The PBY and Zero 3D Model Thread

[Gambit21]

Normally when I start a project I make a WIP (work in progress) thread on a 3D forum, but in this case I figured I'd post progress here instead since it's relevant - sort of.

This is not, and will not be a game model. While beginning in a low-res state, it will end up as a high res subdivision surface (the type of smoothing algorithm used) model

for the purposes of a high res profile illustration print for a local (an hour away from me) PBY Museum in Oak Harbor Washington.

PBY's were stationed here at NAS Whidbey during WWII, along with Wildcats, Hellcats and a few other birds. It's still an operational naval air base.

 

I'm fortunate enough to have access to 2 of the very few WWII restored PBY's in existence, both here in WA State. The closest one, and the one that belongs to the museum for which this job is for, is a work in progress and will never fly due to a cracked frame. The other is owned by my boss's brother in-law and is fully airworthy.

I have full "crawl around" access to both, and pretty sure I can arrange a ride in the flyable one.

The latter I found out about after I got the gig, and had no idea until I mentioned the project to him. What are the chances?

 

Previous to this I've produced several other museum quality renderings/prints based on my 3D models, including the P51-D and F6F-5

My goal with the P51-D was to produce the most accurate, detailed model and rendering of a Mustang to date - my goal with the Catalina is the same.

 

Anyway one or two of you might be interested in how a 3D aircraft model develops - so figured I'd post this here instead of a 3D forum this time.

Keep in mind a game mesh stays at a lower resolution than this mesh will end up as, since I'm not working within the same polygon limitations.

Otherwise development/workflow is the same.

One of the difficult things about making a game model is that detail needs to be represented with less geometry, no always easy.

A higher res model presents it's own challenges.

 

So here is the (almost) blank slate work space with reference placed, and a few nights worth of modeling.

Just establishing the overall shape right now. It's a looong process. Working evenings and some weekend hours the P51 took me 11 months to model, texture and finally render. 

I won't be quite as single-minded working on this model, so I have no idea how long it will take. Overall it seems to possess fewer "how the #@% am I going to do that?" contours.

In any case when finished it will be be a highly detailed, fully painted and rendered museum quality print.

 

(I also have a Zero in progress, but setting it aside temporarily while the Catalina is in development)

[Extreme_One]

Very interesting! I love the Catalina.

 

Do you only model the external contours or do you also model the internals?

[beepee]

Sweet!

 

I am looking forward to watching it progress Gambit.

 

Are you building it in Blender?

 

It looks like a steep learning curve,

 

We have this one down here in New Zealand:  http://www.nzcatalina.org.nz/

 

Edited October 20, 2016 by MF-beepee

[Heliopause]

Nice. In Holland there is also an flying example. http://www.catalina-pby.nl/

Couple of years ago I booked a flight (of 90 min) in it. Impressive.

[Gambit21]

E-one - Modeling just the exterior, and perhaps inside the wheel wells if there's to be a gear down version.

The Mustang was the first aircraft where I fully modeled the gear, inside of the gear bays and doors etc, but no cockpit since

it doesn't show in the render. That took a huge chunk if time.

Fortunately the Catalina is a quite a bit more simple in that regard if I decide to take the time with all that.

 

This particular Catalina will have the radar dome on top, just behind the cockpit. So far the only info/photo I have for a Catalina with interesting artwork that actually was

based at NAS Whidbey is "Radar Rachel" however I'm going to keep looking into that. I'd like to find recreate artwork that was a bit more classical.

 

 

beepee, I model and UV map in Modo then do the lighting and rendering in Cinema 4D.

Heliopause - nice!

Edited October 20, 2016 by Gambit21

[Heliopause]

Btw, I made a clip of the flight

 

https://vimeo.com/106056861

[Gambit21]

Blocking in the fuselage/hull moving rearward.

Still keeping the geometry as low res for as long as possible, only adding an edge loop when absolutely necessary, as adding resolution too early in 

the process just leads to headaches and more pushing of vertexes around, thus compromising the overall accuracy of the mesh.

I save iterations of the mesh as I progress in case I have to back-up and re-start from a certain point - which happens.

Sometimes I lose a night's work because I realize I've gone down the wrong path and would be better served to back out.

 

As work continues, the mesh will be subdivided as needed. In it's current state (absence of features notwithstanding) from a resolution standpoint is probably

85% or better of where a game model would be, at least for a large shape like this.

Depending on the subject and workflow, this subdividing process can mean a greater or lesser amount of geometry from one area to another, just depending on the detail requirements.

Note the difference between the cowl and fuselage of the Hellcat. The cowl was subdivided to allow me to cut the openings for the cowl flaps while still maintain

a smooth form. A game model would not allow me this luxury. Note the game engine does not utilize subdivision surfaces/smoothing, so geometry budget must be carefully

and thoughtfully allocated. (Where can you get away with less geometry/more faceting, where will this be least offensive visually, etc) For illustration purposes, I have a lot more freedom.

 

I have station profiles/cross sections (provided with the drawing) that help me get an accurate shape as I move along the fuselage, they correspond to the letters below the drawing of the fuselage.

These (and the rest of the drawing) must still be cross-referenced with photo's constantly. Good reference is half the battle.

It seems like with every model I end up finding one set of drawings that, without them, the accurate construction of the model would be impossible. 

For the Catalina it was this William Wylam drawing. Without it I'd be floundering as photographs just don't tell the whole story.

 

Note the density of the Mustang mesh - just because I can. 

I doubt the Catalina will be this dense, but areas of it might be, such as the engine cowls. 

 

 

[Lusekofte]

Btw, I made a clip of the flight

 

https://vimeo.com/106056861

 

Thank god you had some internal footage in that Video, remember next time to film all internals like inside the root of the wing base and all stations. In this forum it is inside that counts not the view outside. Good Video compared to flight with B 24 videos I have seen only containing a view over Texas :D 

[Gambit21]

It's the externals that count for my purposes and this illustration. Nonetheless - nice video HP.

[Heliopause]

I was not really thinking about the forum at the time ......

 

How'bout this one?

 

[Gambit21]

What a great experience -   hopefully I'll be able to get a ride in the WWII restored one.

At the very least I'll be able to crawl around inside, but not a priority at the moment since I don't need that internal detail right now.

[Gambit21]

Beginning to cut and re-route geometry for the turret and cockpit area - again, staying low res as long as possible, just concentrating on

the broad-brush shapes. As I get stuck, or bored, or intimidated by an area, I move to another area until I'm ready to tackle whatever geometry

problem is twisting my brain at the time. It's all about edge flow.

 

Smoothing turned on here on the fuselage just for grins, but not really ready for that yet so the lower hull get's too rounded with smoothing on until I add more geometry.

 

The cockpit geometry is rough, and just a placeholder so that I have a point of reference - making sure the fuselage profile is correct in that area.

 

[216th_Lucas_From_Hell]

Looking good!

[Gambit21]

Thanks Lucas

 

Did some work cutting, rerouting and extruding geometry for the tail last night. 

The station profiles I have that were part of the technical drawing you see in the background were not entirely accurate.

 

I visited the actual aircraft yesterday and was able to run my hands over the contour of the fuselage back there toward the tail, where it rounds from the belly

toward the dorsal surface- the radius is much tighter than the drawing indicates...like a basketball almost. 

Not that details like this will matter a ton in an isometric render, but it will affect how the metal catches the highlight from the light source, and I like to be

as accurate as possible...just because.   So adjustments will be made.

[216th_Peterla]

Nice work.

[Gambit21]

Rudder and stabilizer work tonight.

 Almost modeled the incorrect rudder/vertical stabilizer variant! Fortunately I noticed before I got too far along.

I added the appropriate edge loops to both areas to allow proper SDS/smoothing - the fuselage forward of this area

will now get some attention, including the canopy area.

 

 

[Gambit21]

Cockpit reference placeholder is gone, now the real canopy is beginning to take shape on the actual mesh.

Getting this area correct, contours, proportions... and getting it to behave properly when smoothed is probably going to be the greatest challenge on this model.

 

The drawing doesn't exactly jive with my photos here, (of course) so I'm going to have to pay the actual plane another visit and take some measurements, a luxury

I haven't had before when modeling an aircraft. (Well I had access to the Mustang, but it was a hassle) Getting to the PBY is easy.

 

The model is being created to real world scale in the 3D program, so if I measure the window and it's 13", this translates to 13" in the program...things are easier when 

working to scale.

 

 

[II/JG17_HerrMurf]

Is this model something the devs could use as a foundation for an in-game 3D external model when we go to the Pacific? Or is it mostly suited for artistic/static work?

[Gambit21]

Is this model something the devs could use as a foundation for an in-game 3D external model when we go to the Pacific? Or is it mostly suited for artistic/static work?

 

These devs keep things close to the vest with regard to aircraft development - even as far as external meshes up to this point. Who can blame them really,

Oleg had several problems with 3rd parties in the past. Those lessons don't get forgotten.

 

They have their own reasons for this too and I don't want to second guess them here on the forum. I imagine on multiple levels it's just easier for them.

To be able to look over someones shoulder and make changes, give suggestions, or send a part of the model to someone else to begin work on internals or LOD's

for different parts of the model. Not to mention they don't have to worry about someone on the other side of the planet flaking out on them.

Of course, that stuff mostly applies to a model that's in progress rather than a complete mesh. Still I'm not holding my breath that this mesh

will be seen in-game.

 

All that said - yes this mesh will easily convert into a game mesh with a little work...mostly removing or in some cases adding edge loops.

The main difference will probably be in the engine cowls, where I'll have much more geometry present to cut out panels etc, where on the game model

these would be drawn in out of necessity, or represented with very minimal geometry. Whereas I have the freedom to do what I did with the Trojan below.

Notice the difference in resolution between the cowl and fuselage. Both have smoothing activated.

Most of the Catalina base mesh, fuselage, wings...will be lower res. (see the smoothing example below)

 

In any case using re-topology tools over the high-res mesh to re-establish a low-res base mesh is also an option. (think drawing bigger, low res polys over the high res model)

It's a common workflow for converting high res sculpts out of Z-brush to game ready geometry.

 

I save iterations of the mesh as I go however, so I always have the lower res versions as back-ups.

 

So the short answer is yes, it could be used as a basis for the game model...could being the operative word.

 

I think the museum wants the Wildcat after (already did the Hellcat) then Corsair, then on to A6 Prowler...basically aircraft that have been or are based at 

NAS Whidbey. 

 

Below is an example of what the SDS (subdivision surfaces) smoothing does.

Same cylinder, smoothing on and off. The more detail you need in the mesh though, the more geometry you need.

My Zeke mesh above is an example of doing this with a game mesh, the Trojan/Mustang illustration meshes where resolution can go as high as I want.

[Gambit21]

Well I'm stuck on the PBY until I can get to the actual plane and take some measurements next week - so I pulled the A6m2-21 game model off mothballs and began

work converting it into a proper SDS mesh that can be smoothed for illustration and rendering purposes. Still a few details to add as well.

 

Rather than waiting on the Zeke, it looks like I'll be developing both of these models concurrently so I'll use this thread to log the progress of both aircraft.

 

If anyone has any close up shots of any part of the Zeke especially, feel free to post them here. I have tons of reference, but I'm always looking for more.

 

The Zeke has working/folding wingtips, flaps...all control surfaces, the canopy opens etc, although all that won't matter so much with the profile render.

A gear down "at rest" render may come later - who knows.

 

The engine in the Zeke you see in the isometric front view is the Wasp I modeled from the Hellcat - just threw it in there as a placeholder so I don't have to look at the empty space where the engine will be.

Edited November 18, 2016 by Gambit21

[hotshotmike1001]

Here is some pictures for your PBY project Gambit21

 

follow the Link for the FULL Res Pictures

 

https://www.facebook.com/162065864220449/photos/?tab=album&album_id=218151521945216

Edited November 20, 2016 by hotshotmike1001

[Gambit21]

Thanks again Mike.

[Gambit21]

Some work on the ballistics trenches and cowl flaps - edge flow modified to allow proper SDS smoothing and

edge control. Some other small details added that a lower res game mesh will not allow (would have to be painted on the texture map)

This mesh will allow me the freedom to cut these items into the actual geometry - the small hole near the wind screen

is a good example of this.

 

After a trip to the museum to take some measurements, back to the Cat tomorrow night most likely. 

[Mustang_45]

Beautiful work, Gambit21!

[Gambit21]

Thank you!

I was able to crawl around in her a bit today and found out a piece of interesting info.

I thought that "Radar Rachel" was simply an example of a PBY that served at this base...turns out this Catalina

IS in fact the actual Radar Rachel. That is both cool and disappointing. It's neat that I have access to the actual aircraft

pictured, however I was hoping to reproduce artwork that was a bit more "classical" in a pinup sense.

 

No reason though that I can't do an alternate paint job as well if the proper info surfaces.

[Jade_Monkey]

Pretty sweet renders.

[Gambit21]

Thanks Monkey

 

More work on the Zeke, re-doing the intakes as the previous ones were game resolution and just stuck on there.

Now they're extruded from the cowl and fuselage geometry.

 

If anyone has any good shots of underneath the cowl where that exhaust pipe is, I need more reference.

I'm thinking now that the plate just below the lowest cowl flap (the one with the exhaust is poking through) is actually attached to the cowl.

I need to verify this. I've scoured the net collecting photos already.

 

I have one locally that I may need to visit.

 

Having reference problems with the PBY, conflicting drawings/proportions so I'll keep working on the Zeke

until that issue is sorted.