Forum:
Q & A
Thread:
the math behind mapping a sphere to cube faces
|
Re: the math behind mapping a sphere to cube faces
Posted: 16 Dec 2011 at 18:42 GMT
|
go to thread
|
I do not see why there would be any difference mapping onto cubes inside or outside the image sphere as all the projecting rays are from the centre/origin. Just a different scale.
The mapping function from equirectangular to the central cube face I derived:
X-> (Scale Factor) x tan( X x 180 / N )
[ |X|< N/4 ], same for Y
This is for equirectangular size N x 2N, centred on the origin.
Charles
|
|
alert moderator
|
|
Forum:
Q & A
Thread:
running locally
|
Re: running locally
Posted: 9 Mar 2010 at 12:20 GMT
|
go to thread
|
Just noticed you're talking about multires panos, so my previous post won't help.
Charles
|
|
alert moderator
|
|
Forum:
Q & A
Thread:
running locally
|
|
Forum:
Q & A
Thread:
Creating equirectangular projection from 6 faces
|
|
Forum:
Q & A
Thread:
PTgui Technique
|
|
Forum:
Q & A
Thread:
hugin 2009.4.0?
|
Re: hugin 2009.4.0?
Posted: 2 Jan 2010 at 16:10 GMT
|
go to thread
|
2009.4.0 is the most recent version of the source code, which has to be compiled by you to run on Windows. The version you have is the most recent pre-compiled one available for download.
see hugin.sourceforge.net/download/
Charles
|
|
alert moderator
|
|
Forum:
Q & A
Thread:
360x360 or 360x180?
|
Re: 360x360 or 360x180?
Posted: 22 Dec 2009 at 21:23 GMT
updated: 29 Dec 2009 at 20:37 GMT
|
go to thread
|
Well off topic, but instructive nonetheless. Let me have a go at an explanation.
Doug,
May I suggest your difficulty comes from the fact that although a sphere is highly symmetrical, the 2 co-ordinates we conventionally use to specify a point on its surface are not symmetrical. This is readily seen by considering circles of constant yaw/longitude - which are all great circles - and those of constant latitude/pitch - which vary in size.
So the range of yaw and pitch to map the spherical surface uniquely do not have to been the same.
Bear in mind also that we are not mapping a 3 dimensional space onto a 2D image (either equirectangular or cubefaces), but just the view of a 3D world from one point, which has no depth information and is a 2 dimensional entity.
Hope this helps!
Charles
|
|
alert moderator
|
|
Forum:
Q & A
Thread:
Hugin problem
|
Re: Hugin problem
Posted: 15 Dec 2009 at 22:23 GMT
|
go to thread
|
Weird results can come from the autopano-sift generating false control points from similar looking features. Check there are no control points between non-adjacent images.
Also it can help to start by entering initial values for pitch and yaw for each image.
Charles
|
|
alert moderator
|
|
Forum:
Q & A
Thread:
Doub'ts about create a virtual tour
|
Re: Doub'ts about create a virtual tour
Posted: 28 Sep 2009 at 13:46 GMT
|
go to thread
|
|
The Panosaurus can be used for multi row panos. The cheaper Panosaurus Rex is single row only.
|
|
alert moderator
|
|
Forum:
Q & A
Thread:
Viewing panoramas with a head mounted display
|
Viewing panoramas with a head mounted display
Posted: 23 Feb 2008 at 20:21 GMT
|
go to thread
|
My apologies if this has already been covered, but does anyone have experience of using a head mounted display / VR headset with head tracking to view panoramas?
I know there are some very sophisticated (and expensive) VR systems for professional use, but I am thinking of a more affordable unit like the 400 USD iWear VR920 www.vuzix.com/iwear/products_vr920.html designed for video games etc.
|
|
alert moderator
|
|
