Post any defects you find in the released or beta versions of the ImageMagick software here. Include the ImageMagick version, OS, and any command-line required to reproduce the problem. Got a patch for a bug? Post it here.
the supra-white (> 1.0) pixels are not clamped to white but are instead mangled to seemingly-random colors.
(Maybe an integer overflow or something like that?)
It displays fine if you first convert it to EXR format.
IM 7.0.6-2 Q16 HDRI x64 (Windows 7)
Last edited by Dabrosny on 2017-08-10T13:29:03-07:00, edited 6 times in total.
-Dabrosny [Using IM7.0.6 or higher, Q16 HDRI x64 native executable, command line, often invoked from cygwin bash/sh (as of Aug. 2017)]
fmw42 wrote: ↑2017-08-07T15:13:43-07:00
Perhaps you should upgrade your IM to the latest version. 7.0.2-9 is 36 versions old.
As in my original post as well as in the one you are responding to (& my sig btw), i am using 7.0.6 x64 Windows. I said that I *also* reproduced the problem in 7.0.2 x86 32bit for Windows.
I'm still hoping someone will try to reproduce under Windows since you have already told us that you see no problem on your Mac.
-Dabrosny [Using IM7.0.6 or higher, Q16 HDRI x64 native executable, command line, often invoked from cygwin bash/sh (as of Aug. 2017)]
Would anyone please try opening the mentioned memorial.hdr file in display.exe under Windows (with HDRI and/or not)?
I've now *also* tested it on a Non-hdri Cygwin ImageMagick 6.9.1-3 Q16 i686 2015-06-21 (DPC OpenMP) (Cygwin 2.0.4 32-bit) and the issue does *not* occur there.
Fred already said the issue did *not* occur on his Mac (perhaps, like cygwin, this is a unix-like build?).
So far, the problem has occurred for me in these Windows HDRI builds:
7.0.6-2 Q16 x64 2017-07-22; Visual C++: 180040629; Cipher DPC HDRI Modules OpenMP
7.0.2-9 Q16 x86 2016-08-14; Visual C++: 180040629; Cipher DPC HDRI Modules OpenMP
-Dabrosny [Using IM7.0.6 or higher, Q16 HDRI x64 native executable, command line, often invoked from cygwin bash/sh (as of Aug. 2017)]
Mac and Cygwin are unix systems. If you have Windows 10. you could also test its unix environment. So it sounds like a Windows issue. Sorry, I do not have a Windows system to test or I would do it.
I don't use imdisplay. But the image does show as I would expect using imdisplay in both v6.9.5-3 and v7.0.3-5, Q16, non-HDRI, on Windows 8.1, if that's any help.
Can anyone out there try it with an HDRI Windows build?
Sounds like the values aren't being clamped properly in these HDRI Windows builds.
snibgo wrote: ↑2017-08-08T11:52:31-07:00
I don't use imdisplay. But the image does show as I would expect using imdisplay in both v6.9.5-3 and v7.0.3-5, Q16, non-HDRI, on Windows 8.1, if that's any help.
-Dabrosny [Using IM7.0.6 or higher, Q16 HDRI x64 native executable, command line, often invoked from cygwin bash/sh (as of Aug. 2017)]