[HDRI] hdrgen question, systematic error?

[Jack de Valpine]

Hi Mark,

I had something similar occur, however the scene was very different. In 
my case there was a cloud moving past a building. With automatic 
alignment turned on (default behavior) I got a result similar to yours. 
It seemed like the program was tracking on the brightest "source" in the 
set of exposures. I wonder during the time that you took your exposures 
was there a big change in the daylight coming into the space? What 
happens if you turn alignment off (-a included in the command line)?

Or alternatively, are you sure that you are passing only the exposures 
you want to hdrgen? I have done this in the past where I will take a set 
of exposures and part way through shooting the set I realize that I 
screwed something up so I will start over and in the process may change 
the camera position. However I forget to delete the previous set from 
the folder, so they get passed along to hdrgen as well.

-Jack

Mark Stock wrote:
> Folks,
>
> I just used the Linux version of the hdrgen program, and there appears 
> to be a systematic error somwhere.
>
> My first attempts were handheld, and alignment always failed. My 
> attempt today was using a tripod in my living room, and the alignment 
> error is very regular:
>
> http://mark.technolope.org/transfer/dscn6880pfilt.jpg
>
> This was done with:
> hdrgen dscn68??.jpg -o test.pic -r cam.rsp
> pfilt -1 -x /2 -y /2 test.pic | ra_ppm | cjpeg -q 90 > dscn6880pfilt.jpg
>
> uname -a gives:
> Linux localhost.localdomain 2.4.22-1.2115.nptl #1 Wed Oct 29 15:42:51 
> EST 2003 i686 i686 i386 GNU/Linux
>
> Is there anything I can do about this?
>
> Mark
> mstock at umich.edu
>
> _______________________________________________
> HDRI mailing list
> HDRI at radiance-online.org
> http://www.radiance-online.org/mailman/listinfo/hdri
>
>

-- 
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction