Is Max the right package for this type of work

[Neil Woodhouse]

Greetings all and a belated New Year.

 

A quick request for guidance on something.

 

I've recently started doing a considerable amount of work for planning applications, specifically environmental impact imaging for WT's

 

The work involves UK OS Maps/Google Earth and client photographs from various viewpoints.

 

What i am doing is setting the units in max to meters. Going to top view in max, then applying the OS map as a viewport background, placing a box in the top viewport where the WT sits and then using the hard copy to measure and scale off the distances/viewpoint directions for the cameras. I then take the client images and add them as environment & bakground images, so what i get is an image with a 3D WT sat in the picture, which should be correctly positioned in the final render.

 

This seems very clumsy, and i'm worried about perpspective issues, although the client is pleased with the result.

 

I'd like to improve on this.

 

Am i:

 

1. Doing it the right way?

2. Does anyone else do this sort of thing, if so any tips.

3. Is there alternative software packages that can do this better than Max, i.e Photomodeller

 

Regards

 

Neil

[Virgil Johnson]

Neil

 

I take it you are doing what I know as a visual impact study?

 

Not sure what an OS map is.

 

WT = water tank (tower) ?`

 

You could start with the thread at this site dealing with visual impact studies that started here about Feb '07.

 

Since visual impact studies are based on accuracy perhaps you should be using something like AutoCad along with Max.

(Photoshop is of course useful)

 

Digital maps and surveys rule the day.

 

 

 

virgil

[3dp]

don't be fooled by my current location i'm from the correct side of the pennines (and it's red)

having said that you may want to look at this

 

http://www.zmapping.co.uk/

 

it's laser scanned 3d site model service - not cheap but a client should pay for this service if he's going in with the services of a planing consultant

 

having undertaken many planning applications in the past and applying verified view methodolgy this really eliminates a lot of debate with designer/planner and client as although the z mapping is not 100% accurate sme of the site i'd worked up in Bradford/Halifax in the past which would be tricky to survey this means you spend time of site photography and place your 3d camera in the model where you were on site and camera match it.

still need to do a little fudge work but this takes away lot headaches on tricky sites

it really depends on you clients wallet size and a planners quest for accuracy

[Neil Woodhouse]

Thaks for the replies so far. To answer a couple of points.

 

Virgil, yes these are visual impact studies. WT means Wind Turbines and OS Maps are the UK Ordnance Survey Maps.

 

I've looked at the zmapping site and i'm not sure it's what i need as the photographs i work with are fields with the occasional wall, so it's very much a case of fudging the 3d model in the image to match the odd ref point.

 

I shall have to have a re-think.

 

P.S. Since when has the red side of the pennines been the right side LOL.

[Virgil Johnson]

Neil

 

Just some quick thoughts.

 

Get as many measured reference points as you can. (xyz -distances and azimuth - remember the difference between magnetic and true north )

 

I locate them on a separate model (I use AutoCad first and then take into Max) and then bring in the photo to see what lines up with what.

 

That is - take the rendered reference points picture and overlay onto the field photograph and see what lines up.

 

Do the modeled reference points match up with the reference points in the field photo?

 

This is best done in Photoshop I think.

 

Not all the measured (modeled) reference points will match.

Some are not accurate to begin with.

And other will suffer from lens distortion - especially as you move away from the focal point of the lens.

 

Use the more accurate modeled reference points to change the max (WT model) camera model render output to match the reference points.

 

Hope this gets an workflow idea going for you.

 

virgil