Sun study

[Ignacio Azpiazu]

I am needing to produce a sun study showing the net new shadow thrown on a park by a new building. If anybody knows any procedure better than what I figured and describe below, in any software, please let me know.

The process requires calculating the shadowed area on the park as existing, the shadowed area on the park with the new building included, and subtracting the first value from the second value, for *about 1400 moments in the year* (one shot every 15 minutes for 27 days each representing a week in a half year).

The 54 sun studies part (27 existing, 27 new) gets taken care of by ArchiCAD, I guess exporting as 2D lines. Then there is manual magicwanding or whatever to highlight the area of shadow on the park itself by getting it into a polygon or zone, and checking what that value is. My optimist estimate is that each moment may take between one and two minutes, which makes the job a 24 to 48 man hours job.

I find it hard to believe that every single sun study provided to the city required a week or more from some unfortunate guy, and I hope there must be some better method.

[Anonymous]

It seems to me that you've figured out the best method. It laso seems that this is a pretty onerous requirement. I would think that a smaller number of carefully selected studies should suffice. Especially when combined with some animated visual studies that AC can produce semi-automatically.

One thing I would suggest is to see if there is a consistent mathematical relationship between the different moments. If so you might only need to to measure some number of key frames and interpolate the others by calculation.

[Aaron Bourgoin]

I am writing to invite anyone who has used Japanese versions of ArchiCAD - or to anyone at Graphisoft, for that matter, to comment on this thread.

Some time ago I was shown an Add-On (in the very early days of Add-Ons) that performed shadow studies in ArchiCAD and prepared data to submit data to the (Tokyo?) building authority. Apparently their requirements were quite complex. What I also recall from this tool was the speed with which it did all this on what would be a dinosaur of a computer by today's standards.

Bearing in mind that the Zone Tool was originally conceived for the German market and addressing Berlin's rigorous space type and quantity requirements for housing projects, would it be too much to consider finding this amazing tool and brining some of its functionality to the broader user base.

[Karl Ottenstein]

Here is a pretty lame idea, but I wonder if it will work to get an image showing the total shadow area. How about setting up a (digital or other) camera on a tripod in front of your monitor and taking a time-lapse photo (basically, keep the shutter open) of the entire sun study animation? Use the smallest aperture and you might need a filter too if you are still overexposed if the animation is very long. You would want to carefully choose your scene materials/colors so that the result gives the mask that you want. (Minimize ambient light I think to boost contrast.) I think the result would be an image with near-black only in permanent shadows, and shades of grey indicating shadow strength as you move towards unaffected areas.

A similar option, taking more computer time but no human time, might be to create an automation script in Photoshop that takes a folder with all of the individual frames of the animation and merges them, one by one (bring new one in as a layer with appropriate blending mode, merge layers, bring next one in, etc.)... ??

Thinking sideways...

Karl

[Anonymous]

Or take Karl's idea about photoshop and think about using a vectorial sunstudy, parallel projection top view. Vector data.

Generate the PMK's and place them in register in your layout. Plot result or draft over the end result showing the net yearly shadow area. How many samples do you really need? Interpolate between dates and times.

[Rakela Raul]

there was a comment from somebody in europe on a city's reqt regarding same subject, no more than a 1 yr back.....maybe it can also help....

[Ignacio Azpiazu]

The requirements I described in the first message (number of snapshots, etc.) are those of the city of San Francisco.

In the presentation the graphics are meant to support the shaded area calculations, and the numbers are what count (they go into a spreadsheet and from there to the Parks authority or whatever). So that movies or pictures won't work unless they are vectorial and with some tracing let me calculate the area using the zone or polygon tool.

[Anonymous]

Well that's San Francisco for ya. Lot of luck with that jurisdiction.

[Anonymous]

Burginger wrote:
Well that's San Francisco for ya. Lot of luck with that jurisdiction.

I assume he is referring to San Francisco Argentina. Is that what you meant or are you thinking of Bagdad by the Bay?

Ignacio wrote:
In the presentation the graphics are meant to support the shaded area calculations, and the numbers are what count (they go into a spreadsheet and from there to the Parks authority or whatever). So that movies or pictures won't work unless they are vectorial and with some tracing let me calculate the area using the zone or polygon tool.

It still seems to me that there are probably some reasonably simple mathematical relationships that should allow you to calculate values between a few measured key frames.

[Ignacio Azpiazu]

Matthew wrote:

Burginger wrote:
Well that's San Francisco for ya. Lot of luck with that jurisdiction.

I assume he is referring to San Francisco Argentina. Is that what you meant or are you thinking of Bagdad by the Bay?

It's the San Francisco Burbinger is thinking of. Luck and temperance.