I am creating a WWII german u-boat for my video project and i would like to ask a few questions about workflow and level of detail (LOD).

https://newsimg.bbc.co.uk/media/image...boat_300pa.jpg

I need to make 3 versions of the boat, low poly for far away shots, medium poly for mid range shots and high detail for the close ups. There are various tubes, hand rails, nuts and bolts visible on the high detail version that obviously won't need to be as high poly on the distant shots but they will still need to be included.

As far as workflow goes would it be best to start with the low poly version making handrails say for example (cross section) 3 or 5 sided polygons and then punching it up to 20 sides for close up so you don't see any 'facet'ing' (sp). This is of course proportional to the render time.

is it best practise to start with the low poly version to get sense of proportions and then use it as a template for the high detail version ? If i leave the components in the correrct places then I should be able to (providing I havn't deleted history) punch up the side edges to increase the details. So i could leave some piping or torpedo tubes at 5 sides and then increase the divisions up to 20 for the close up version.

I'm trying to avoid building 3 models from scratch, any thought on using Mesh > Smooth ? I have tried this but it can alter the shape too much and then I'll have to add edge loops to tighten then edges up.

Let me know your thought on this project.

Many Thanks