After having written the last tutorial – How to model a building, I have been asked by a few architectural illustrators why I don’t use Autocad to model buildings, since “it is more accurate and takes less time”.

Basically these are the 2 main arguments that I heard in favor of modeling in autocad, and after the discussion that I had with these people, I figured that this could make a quite interesting article.

I will try to talk about each argument systematically:

1) The precision issue
First of all, regarding the accuracy thing, it would be foolish of me not to admit that autocad is far more accurate than 3d max. This is why autocad (and other similar software) is used to draw plans and elevations that are used in construction and other areas that require precision, while 3ds max is used in areas where the “visual factor” is more important.
I think that we can all agree that if you are working in architectural visualization your main goal is to create outstanding renderings.

However, I’m also aware of the fact that this type of activity has a very technical component, which is to create the 3d models of the buildings, following accurately the cad files provided by the architect. But what you would define as “accuracy” in this case? If you want to take it to “microscopic” level, than 3d max is definitely not the proper software. The truth is that you don’t need this kind of precision when modeling 3d buildings; who would notice if a wall has 7 meters in the cad drawing and the geometry that you’ve modeled is 4 millimeters longer?

2) Speed
Some may argue that modeling 3d buildings in autocad is faster than in 3d max. If you have a solid cad background, like lots of architectural illustrators, it is absolutely normal to work faster in cad than in max. This is not valid for others, so until someone will be able to prove me wrong, I stand by my opinion that this only depends on each person’s background.
Furthermore, if you need close-ups on some details of the geometry, you may want to chamfer some edges, adjust some smoothing groups, etc., in order to avoid ultra sharp angles between faces that tend to give the image a “CG-ish” look.
If you try to do this in autocad, you have a lot less tools, and even if you manage to do it, there is no guarantee that the result will look the same after being imported in 3ds max. Which brings me to the next issue…

Importing 3d geometry from autocad can be really frustrating.

Inverted normals, chaotic smoothing groups and messed topology of the geometry are some of the most common problems that I have encountered when I needed to import cad geometry provided by the clients.

Basically these are the main reasons why I don’t use autocad to model buildings and I prefer to do it directly in max.

I look forward to hearing your comments/arguments if you don’t agree with what I have written, or if you just have something to add.

In my experience as an architectural illustrator I have found out that in some situations one can obtain a lot better results by composting elements cropped from photos into 3d renderings, rather than actually rendering them.

Many use this method to add entourage elements such as people or vegetation to renderings and maybe sometimes cars. However, this method also works well for some other things. There was a time when I used it intensively in my renderings for placing almost anything from office clutter, interior plants to curtains and even restaurant tables.

I have found it to be very useful for several reasons.

1)-the rendering time was considerably lower than it would have been if I rendered all the objects directly in 3d

2)-I spent a lot less time compositing elements from photos to the renderings that I would have spent on modeling, texturing and material settings for every specific object

3)-sometimes one may find it hard to model something in 3d, especially if that it does not have enough info (like correct front views, side views and top views), or simply because the object is too difficult to replicate in 3d due to the complexity of the model or “special materials”so the results that can be obtained by using a composting method may actually look a lot better than the ones obtained by the traditional 3d methods.

4)-I have avoided a lot of ram memory errors. These type of errors are quite frequent if have a 3d scene overloaded with polygons and textures from lots of objects (if you don’t have a very powerful workstation)

5)-the feedback from my clients was great!

In conclusion I would like to say that no one should be afraid to use photoshop whenever it is possible. I know that a lot of 3d experts out there actually recommend to model and texture everything from scratch (and to some extend they are right) but again, there are certain situations
when photoshop can be a real life saver. You will find some tutorials on this blog on how to add different elements like hedges, restaurant tables, vegetation and curtains. I hope that you will find them useful