Lockheed Constellation (new developer)

[dzignguy]

Hi,
I am new to this forum. I am a retired architect and have started doing a full scale 3D digital model of the L-049 -L-749 Constellations. I have been an AutoCAD user for many years in my profession and have recently moved over to MAX for more dynamic modeling and visualization presentations.

I am also an old pilot that flew for many years and had to hang up his wings due to a health condition. As a kid the one airplane that fascinated me most was the Lockheed Constellation. I had the fortune to travel with my grandparents several time to Europe on a KLH Connie and a TWA Super G. My Uncle was a pilot for TWA and flew the New York to London route first in L-749's and later in a Starliner. So my first aircraft development will be the early Constellation.

I have actually modeled most of the exterior of a L-049 Constellation with animateable landing gear with steerable nose gear, control surfaces, flaps and gear doors. Im just starting to model the engine and starting the texture work. Any way to make a long story short I want to take this model and some subsequent models into MSFS and possibly X-Plane. I can model but I have no idea what is needed to get this thing ready for FS2002-2004.

I understand UVW mapping and have done texturing for architectural visualiztion but have no idea how FS2002-2004 maps. Any pointers here will be helpful to say the least. I am in the process of taking my High poly model into GMAX and getting it more consistant with a low poly model. Can someone tell me what the limits for poly and texture sizes are in FS? I'm shooting for under 4000 polygons for the model but dont know how easy this is going to be since my current model has in excess of 20,000.



Anyway, I will be glad to share my files with anyone that shows true interest and qualification for this project. This will be a freeware aircraft when done. I already have the high poly model in GMAX and am about two days away from having a low poly model ready. Any help or advice will be appreciated.

John

BTW I plan on texturing at least 5 liveries for my early model Constellations and will eventually do the Super and Starliners as well.

[Firestriker]

John,

If your model is 20,000 polys you well within any limits for FS2002. Some like the Meljets are approaching the 100,000 mark. Real frame rate killers on low and medium range PCs.

If you are using 3dsMax all you need to get your model in to FS2002 is the Panda X file exporter. Then you can compile to a *.mdl (flyable vehicle) or *.bgl (scenery object) using the Makemdl.exe compiler from the SDK.

Texturing is not to bad... start small and work your way up. The main thing you have to remember is that textures need to be square and divisible by 4 for the most part... 4x4, 16x16, 32x32, 64x64, 128x128, 256x256, 512x512, 1024x1024. Also keep them simple at first, by that I mean 8bit, 256 color bmp format. As you become more accustomed to working with them you can move up to the more complex types of files and will learn what is used where and how they function in the sim. Just remember, you CAN'T use 24bit bmp files in the sim.

Lou

[dzignguy]

Lou,
Thanks for the good info. I went and downloaded the Panda X exporter and it seems to do exactly as advertised. Nice stuff!

I have been doing architectural visualiztion for a number of years so I have a very good grasp of UVW mapping and texturing but was kind of unsure of how MSFS handled this issue. Thanks for your explanation!

I also am still a little unclear on smoothing and how MSFS handles that. I have a number of options in MAX as to how I handle smoothing. However I'm not clear as to how those options will effect the way the model is rendered in the 2002-4 versions of MSFS.

I slightly understated the size of my detail model. 600,000 polys and climbing, as I dont have all exterior features done yet. The engine model could be that much and more before I finish. Additionally, I havent even started on the interior yet. On the good side, the low poly edition is now down to around 34000 and there is more room to reduce further. I will use maps instead of polygons to show some less important features like wheels, hinges, vents, scoops, doors, links, nuts bolts, etc and landing gear detail. A reduction of another 20% or more.

BTW, the one advantage I see that Gmax has is the ability to colaborate with those that dont have MAX. MAX 5.1 is far more advanced but Gmax is a great tool for the flight simulation groups. The quality of models available since it's inception has grown measurably. And to those that think the Gmax route is cumbersome.........Just try and do a great model in X-Plane's Plane Maker and one knows that theres got to be a better way...........That system needs a *.max or *.dxf import filter BAD!
You can sure tell there are a lot of left brained people in charge of that system!

Thanks again for the pointer to Panda!

John

[xfeed]

Sounds interesting! What I would recommend, since you have Max5.1 with all the modelling/texturing tools, is to use texture baking. Take your high poly model, nicely unwrap it and render to texture using Max's advanced lighting (or radiosity if you like). Then you have a superb texture which you can apply to your low poly model. Using this technique I would estimate your polygon count to be as low as 25000 or so and the Connie would still look great.
Check for following keywords in Max's online help: "texture baking", "render to texture", "advanced lighting", "skylight", "unwrap uvw".
That should get you going. ;)

[dzignguy]

Firestriker:
Great project! I like that bus! At least you can get yours to appear in MSFS. I'm not even close yet! Oh well thanks for the encouragement I will keep on pluggin till it goes. My current model is going to have to stay Hi detail and I am building a new Flight Sim model more in the Greg Pepper Style of work. I have flown several of his GMAX aircraft and they all work very well so if I get anywhere close I'll be happy. A couple of more days and I'll have it ready to start plugging in a mdl file.

XFeed:
Very good idea, I'll give it a try. We actually tried to load the low poly model last night and it is throwing so many errors that I think there is a problem with the way I constructed it. I really want to learn all of this process from front to back so I wont gripe about having to do a new low poly model. I'm familiar with texture baking and how to unwrap UVW so that isnt a problem for me. I just dont have a good knowlege of how MSFS does the mesh smooth if any at all and if it will use my smoothing groups as it is in the file. This information is real important to me as it dertermines what strategy I use to make the model.

I'm a real fan of the Sub division (SUB-D) method of modeling but I have a real suspision that this is not useful in MSFS. If I'm wrong please correct me. None of the GMAX models I've seen, that were done for MSFS, are compilations of mesh objects assembled into an aircraft. Not bad, just a little more crude than I expected..........In any case thanks for the reply and useful pointers.

[xfeed]

Well, the SUB-D helps insofar, as if you stick to the rules for Sub-D (only Quads), your model will look just fine in COF.
I just went through the (somewhat painful ;)) process of modelling in Max5 for COF. Usually the errors (I believe you refer to the errors when using MakeMdl) fall in two categories:
1. fatal: a) when two vertices are colocated (weld before you export)
b) when you applied scaling to an object (use "affect object only" and then "reset scaling")
2. non-fatal: The faces look flipped in COF. This happens when you mirrored an object. Use "Reset XForm" and then collapse to editable Poly.
That's all there is to it. After following the rules to avoid these two problems everything should turn out fine. ;)
Some more ideas:
Smoothing groups work fine. Multi-/Subobject materials work fine. Composite materials don't, like everything that is dependent on Alpha-Maps (except for the props). So better solve windows by geometry and assign another mat-ID to them.

[dzignguy]

Martin,

Great answers! Ok, first, pardon my ignorance but COF? Anyway, Most of the errors were solvable. However, at the end I kept getting fatal errors and I can only say that they were probably due to scaling during part of my original construction of the fuselage and wing areas. Another problem was the mirror porblem if you dont use the modifier from the drop down list instead of the tool bar mirror. Didnt remember that one until Tom Gibson brought it up via Email.

I know I can do the Sub-D fairly accurately except that it's implamentation in MAX is a little quirky and you end up doing square corners in a way that requires more polygons if your just going to use smoothing groups and not go for the nurms smoothing in Mesh Smooth. Since that doesnt appear to be part of FS8 or 9 it seems to be easier to go the other route. I just finished a fuselage and wing assembly by going the simpler route and I cut polys by 60%. If I'm wrong please get me squared away...... :) Thanks again! This disccourse is very helpful!

john

[xfeed]

COF = a Century Of Flight
If you check the error log of MakeMdl it pretty accurately describes the problems: if you see something like "colocated" you need to weld vertices, if it complains about "has scaling matrix without animation keys..." you need to reset the scaling.
I can't follow what you mean by "square corners" that require "more polygons". What I meant is, that you don't use Sub-D for the models, just ensure that your basemesh follows the same modelling principles as if you would lateron apply an iteration or two of MeshSmooth over it and it would still look ok. You are right, Sub-D isn't supported by FS9; your models appear as in the Max viewports.
Regarding polygon-count: possibly I'm a bit picky here, but as far as I can see a lot of models use too many of them. Especially with the superb texturing tools of Max there often is no need to solve minor details by geometry. I would consider anything above 10K as high-poly and unnecessarily detailled for in-game application. ;)

[dzignguy]

Ok, let me see if I can explain this. When I model with the anticipation of using one or two iterations of Nurms Mesh Smooth via the Mesh Smooth modifier there are several things that need to be done to keep extruded holes, such as gear bays, etc. from having rounded corners. I realize that there are other strategies to model these features but it you want one, homogeneous part, there are hoops you have to jump through, mesh wise, to get square corners to look square. This is if they are "mesh smoothed" via the pull down modifier. I'm sure from your responses that you understand this reality....

Anyway I have a picture of the first trial load of the bard fuselage/wing/empenage assembly in FS8. I have no idea what's going on but I have an idea that the symetry modifier and some of the techniques I used to extrude poly's in this model are causing this problem. What do you think? At least I know my Connie is going to fly someday soon!

http://webpages.charter.net/dzignguy/images/connie2.jpg

Best Regards,
John

[dzignguy]

Martin,

Great Bonanza! Excellent strategy on your poly placement. What's the total on that beauty? I have learned so much from other modelers wire frames. They are just like a puzzle but if you study any of them enough, you can pick up a little hint here and there and eventually you get better and better. Your Bonanza has some really nice ideas for me. Thanks for the display! I love to model and for years I kept to just solid modeling architectural structures but now that I've ventured out into this organic, and mehanical world things are so much fun again!

Ok, I have a problem with doing what you describe with that particular model. It is a revised edition of a much higher poly, detailed model done over the last couple of months and this model's stack was colapsed before moving into Gmax. Fortunately, I have many iterations of the model during the building stages, before the collapse but they are still way tooooooo big. 400,000 polygons without the meshsmooth. This of course includes all the Windows, control surfaces, landing gear, wheels, doors, cowls, cowl flaps, gear bays, doors and many more features. Taking this back to a low poly model is not feasable because the strategies I used here will cause me more time to rework than I can build a new MSFS model. The other advantage is that I learn more and more each time I do a rework. Sometimes I get bored and then the project is set aside. However, the Connie is a very dear project for me and I'm not even close to that point yet.

I have tried turning the normals on faces and and rebuilding other areas that were wierd. The image above is one that I had worked on for hours before I even got that to display. I started with just a few polys here and there so that I couldnt even recognize what the heck it was.

I am almost back to the full fuselage, wing and empenage assembly with only 3000 polygons and I will start testing it today to see If I have it figured out. I will post what I get then. Thanks so much for the suggestions, it helps alot.

John

[xfeed]

I move us back to the left before we are squeezed into a single column... :P
Poly count on the Bonaza so far is about 2500, but at a single time some hundred less are displayed in MSFS (e.g. there are three variations of the prop depending on speed, wheelwell only shows when the geardoors are open etc.).
You're doing the right thing by starting from scratch - it would be nearly impossible to reduce a high-poly-model to one with a MSFS-friendly count.
What I would suggest, to get a feel for the workflow when dealing with polygon flippings, is to start with a simple box, clone-mirror it and export both to MSFS to see how they look (use different colors on them!). Then apply the Reset XForm (it is found in the Utilities Tab, not Modifier Tab) to the cloned box and you will see that it appears inside-out in the Max-viewport as well. Convert it to an editable poly and flip the faulty polygons and export again. Both boxes should now appear flawless.
Looking forward to an update of the "best three-engine aircraft" (they had a reputation for failing one engine on many flights). ;)

[dzignguy]

Martin,

Another great answer! I actually thought of doing this yesterday, testing the simpler form and strategy, but didnt know if it was feasable. I will do that this afternoon, I have my anual physical this AM so I will not be on the work station till then.

Yep, that three engined phrase was very common among Connie pilots. I flew on Connies many times as a boy and young man, what a great ride! My uncle, Capt. Jack Martin, had 15,000+ hours in various versions of the Constellation and aside from the early problems with the 3350's had the opinion that it was the best powerplant to ever power a passenger prop-liner. I flew on Uncle Jack's New York to London route a couple of times and on the first, which was in one, 1952, of the early livery TWA 49's we lost an engine about 2/3 of the way across the Atlantic. Quite an experience but of no consequence as we arrived 5 minutes late to the gate. Boy, the days of Wine And Roses!

Ok, I will try your test today. Is there anything I have to add to the directory in MSFS2002/aircraft to make this box show up? Is it just a matter of compiling the box into mdl and placing it with a config file in a special directory? Sorry about the stupid question, but I'm really green when it comes to working with the mdl output and how to set up in FS8. Anyway, thanks again Martin, youve been a great help! Whens that Bonanza going to fly?

John

[xfeed]

Easiest way would be the following: Make a copy of any complete aircraft folder as a backup (choose an aircraft you're not going to use the next few days anyway). Then replace the .mdl-file in the original folder with your new .mdl. It will appear in MSFS under the original name (e.g. Beech Baron 58) with the same cockpit, flying characteristics, sounds etc. as the Baron.
The Bonanza is flying alright. I was lucky enough to keep all the technical information from flightschool (flew it back in 1989 in Phoenix, Az.), so the characteristics are pretty accurate (at least as far as I remember them). I just joined a local club that owns a F33, and as soon as I get some reference photos I will continue to texture it. Right now I'm trying to work out how to create the gauges. ;)

[dzignguy]

Sorry for the delay in getting back to your last reply. I was off on a trip that was supposed to be a photo trip for existing Connies but turned into a family emergency trip to Texas. SIH... maybe I get to go photograph Connies in October.....

Anyway, I have redone the connie exterior into a Low Poly model and got it into FS8 without gear and doors and it works! There are a couple of areas that need polygons flipped but overall 99% is visible just as I designed. One problem, the model is as big as the airport and partially burried into the ground. I reset scale and system units and still no joy. Now I cant get anything to display. Oh Well I will walk away for an hour or two and I'm sure it will come to me.....

You mentioned in you last post something about not having your gear bays render until the gear doors open. What is that all about, I think I like that idea. I'm working on textures today so I will try to get this thing textured for an intitial showing tomorrow.

[Firestriker]

Are your System Units set to 1 modeling unit = 1 meter? If not, this is what usually causes objects to be way out of scale inside the sim.

Lou

[dzignguy]

Yes, I have 1 unit = 1.0 meters in the system configuration panel and I have tried everything encluding opening a gmax file that used to show properly in FS8. It also now shows up as a giant aircraft that takes up the whole airport....... Ive gone through all kinds of kinks to get this model to appear properly, including the fact that it was flying backwards initially.......:) This is really FUN!

[dzignguy]

Ok, here's another little problem. The ailerons and elevators operate properly but the Rudders are rotating around the wrong axis.

http://webpages.charter.net/dzignguy/images/MSFS%20connie-3.jpg

Real Cute! Unique! Anyway I tried rotating the Pivot point in all three directions but still this thing is whacky. Ideas? This process is really working my over. My IQ gets lower with each new task...:)

Here's another view of the basic airframe above San Luis Obispo, CA
I know, I know, all you verterans are chuckling internally at the newbies new toys. Hey, I'm having fun!

http://webpages.charter.net/dzignguy/images/MSFS%20connie-5.jpg

This Connie will fly and in TWA colors soon!

[xfeed]

Looks great so far!
Scaling: Should work fine if you scale first and then, with "Affect object only" activated, reset scaling again.
Axis of rotation: I could adjust the rotating axis by simply rotating the pivot to the correct orientation (I'm currently not at my computer, but I believe it is the z-axis, around which the parts rotate). I'll check again when I'm home.
Wheel-wells: Check the Makemdl SDK for the proper animation tags - if a part has a name like (again, I'm not completely sure about the name) l_wheel_well_xxx FS only shows it when the gear-doors open. ;)

[dzignguy]

Well I figured the rudder thing out. I guess I had linked all three rudders to the center previously when experimenting with animation. However once I got the rudders unlinked they still rotated in the wrong direction. After changing the pivot to the X axis things worked properly. Hmmmm odd that they would use the X axis for that......

xfeed, another question.

Does Fs8-9 use the bump channel in materials? If yes, how do you name those associated bump images? If no, then how does FS8-9 display apparant depth without modeling those features?

Perhaps when I get this first model done I would like to get several people together to write a overall primer on MSFS project development. There are tons of tutorials all over the place but no single, deffinative, document. The modeling stuff was easy for me but all the requirements of nomenclature /hierachy and tags is really hard to come by...... I've had to go back and redo so much because I didnt understand the way things had to be in order for them to work in the SIM. Anyway just a thought

[xfeed]

Yeah, thought about the Tut myself already - Good idea!
Bump Mapping is not supported. As far as I can tell from the stock aircraft textures, they are using Texture Baking (Render to texture) as I mentioned earlier. This leads to somewhat inaccurate highlight/shadows-areas depending on the time of day, but looks good enough in FS. If you use Advanced Lighting or Radiosity in your high-quality Max-aircraft (from which you bake the textures), your textures will be as good as those on expensive add-ons. ;)

[dzignguy]

Well, I'm way behind on this Connie but I am making some progress. Sometimes my real life gets in the way of the good stuff. Any way, I now have a problem with my gear, doors and cowl flaps. I did finally get both sections of the trailing edge flaps to work properly but the doors on all the gear except the rear two on the C_gear dont seem to be opening. I know they work as the key frames in Max are correct. At least as best as I know.

Also, why is the left gear not extending, the right works fine, without the doors? I know the nomenclature is correct. I'm including a couple of screen captures within FS8. At least it flies OK and with an updated air file it should be pretty close to original. I have an Uncle that has at least 15,000 hours in various versions of these babies and he is working with me to get it right. However I need help with this model. I'm about to finish baking the textures for the original TWA Connie and will start on the Western and PAA next. All and any help will be appreciated!

http://webpages.charter.net/dzignguy/images/msfs%20Connie%20gear%20down_4.jpg


http://webpages.charter.net/dzignguy/images/msfs%20Connie%20gear%20down_2.jpg

I will have it with textures in the next day or two.

[argueman99]

That Connie looks great!! My favorite propliner for sure.

Really looking forward to this!

[dzignguy]

Thanks, it wont be long until I have an update posted. All of the exterior issues have been solved and I am now doing the VC. The debut of this Connie will be on Tom Gibson's, California Classics web page.

Thanks again,

john

[southerndandy12]

dzignguy-
Your Connie looks really sweet! The connie has always been my favorite propliner, probably cause thats what my dad came over from Germany on, I beleive it was a TWA Super Constellation.

[dzignguy]

southerndandy12,

Thanks, I'm a little behind on this project due to some personal health issues. Fortunately that is now subsiding and I am back at work on this Connie. I now have all the exterior issues pretty will worked out. I am currently concentrating on finishing the textures and the VC / interior.

I have also been working on upgrading the flight model and air file. Thus as an old pilot I am having a lot of fun flying this creation. I have the help of a couple of people who have lots of hours in various versions of Constellations. My Uncle Jack Flew Connies for TWA from 1948 - 1958 before moving over to American Airlines and Jets. Most of his Connie time was in International flights between the US and Europe. Who knows He may have been the pilot of that Connie that brought your Dad here......... He personally liked the early birds as they were very much on the cutting edge of technology for the day. He says the Super Connies later became more like big flying busses. However, he says that the big jet carried that idea to the limits.........

Anyway I'm having big time fun with this project and will do many more in the future as I have found something that I truly enjoy!

Here's A couple of updates. The first is the bird in full flaps and flared for landing. I'm glad you cant see how much runway is left here as I cut this landing a little "tight".

http://webpages.charter.net/dzignguy/images/msfs%20Connie%20_25.jpg

This is a take-off shot that shows the gear fold sequence. The downlock struts fold outward instead of inward. I originally had them folding inward and my mentor pointed out that they should fold to the outside. I had tons of pictures showing this but never picked it up............ Cant see the forest due to the trees.....Please forgive the Fuselage texture. This is just a test of the Cylindrical UVW map applied to see how well it works. The next update will be much closer to the final version.

http://webpages.charter.net/dzignguy/images/msfs%20Connie%20_28.jpg

This bird is really easy to fly once you get the feel for the Connies dynamics. She has a nice slow sink with full flaps and very little power on approach. Once you have the timing down and get her lined up with the runway she just floats down to the runway in a very controlled asscent. Of course a Cross wind and turbulance will add some other issues into the equation but still a very easy aircraft to control... Uncle Jack says that's what he loved about the L049-L749 Connies. They were very tame and controllable on landing and take-off but flew very much like fighters once in the air. Remember that the P-38's top speed was about the same as the Connie..........The Connies wing was taken from the P-38 so you can understand why she was such a radical upgrade to passenger aircraft.....

[southerndandy12]

happy that your feeling better, the pics look great so far.

[dzignguy]

Hi All,

I just wanted to revive this thread for one last time as a way of thanking all of those that helped me get started with my Dream project. I recieved so much help from the people here and all of it aided in my continuation of the ultimate goal.

The Lockheed L049A, N90831 TWA, Star of Switzerland, is now being redied for release next week to the public as fee ware. All of this is due to the generosity of spirit here in this forum.

Thanks,

Best Regards,

John Howard White
dzignguy

Here are a few release photos of the "Star" that was started through the help on this forum.

http://fsdzigns.com/images/silver%20bird_20b.jpg

http://fsdzigns.com/images/silver%20bird_04a.jpg

http://fsdzigns.com/images/silver%20bird_07a.jpg

[n4gix]

John, that is a stunning model! I can't wait to get that bird and fly it! :)

Although I build biz jets for Eaglesoft, my heart is still in light GA and vintage big birds... :)

[dzignguy]

Thanks, I appreciate the comment and just know that I still have a lot to learn in this hobby! Most of what I have already learned came from this Forum or Tom Gibson's California Calssics. I just hope to find the time as I gain more knowlege to help others trying to do what I have done.....

There are those that contribute on a daily basis and that takes alot of effort and dedication. Actually I think that is just as important to the hobby as the actual production. I am not fast enough to be able to do both things so I have to concentrate on the model and its eventual completion.....

My hat is off to all those that help get beginners started and see them through the process; that takes enormous energy and patience.
I guess the big payoff is that from time to time we do produce something that will be used for the better in this hobby.....

All of you who have helped this rookie, yours truly, you know who you are and I thank you all!

Best Regards,

John
dzignguy

[n4gix]

I am particularly fond of the Connie, because when I was seven, my mom and I travelled TWA from KMIA to KLAX on one.

It was big (at least to my seven year old eyes!), noisy and exciting! I remember especially the GOOD FOOD served on china, with silver service, crystal glasses and linen napkins. This was a level of graciousness and gentility, which is sorely lacking in today's plastic world of peanuts and canned soda pop. :(

[Firestriker]

John,

I have also gleened a post and put it in a web page for the export settings that work really well.

http://www.oregon-coast.net/FS2002/3DSM-PandaX-FS.htm

While my projects are not nearly as ambitious as yours I have had good luck with the process.

Lou

[xfeed]

I see. Well, there are two ways to prevent MeshSmooth from smoothing certain areas: 1. do a chamfer along the corner and 2. assign different smoothing groups/Mat-Ids to neighbouring Polygons and activate either separate by materials or separate by SGs. The second approach is the one I am referring to and that one does not depend on more polygons.
Now to the solution of your problem: The symmetry modifier is nothing more than a mirroring applied to an object with following welding of the border vertices. Mirroring causes the faces to appear flipped in COF, so you need to employ the approach I described above: Assign an "Reset XForm"-Utility to your aircraft, then change it back to an editable poly. Now all the polygons will appear as in COF (i.e. with partly flipped polygons). Now comes the fun part: Select and flip the misoriented polygons, export again and then you're done.
Funny, but as long as you keep the symmetry mod on top of the stack and export, everything worked all right with my aircraft. The Bonanza with the shown modifier stack shows with no problems in COF. ;)