Challenge day…oh…feck…

It seems more complex to work seven straight days than to make the scenery itself.

Daily affairs™ did not let me enough time to work on the scenery, and I think in the next weeks neither. So, again, the weight of reality falls over me in all his glory.

Frustration +10

Advertisements

Challenge day 03: playing with autogen

Beside having finished the watermask, today I’ve started to make the texture of autogen buildings (using googleEarth as reference to create my own ones) but the most important (and fun) task has been some test with the autogen vegetation:

(some custom autogen models without textures. No library objects has been used).

As I said before, I can’t distribute this testing because I have not the enough C++ skills to make an installer for this kind of feature. This method consists to add my own autogen models to the fsx. To do that, I must decompile the .spb file which contains the default autogen data into a .xml file, and then add my models. Finally recompile the modified xml into .spb again. To avoid conflicts with the default autogen and other installed addons, the ideal scenario it would be to make it automatically but sadly this task it´s not feasible to me.

As you can see in the screenshot, the autogen trees are not flat as the default ones, they are 3D lowpoly models, adding a lot of more depth and realism to the scenery without any lack of performance.

To generate the textures I´m using the Blender´s tree addon and late I baked them into a low poly object:

As you can imagine, it´s not a easy way and It requires a lot of texturing work to get the final texture.

Challenge day 02: nite textures and autogen

Yesterday I could not work on the scenery (RL matters) but today I´ve tried to hurry up and I’ve finished the night and spring variation and…

…I´ve started to put the autogen:

As you can see, the most representative buildings (the hotel, Industrial area…) haven´t autogen model because I want to make a small library with my own lowpoly models. 

Tomorrow I will make the textures for the building autogen, a little experiment with vegetation autogen and grass, the watermask and some flatten stuff.  

Challenge day 01: orthophotos & recoloring

For first course we have some “delicious” recoloring works

Original orthophoto from NAIP:

…masked:

And after some hours of cosmetic retouching:

Obviously, I could not resist to try it over Orbx´s PNW (late winter):

… in summer:

…and hardwinter:

I´ve made fall season too.

Tomorrow  it will be the turn of spring season and night variation (and other stuff).

But now, I’m going to watch Game of thrones new season 😉

Preproduction day 03: vector quest.

Today I’ve found some interesting sources for vector data:

The first one: a vegetation classification of Gilford Pinchot National Forest:

I will have to process it to use it as autogen (what kind of trees and all that stuff). As far I can see the main class is Hemlock trees. There is no a lot of diversity.

Although I have not yet a way to distribute it, maybe I will play, a little, generating my own custom tree textures.

I’ve found information about streets and roads from openstreetmaps and USGS but I haven’t checked them yet. I will use it to generate a mask in photoshop for the nocturnal texture. In this way I will gain an extra time generating that variation.

The downside of the day: I have not found yet any hangars reference, so I’m afraid that I will have to create “an interpretation” of them.

Preproduction day 02: references

Although my heart asks me to re-make Flying M Ranch, I’ve decided finally to choose Packwood. The reason is I have more information of this small airport and I undertaken a commitment with this location: In 2006 I tried make this scenery for fs9 but due the lack of time, references and knowledge I was no able to finish it.

In this first stage I’m searching references, not only images but also videos in youtube. And I’m amazed because there are several good examples:

Landing at Packwood (55S) by andyw248

Landing at Packwood, Washington by kodiak68jones 

Cessna 170 wheel landing Packwood by fishdoc55 

At the same time I´m checking the quality of the orthophotos from NAIP program. They are a little outdated (2009 and 2011) and they will need a deep color retouch but I hope that won’t be too much problematic.

The main drawbacks are the two hangars. As it happened to me in 2007 I am not finding photos or references. Nor the buildings around the airport. Streetview only shows Packwood´s main street (great source for some small buildings) but nothing from the airport surroundings 😦 I will try to contact with the authors of the videos).

More tomorrow: vectorial resources!

scenering jam session: preproduction day 01

ImageWhen I´m angry & frustrated I need let go a lot of energy. The best method I know is do it in a creative way. Yeah, due boring several reasons my frustration level is at its top, so I’ve decided to push myself into a strange-alike-jazz-jam-session but “à la scenery design”.

The idea is stop for a while LEMT´s development (I need step back little from it) and gimme 7 days to make a little scenery. Yep, you have heard correctly: to make an scenery in 7 days. Obviously it won’t be at the same complexity level that LEMT but I want to challenge myself to make something more interesting than to put some objects from a library over a orthophoto.

So this week I’m going to make a preproduction task: I want to collect all the possible references from a small airstrip, all the inspirational photos and videos, testing some methods with several scenery design tools, and the next week, close the door of my room until I get an interesting location for FS.

Briefly, I want to challenge my texturing/modelling skills and all the learned knowledge in the last year under over-stress deadline.

I have two candidates: Packwood (55s) or a remake from Georender´s classic: Flying M Ranch (OR05) (BTW why did he left us behind??? He was the best! I don´t forget his terrific platform´s texture in Emma Field. I spent so many hours studying it).

And now, I need to study carefully which will be the choose one. Tomorrow, the solution.

February update

As some people asks me if I have thrown the towel I´m going to post a quick update to deny all those concerns 😉

All the snapshots show the framerate. My workstation´s specs are:
Intel Core Duo T8300@2.4Ghz
4Gb RAM
Quadro 3600 (similar to 8800) 512Mb.

This my FSX work version. I don´t apply it any kind of tweak (nor bojote, nor bufferpools, nothing). I try to work as much as possible in a “default enviroment”.

Before to start I have to clear something up: the main goal of this project is to learn and improve my 3D modelling and (mainly) my texturing skills. So to me, it´s a priority to experimenting and testing different methods instead of release it as soon as possible.

That is the reason because I did not post nothing in the last two months. My initial intention was to have now an alpha version, with all hangars textured, to make some performance tests. Sadly, as a noob in these subjects, I misjudged the difficult of the task.

As I said before, my curiosity to explore and learn new ways to achieve the quality that I´m looking for has been the reason to this delay. And the main guilty is…. this mound:

As you can see the airport´s main building/tower is over a little slope. The easy way would be to take away and put it over a flat terrain (as the others hangars are) or, as I made at the beginning, to create a very basic low poly model and fake it later with a high resolution texture. But the results were just crap.

Then I began to search for other ways to get better quality. One of them was the blender´s sculpting feature. I´m not going into details (youtube could be your best source if you want to know more) but, briefly, with this feature I can create a high poly mesh with a lot of detail and then baked it to a “friendly fps” low poly. A picture speaks a thousand words:

Not bad for a newbie 😉

As you can imagine this method is not easy to learn nor something that you can achive only looking one or two tutorials but it´s not the holy grial of 3D development. This is very common technique in videogame development. But it took me some nonstop journeys to get reasonable results to use it in FS (as any other mesh). BTW, the method it´s very, very promising to get a very realistic textures for ground polys.

As I said before, you can find a lot of basic tutorials about sculpting in blender or any other similar software in youtube.

At the same time I´ve been texturing and modeling other stuff. All the hangars are modeled and their basic shade is done (ambient occlusion layer) . I´m still texturing but at low rate because right now I´m searching new ways for improve the quality and, very important for me, the “productivity” generating them. I get older and it´s very sad to discover that you are not so fast learning new stuff like in the past 😦

Be aware: The next hangers are not textured, it only are shaded by a basic AO layer. I repeat: these images are WIP 😉

So, the big question: when will I finish it? it will be done when it´s done (I love that sentence XD). I really don´t know. From now on I can´t be full time on this (I must pay bills and all that boring stuff 😉 ). But, anyway the the plan would be once the hangars textures are finished I will release a close alpha to check the performance and in the meantime I will start to model some hangar´s atrezzo, stactic aircrafts, fences, cars, P.O.I in the area (Navalcanero´s silo, Xandadu shopping center, Batres´ Castle, some bull fighting ring) and, obviously… ¡grass!

And what happend to Madrid Photoscenery??? Well I made some progress. All the southern Madrid is recolored (I tease you LECU and LEGT won´t be censored this time). But as I will have so little spare time and my “machine” spend a lot of time opening these so huge files (more than 2Gb per file) this is going take a while.

Anyway, as soon I get further progress I will post them here.