler out there to see if anyone else would be interested... Thx, -Nate

@Martes - So I'm revisiting Rose a bit this morning... gotta say, thanks to yourself and others on this forum, my nautical/historical understanding has come a long way since I started this project. For example, you once asked me what the rectangular linen boxes were inside the hammock cranes as hammocks were packed into cylindrical containers... I now have the correct answer: The USS Brig Niagara, like the USS Constitution, is an operational to this day, but unlike the constitution, Niagara offers day trips to the public. Thus, she is subject to the laws of the USS Coast Guard. So, in an effort to preserve the historical feel, adult lifejackets are stored in the hammock cranes in sacks that mimic the linen of the time... A nice touch by the Pennsylvania Historical Society, if you ask me. I am still removing them from the final model, but just thought I'd share

@Maddog Shipyards - Excellent question re: UV unwrapping... So I'll give you a few different answers/things to consider. The whole 'low-poly' concept for real-time applications (i.e. Unity, Unreal, etc.) is widely misunderstood. Keeping a low poly count is mainly important for models used in particle systems like trees/vegitation, or not letting it get to a crazy-high number in models that are complex by nature such as people and animals. Just FYI - buildings, unless you're re-creating a gothic cathedral, should be easy to keep low. Having said that, there is an argument to be made that a game that takes place on the ocean would need all environment models (i.e. boats and ships) to be low-poly because the ocean itself has to be fairly complex geometry in order to look good. Luckily, there's a very convincing work-around: The trick is to layer 2 ocean modifiers on top of each other. The first is a large scale, low-poly plane that will act as the underlying 'waves' or 'swells'. The second is a small-scale, super high-poly ocean modifier that mimics the ripples. Both are baked into a displacement map, but only the large scale displacement is used to displace geometry while the high poly is simply a bump map. The results look like this, and surprisingly - the entire ocean has far less polys than the ship itself: The 'professional' 3D workflow used by major studios is modelling -> re-topology -> UV unwrapping/texturing -> 3D rigging (not to be confused with rigging on a ship) -> animation (or baked animations - which Blender calls 'actions' - for use in video games). Professional studios have teams of people whos full-time jobs it is to do a specific step in the above workflow. Ergo, as an individual artist, I actually find it easiest to use this rule of thumb: UV unwrapping/texturing should be done in a way that prevents duplicating work. Therefore, I will hold off at least until I'm satisfied that a particular portion is finished from a modelling perspective. However there are two scerarios where you want to texture earlier in the process. they are: Things I'm going to reuse multiple times. (Example: anchors, barrels, bollards, cannons, cleats, columns, deadeyes, grates, lanterns, staircases... you get the idea) Anything that requires an array modifier. (Example: chain links and rope segments... plus maybe a couple little things I'm missing) Hope this helps, -Nate

Thank you Maddog! So years ago, one of my 3D mentors once told me, "At the end of the day, there is no substitute for actual geometry." Weighing in at 6,076,131 vertices and 6,085,643 polys, Rose (a fictional brig - named after the original HMS Rose replica, titled HMS Surprise for the Master and Commander film) is proof of that idea. Even so, I think I would amend that to say, "There is no substitute for actual geometry... where it counts." As to figuring out precisely where that is, well, there in lies the rub, as they say. Best, -Nate