Hi all,

Hopefully this is a simple question:

Is it possible to get the focal length of a camera view from within

Radiance?

I know that the view angle specification makes this unnecessary as an

input parameter but was wondering if it is possible to reconstruct the

physical camera properties from only a Radiance view description.

Cheers

Rich

(Hoping to have a decent animation and the Maya2Rad tools online

soon.promises, promises ;-))

## ···

____________________________

Richard Gillibrand

Department of Computer Science,

University of Bristol,

Tel: 0117 9545256

____________________________

Hi Rich,

Marcus Jacobs asked a similar question about focal lengths last month on this group (Oct. 19th, in fact).

For a 35mm frame (which measures 36x24mm on the actual exposure), the equation for the horizontal and vertical view angles (in degrees) is:

h = 2*180/PI*atan(36/2/f)

v = 2*180/PI*atan(24/2/f)

To derive the focal length from the view angle, you just invert these:

f = 36/2/tan(PI/180/2*h)

f = 24/2/tan(PI/180/2*v)

As you can see, there are two formulas above, which means that you can derive the focal length from either the horizontal or the vertical view angles, but they won't necessarily agree. This is obvious because any given Radiance view may not have the correct image aspect ratio to match 35mm (or any other format) film. You have what is known in math as an "overdetermined problem." You can choose to ignore this and simply average the results from these two formulas, or pick the larger or smaller of the two, depending on whether you wanted to fill the frame completely or keep the image with black borders (i.e., "letterbox" format).

-Greg

## ···

From: R Gillibrand <gillibrand@cs.bris.ac.uk>

Date: Wed Nov 12, 2003 4:10:34 AM US/Pacific

Hi all,

Hopefully this is a simple question:

Is it possible to get the focal length of a camera view from within Radiance?

I know that the view angle specification makes this unnecessary as an input parameter but was wondering if it is possible to reconstruct the physical camera properties from only a Radiance view description.

Cheers

Rich

Greg Ward wrote:

Hi Rich,

Marcus Jacobs asked a similar question about focal lengths last month

on this group (Oct. 19th, in fact).

For a 35mm frame (which measures 36x24mm on the actual exposure), the

equation for the horizontal and vertical view angles (in degrees) is:

h = 2*180/PI*atan(36/2/f)

v = 2*180/PI*atan(24/2/f)

To derive the focal length from the view angle, you just invert these:

f = 36/2/tan(PI/180/2*h)

f = 24/2/tan(PI/180/2*v)

As you can see, there are two formulas above, which means that you can

derive the focal length from either the horizontal or the vertical view

angles, but they won't necessarily agree.

Actually, the focal length is determined by the largest

dimension of the picture, the diagonal, which makes it

somewhat independent of the aspect ratio.

The normally used formula for 35mm film pictures

therefore goes like this:

f = 43.27/2/tan(PI/180/2*h)

-schorsch

## ···

--

Georg Mischler -- simulations developer -- schorsch at schorsch com

+schorsch.com+ -- lighting design tools -- http://www.schorsch.com/

Ah!

Thanks Greg, I thought the question was familiar when I asked it!

Rich

## ···

____________________________

Richard Gillibrand

Department of Computer Science,

University of Bristol,

Tel: 0117 9545256

____________________________

-----Original Message-----

From: Greg Ward [mailto:gward@lmi.net]

Sent: 12 November 2003 16:45

To: radiance-general@radiance-online.org

Cc: R Gillibrand

Subject: Re: Getting the focal length of a Radiance camera

Hi Rich,

Marcus Jacobs asked a similar question about focal lengths last month

on this group (Oct. 19th, in fact).

For a 35mm frame (which measures 36x24mm on the actual exposure), the

equation for the horizontal and vertical view angles (in degrees) is:

h = 2*180/PI*atan(36/2/f)

v = 2*180/PI*atan(24/2/f)

To derive the focal length from the view angle, you just invert these:

f = 36/2/tan(PI/180/2*h)

f = 24/2/tan(PI/180/2*v)

As you can see, there are two formulas above, which means that you can

derive the focal length from either the horizontal or the vertical view

angles, but they won't necessarily agree. This is obvious because any

given Radiance view may not have the correct image aspect ratio to

match 35mm (or any other format) film. You have what is known in math

as an "overdetermined problem." You can choose to ignore this and

simply average the results from these two formulas, or pick the larger

or smaller of the two, depending on whether you wanted to fill the

frame completely or keep the image with black borders (i.e.,

"letterbox" format).

-Greg

From: R Gillibrand <gillibrand@cs.bris.ac.uk>

Date: Wed Nov 12, 2003 4:10:34 AM US/Pacific

Hi all,

Hopefully this is a simple question:

Is it possible to get the focal length of a camera view from within

Radiance?

I know that the view angle specification makes this unnecessary as an

input parameter but was wondering if it is possible to reconstruct the

physical camera properties from only a Radiance view description.

Cheers

Rich

R Gillibrand wrote:

Ah!

Thanks Greg, I thought the question was familiar when I asked it!

Rich

btw: there's a search engine combing though the mailman archives every night: http://www.radiance-online.org/search.html

## ···

--

pab-opto, Freiburg, Germany, www.pab-opto.de

Peter Apian-Bennewitz wrote:

R Gillibrand wrote:

Ah!

Thanks Greg, I thought the question was familiar when I asked it!

Rich

btw: there's a search engine combing though the mailman archives every night: http://www.radiance-online.org/search.html

... and The Mighty Google also crawls the site. You can restrict a Google search to a specific domain by adding "site:domain.com" in your search string:

Richard, you could locate the juicy bits of your recent request by entering the following srting at Google:

focal length site:www.radiance-online.org

Have fun, and thanks again to Peter for hosting this site & list. It's a great resource!

## ···

----

Rob Guglielmetti

e. rpg@rumblestrip.org

w. www.rumblestrip.org