After all the work I did on the COLLADA and KMZ importers for ModelConverterX, I thought it would be a good moment to try to model a bit more in SketchUp myself. So about a week ago I cycled to the nearby town of Abcoude and made a bunch of pictures of their local church. With those as my reference and some maps I found on the internet, I started modelling the church today. In this blog posts and those that follow until I have finished the object, I will try to give an overview of my adventures while learning more about SketchUp.

Being a bit of a number-head, one of the things that worried me about SketchUp was if I could easily enter numbers for sizes of objects and offsets. I know many people do not worry so much about this, they are mainly concerned with how the object looks, but my style of modelling involves using paper and pencil to calculate distances and things like that. And although it works different from GMax, I must say that I am very happy with the way SketchUp allows you to enter distance when drawing lines or when moving objects.

Below you see a picture of the church and drawing the roof was quite a nice learning experience. With all the SketchUp functionality to draw lines parallel or perpendicular to other lines or the axis, it was in the end quite easy to make this roof. Even though the walls supporting it have different heights.

While modelling the church I also experimented with the match photo function of SketchUp. This function allows you to align your view so that it matches with the position from which you shot your photo. Afterwards you can then model in that view and make the object match the dimensions as show on the photo. At first I had some troubling setting this up correctly, but I think that was many due to make lack of good overview photos. Due to all the trees planted around the church, I mainly made detail photos, while not having many good overviews. My impression is that the match photo function works best with pictures that show a big part of the object.

But in the end I got it setup for some shots I made and it turned out to be a really easy way to determine the height of the roof and the tower. Below you see an picture of the object being aligned with the photo (I made the object transparent so you can see the picture better).

So as you can see in the pictures, the basic shape is done now. I will probably have to add some minor details to gables and that kind of stuff. But I am quite satisfied with the result until now. The next step will be texturing the object. I have already done a little experimenting with projecting pictures on the object (after I used the match photo function), but as you can see in the picture below the results are not good enough yet. I am planning to see how far I can go with projecting the pictures and then using the Drawcall Minimizer of ModelConverterX to make a nice texture sheet. If that does not work, I will go back to the traditional way of making the texture sheet myself.

Another thing I hope to learn from these experiments is how to model most efficiently in SketchUp (from the point of view of FSX). So how do you keep the polygon count and the drawcalls within limits. Using ModelConverterX I can get nice statistics about all those performance related numbers. So more on that and the texturing in a next blog post.