Sun angles: accurate?

[George Smith]

I have noticed on a few other threads that some have mentioned the sun angle changes between basic & advanced rendering.  
So, then, does anyone know which one is accurate? Has anyone spent a few minutes calibrating it with a couple renderings?
I don't know enough about the science behind the changing sun angles to do this myself (yet).

From the first time I tried rendering a building I was very excited that I could punch in a date & time, and have the ability to see what the shadows would look like in real life!  It would be terribly disappointing to find out the whole thing is not really accurate, just a ballpark guideline.  The reason the accuracy is important to me is: I am using TFP to design houses, including exterior shades/canopies in order to conserve energy by only allowing the sun in in the winter but not the summer.  If it's inaccurate then this is an exercise in futility.

Thanks
George

[Allan Chesney]

Hi George,

It is not out very much between the two but which is the most accurate I do not know.  From what I recall with the earlier tests I did (purely a visual "calibration" - just a vertical pole from overhead on the terrain), it would only be out about a dregree of two - not enough to make any difference in what you are doing I don't think. 

Sunlight changes continually during the day of course as the sun is moving up and down as well as east/west (not that the SUN moves of course!) so the shadow will be different all the time.  Anictodally the vertical movement appears close to correct from the modelling I did the Home Renovator on the Synergy website (my former employer), which was all about sun angles and shading windows.  Actually you may find that of some interest to you now, since it deals with that subject and how to make an existing home energy efficient. It is only the basic principles that you undoubtably know but unfortunatley many seem to overlook. Although it is talking about renovations it all applies to a new home as well of course. www.synergy.net.au/renovator.  All the movies and animations were done in TFP.  A few still shots from it below and some others in the Gallery http://forums.turbofloorplan.com/index.php/topic,1011.0.html and a couple of animated GIFs http://forums.turbofloorplan.com/index.php/topic,1012.0.html.

You may also be interested in another TFP project I did on building a new energy efficient home. It starts with buying the block but you may already have done that. Also it relates to the southern hemisphere but the principles are the same of course,  just replace North with South!! ( www.synergy.net.au/envirohome ).  In the Renovator I left out reference to North/South so it applied internationally.

All the best with your project.  If you decide to do some proper sun angle calibration don't forget to post the results!! Incidentally the new version of Envisioneer allows you to add a new city by typing in the logitude and latitude, so I assume TFP V15 will inherit that.

Allan

[Doug.S]

I noticed the problem of sun shade errors when I tried to match a TFP design with an actual photo.

I never did any research on accuracy of TFP tho as it was fairly close.

However I'd be willing to collaborate on some simple testing to compare various TFP results with another 3D program I use to make sun shade movies...as it does the sun calcs too and also export movies and I do side by side stills for comparison....for example sunrise to sunset for exterior and interior views at various times of the year (equinox, summer/winter solstice).

The program I used was Carrara Studio Pro....importing the house from a TFP export. Also did a small set of work using SketchUp.

With a standard design of maybe a pole/square pipe sitting in center of a circular graph (N at north, with markings to see where the shadow falls)....at some good level of image size & graph lines to check accuracy? Made as a .3ds file so all have the same exact shape....or a dimensioned drawing to use for construct in any 3D program???

I have several links to web sites to retrieve sun data. Like here for a sun shadow graphical display...see "sun shadow applet"
http://www.jgiesen.de/sunshadow/index.htm
or
http://www.pilkington.com/resources/pilkingtonsunanglecalculatormanual.pdf
or Google search as there is an abundance of info/sites

Also using TFP you can predict where shadows from trees will fall during the year and as they grow over the years.

Fun stuff but does take some time to work it all out. But a simple test point may be sufficient to see how much error there is....and as Allan has said: a degree or two error is probably not noticeable ....especially for one day of the year (min./max.).

Another simple test is to construct something in TFP that matches something in your yard and compare the TFP result with the actual scene for the same day/date/time.  Of course you will need exact long. & lat. for your location but that is easy to get via www too.

Doug.S

[George Smith]

Ok. I ran some tests.  Here's what I found.
First of all, TFP is pretty accurate, all things considered.
I picked an arbitrary spot in the US, at a certain latitude & longitude.  I built a quick 3ds sundial in Turbocad (not a real sundial, but what I figured could work for a simple test board) so I could check the angles.  I made a vertical post, and figured out the angles its shadow should cast, and the length of shadow a 10" tall post would cast.  I chose the summer & winter solstices, at solar noon and another later time. I've attached the results, but could only attach the maximum 4 pics.
1. summer solstice, basic: solar altitude angle is really close (within a couple degrees), azimuth angle is perfect
2. summer solstice, advanced render: solar altitude angle is really close (within a couple degrees), azimuth angle is a bit off
3. winter solstice, basic: solar altitude angle is really close (within a couple degrees), azimuth angle is perfect
4. winter solstice, advanced: solar altitude angle is a bit off, azimuth angle is perfect
5. winter solstice, basic: solar altitude angle is really close (within a couple degrees), azimuth angle is perfect
6. winter solstice, advanced: solar altitude angle is a bit off, azimuth angle is perfect

It looks like basic render is great, very accurate.
Advanced render seems to lower the sun in the sky a little bit.

As a side note, the settings for daylight savings time are wrong.  TFP changes the sun by one hour all year round instead of only part of the year.  So checking the daylight savings time box would give an inaccurate render for roughly November thru mid-March.

Here's an excellent resource I found with some really cool free tools:
http://susdesign.com/tools.php

So the bottom line is, Allan's right: TFP is plenty accurate for just about everything we might use it for.  Basic render is more accurate, and watch out for DST errors.  I still wouldn't use it for designing overhangs, though, because a few degrees could make a difference, however slight.

[George Smith]

plus the other 2

[Allan Chesney]

Thanks George for doing those tests and posting the results. As is supsected it is pretty close for all normal purposes. I does not surprise me that Basic is more accurate, as Advanced is really designed for interior rendering, hence the issues with 'black walls' etc I suspect. http://forums.turbofloorplan.com/index.php/topic,997.0.html.  Move the camera inside and it is OK.

I presume the daylight saving time setting is expected to be manual, turn it on if you have DST and turn it off if you don't.  It probably has to be basically that way because of international variations.  Our daylight saving times are the opposite time of the year of course (Southern Hemisphere) but also they even vary so much between the states in Australia as far as start and end dates are concerned.  They would have to try to program each city, as some even adopt the time of the adjacent state for convenience. They also vary from year to year also. We in Western Australia do not have DST this year but have had for the previous 3 years, so I think they would be hard pressed to cope with such variations.   Even in the world of international timing they can't keep up!  Becasue of DST being 'voted out' in a referendum starting this current summer, I have had to set my phone to Singapore time, because the phone still thinks we have DST, so is an hour out! At the same time alarms going off an hour early are ideal for getting to meetings on time!!

More significant, but presumably not easily fixable, is the seasons for the plants. TFP ignores the southern hemisphere, so has Australia with snow in mid Summer, brilliant sunshine mid Winter, trees losing their leaves in Spring and bulbs and other plants flowering adundantly in Autumn (Fall)!!!  I have to remember to set the calendar six months wrong to get the seasons right, but of course that is not hard to do!

Those were interesting and useful tools on the web link your provided, so thanks for posting that too.  

Allan