• New Horizons on Maelstrom
    Maelstrom New Horizons


    Visit our website www.piratehorizons.com to quickly find download links for the newest versions of our New Horizons mods Beyond New Horizons and Maelstrom New Horizons!

WIP HMS Bellona (1760 version)

I'd say that you have enough polygons ;)
Maybe spending 2 or three more on the bow and deleting every second vertical edge on the side? Seems a bit overkill there and you wont notice any difference
 
The only problem I found with reducing polygons is that the sheer may look odd if you remove vertical edges as there may be a noticeable straight edge in the curve especially in a large vessel such as this so care needs to be taken when making reductions.

Regards,
 
@Armada: I like the result very much! It looks like that the Nurbs produce a verry great result of edgeflow unter the sternpost. This is real great! The lofting options can with a few tricks result near the same situation but the edges of the sternpost are ever a complicated part, that looks very, very great solved! Im very surprised that you have setup the CV curves to result with Nurbs THIS. I have seen many techniques but really: Im interest so see some screenshots how you get this result. It is possible that you have found a way to bypass some hard parts at lofting (specially sternpost) when change a Nurbs shape with a special setup of curves to polys. And this at ?one? piece? Im real interest. Most coincidences gives the best ideas.

For Polys: I say in contrast: The area under the stern post need more polys, so that the round part (dont know the english word) looking more round. Easy job so far. The rest I would stay exactly how it is. The "saving" of polygones is a joke, compared to the rest of high poly modells at the ship! Think: A gun have at LOD0 (rendermesh) 30.000. With this count you can create 8 high till very high polycount hulls. I would never save too many polys at the hull. One loop here and one loop there gives most a joke reduction of ~1000 polys. NOT WORTH in my eyes! All easy to compensate woth LODs.

Very great work!
 
Last edited:
Marion - Thanks for those references! Based on that, and as I'll need to define the waterline on the mesh for coppering anyway, I'll probably keep the white stuff to below the waterline.
Personally I think it looks better that way, too.

SeaNorris. said:
Looking good,

At the hull around counter there may be a need for a few adjustments of the vertices as it seems to be indented (if you follow the top edge of the counter to the side you should see what I mean). There is also a small anomaly which could just need triangulating (counting the top edge of the counter as one, six lines down then follow to the side should locate it).

Also it may be a good idea to have the polygon smoothing group of the counter different from the rest so as to define it more.

Also out of curiosity what is the triangle count?
Click to expand...
I think I know what you're referring to, and I've fixed the indentation and geometry error accordingly. See below, with red lines as it used to be, and green as it is now.
Bellona_WIP20.jpg

I've also adjusted the smoothing at the bottom of the counter as you suggested, though it's a little hard to see here.

The tri count of the outer hull by itself is 8464 tris, excluding the keel and as-yet omitted inner hull.
For those that think this is too much: the headrails and transom seen in my earlier WIP pics, which will be added to the new hull soon, comprise nearly 40,000 tris alone.
Wedori is right: a high-poly hull really doesn't add much to the overall poly count once the rest of the model's detail has been added. :p


Wedori said:
@Armada: I like the result very much! It looks like that the Nurbs produce a verry great result of edgeflow unter the sternpost. This is real great! The lofting options can with a few tricks result near the same situation but the edges of the sternpost are ever a complicated part, that looks very, very great solved! Im very surprised that you have setup the CV curves to result with Nurbs THIS. I have seen many techniques but really: Im interest so see some screenshots how you get this result. It is possible that you have found a way to bypass some hard parts at lofting (specially sternpost) when change a Nurbs shape with a special setup of curves to polys. And this at ?one? piece? Im real interest. Most coincidences gives the best ideas.
Click to expand...
I actually used two NURBS surfaces: one for the majority of the hull (shown in green below), and one for the stern counter (shown in white).
I converted them to polys independently, and then merged the vertices where the two meshes meet. The high-poly mesh is faintly shown as blue, for reference.
To improve the edge flow by the stern post, I then modified two of the edges by moving them towards the stern and merging some of the vertices (see the red line and circles below).
Bellona_WIP19.jpg

I hope that explains it, but do ask if you want me to describe my method more clearly. :)
 
Armada said:
Marion - Thanks for those references! Based on that, and as I'll need to define the waterline on the mesh for coppering anyway, I'll probably keep the white stuff to below the waterline.
Personally I think it looks better that way, too.


I think I know what you're referring to, and I've fixed the indentation and geometry error accordingly. See below, with red lines as it used to be, and green as it is now.
Bellona_WIP20.jpg

I've also adjusted the smoothing at the bottom of the counter as you suggested, though it's a little hard to see here.

The tri count of the outer hull by itself is 8464 tris, excluding the keel and as-yet omitted inner hull.
For those that think this is too much: the headrails and transom seen in my earlier WIP pics, which will be added to the new hull soon, comprise nearly 40,000 tris alone.
Wedori is right: a high-poly hull really doesn't add much to the overall poly count once the rest of the model's detail has been added. :p



I actually used two NURBS surfaces: one for the majority of the hull (shown in green below), and one for the stern counter (shown in white).
I converted them to polys independently, and then merged the vertices where the two meshes meet. The high-poly mesh is faintly shown as blue, for reference.
To improve the edge flow by the stern post, I then modified two of the edges by moving them towards the stern and merging some of the vertices (see the red line and circles below).
Bellona_WIP19.jpg

I hope that explains it, but do ask if you want me to describe my method more clearly. :)
Click to expand...


My point of interest was the Curves and how you loft it around the sternpost. I have another way here but I have seen your edgeflow and think that you must use another way. Thank you for the pictures, it answeres in another way inderect how you have setup the hull. What I see explains me HOW you loft BUT this way is clear to me, Im much more surprised when I look at your curves. Here I found the secret :D The vertically ones for the timbering I know from close to every plan (Bellona Book Page 31) BUT HERE IT COMES: You have setup horizontal ones, that rise till the stern counter. I know from plans only plane horizontal curves (for example in the Bellona Book Pake 30/31 drawing A1/2) but I have never seen those horizontal ones, that rise in the third dimension till the stern counter. And thats the answere to my question. You have created the curves for the timbering and add your setup these special curves. And here my question comes: What was the base for the horizontal curves? I suspect that you have created Bellona A1/2 AND A2 Curves and created, based on these two shapes a new special one, who fits both intersections. Im very eager to hear your answere becauue the way you use can be a perfect supplement to the loft technique.

EDIT: So, perfect work so far, really!
 
Ah, I see what you mean. The trick here was, I didn't use the plans from the Anatomy of a Ship book, but instead used high-res ones from the National Maritime Museum that SeaNorris gave me.
These plans include planking lines on the front and back views. I placed curves along these lines, and then added extra curves between them to define the shape more accurately.
(I inverted the colour of the plans, to improve the contrast. Picked up that tip from Bava ages ago. ;))
Bellona_curves_plan.jpg
The rest, I pretty much described in this previous post:
Armada said:
I started by placing a curve along the lower edge of the main wale (red curve below) and duplicated and adjusted this curve towards the top and bottom of the hull:
Bellona_WIP15.jpg
The yellow crosses represent the Edit Points of each curve, which I aligned with the hull ribs on the plans.
Above the red curve, the others are mostly parallel except for two points on the stern (circled red below), where I bent the curves to touch two corners.
Bellona_WIP16.jpg
With the curves in place, I then lofted through them from top to bottom, creating a NURBS surface instead of a polygonal mesh.
This worked well overall, except three areas around the stern (circled red below), which understandably had problems following the hull's complex shape.
Bellona_WIP17.jpg
Luckily, with a few small tweaks to the surface's control points, I managed to straighten out these areas to get the correct shape.
Bellona_WIP18.jpg
Click to expand...
I hope that helps.
 
Thank you. This is a great tipp with the special plan curves! I didnt notice this before and need to take a look into my books if I find comparable ones. This technique have the big advantage, that the extrude below the stern post ist much better to loft. I will try (after all other work) to advance my rib-extrude-technique with a combination of both workarounds. Curves, based on the ribs, have a big advantage but I see when you use the special plan curves, you get a better flow at the back of the ship -> and lofted in one piece. My base loft, polygones techniques, have in my eyes the best edgeflow and when cutting the upper shape later, the hull is real great smooth BUT the disadvantage till today is to create the area below the sternpost with a second curves setup. Here I will try to combine with your idea and workaround.

You should write a WIKI for the technique. It is indeed a great useable workaround!
 
PROGRESS... :dance

The new hull has been properly shaped and textured for a while now, but I've only recently got back to the task of cutting the gunports.
I've devised a system using Booleans where I can generate the gunports and the lids in one go (albeit a very convoluted procedure).
It seems to work, but requires a lot of fixing mesh errors caused by the Booleans, hence it's taking a while.

I tested this on the lower gunports, as you can see below:
Bellona_WIP21.jpg

Bellona_WIP23.jpg


Bellona_WIP24.jpg

Here's a close up of one of the ports, showing the difference in thickness where the lid sits:
Bellona_WIP22.jpg


I'm gauging the difference in thickness by eye, based on the NMM wooden model.

I'll post some pictures of the lids once I've sorted out their pivots, but that's less of a priority until I've got the rest of the ports cut.
My current plan is to finish the gunports and decks so the model can be exported to Unity for testing purposes fairly soon.
I'm doing this quite roughly and skipping some details for now, so apologies if the perfectionists out there see some glaring omissions. :wp
 
Looking damn good! The gunports seem quite precise, you'll have to teach the rest of us the method once you've finished it up:napoleon its almost always a convoluted process with the interior ledge, and it's so easy to get it slightly messed up and then have to do lots of tedious adjusting.:modding Sounds good about the testing! I know Captain Murphy might be in need of something larger than the St. Albans, and I should think this would qualify. :p
 
Before I go any further, I'm downloading Maya 2015. Supposedly it has better Boolean algorithms, among other enhancements, so I'm hoping that will reduce the amount of cleaning up I need to do after performing Booleans on the hull. The cleaning up phase is what's making me hesitant to recommend this method for now, because it's so time consuming.

Briefly, this is the method I'm using to cut the gunports:
  1. Create cuboids covering each gunport, with plenty of length either side of the inner and outer hull.
  2. Merge cuboids into one object, and create TWO copies (cuttercopy1 and 2)
  3. On cuttercopy1, reduce the size of the cuboids to 90% of the originals.
  4. Perform Boolean intersection between a copy of the hull (hullcopy1) and cuttercopy2 to form intersection.
  5. Clean up the excess edges and vertices on intersection.
  6. On the intersection faces spanning the hull thickness, add a vertical edge loop approximately 15% of the length from the outer hull faces.
  7. Remove faces between this new edge loop and the inner hull faces, and close the remaining mesh.
  8. On cuttercopy1, create an edge loop on each cuboid, ensuring it sits inside the intersection mesh.
  9. Remove faces between this edge loop and the outer hull side.
  10. Perform Boolean union between intersection and cuttercopy1 to form newcutter.
  11. Create a copy of newcutter named lids.
  12. Perform Boolean difference between hullcopy2 and newcutter to form newhull.
  13. Clean up excess edges and vertices on newhull.
  14. Use the same technique as step 6 to reduce the thickness of lids.
All of this is done on half a hull, and with textures applied. I realise it's quite vague, so I'll explain with images later.
 
Well explained, but If you could take some screens while you cut the upper ones that would certainly help :yes Also how are you getting maya 2015 on a student license? I looked into it because I only have 2013(which won't open 2014 files) and it seemed an upgrade was impossible.
 
I'm just downloading Maya 2015 from the Education Community as normal. It won't replace my previous install, but will install alongside it.
As it happens, my college is still using Maya 2014 (and other Autodesk software), so I'll probably need to keep the earlier version installed anyway.

The upgrade should be worth it, even going by the new modelling stuff alone.
The UV editor looks like it's received a much-needed update... :type1
 
Unfortunately, the new Boolean algorithm is still leaving the same excess edges and vertices as before, so while it's improved for many cases, this isn't one of them.
This is probably down to the complexity of the hull mesh, so I'll have to live with it.

Anyway, I'm working on the rest of the gunports now, as well as plugging a few gaps in the transom and fine tuning here and there.
I'm hoping to get as much done as possible before next week, which is when the real work begins for my degree studies. :read
 
Phew! Seventy-four gunports and a couple of enhancements later...

Bellona_WIP25.jpg


With an improved (but not finished) transom:

Bellona_WIP26.jpg


Bellona_WIP27.jpg
If you see small grey sections clipping through the model, those are parts of the decks. They're still shaped to fit the old hull, so they need a bit of fixing. And the lower gundeck needs to be added.
And what you can't see is the messy geometry around the new gunports, so that still needs to be addressed.
Aside from that, all I need to do is create the physical proxies and the buoyancy mesh, and then I'll start looking at getting her into Unity.

I did take some screenshots while I was cutting the gunports this time, but I'll post them with explanations... when I get the time.
 
So I should be seeing the completed model sometime this week then? :razz
 
Latest progress:

I've now split the hull along the waterline so the 'white stuff' is only on the underwater side.
I also slightly recoloured part of the headrails, and moved them to rest against the new hull (they were still offset because of the old hull's shape).

Bellona_WIP28.jpg


All stern window frames have been replaced with thinner, lower-poly versions and now have glass panes.
The old frames were too thick and had an unnecessary taper in the centre.

Bellona_WIP30.jpg


I've improved the screen bulkhead too, adding more detail and two doors leading to the balcony.
You can see the doors on the far left and right sides, with the 2x3 windows. They're currently built to approximately the right width, but they might need to be expanded for gameplay purposes.

Bellona_WIP29.jpg

Finally, all decks down to the main gundeck are now installed. I haven't included all the bulkheads yet, but we probably won't need them for testing purposes.

Bellona_WIP31.jpg

The main things I need to add now are the companionways, to allow movement between decks, and the rudder.
Aside from that, there shouldn't be any other modelling required before I prepare the ship for export.
 
You must log in or register to reply here.
Back
Top