[Moved]octave 8 bit pixels

Question

I installed octave from the repos but when I use the imread function I get the following:
warning: your version of GraphicsMagick limits images to 8 bits per pixel
I can't be restricted to 8 bits per pixel and from what I have read the default is 16, so I am not sure why the version of GraphicsMagick in the repos is compiled using 8. I tried to recompile both GraphicsMagick and Octave by slightly modifying the PKGBUILD I downloaded from the x86_64 extra repository. The only change is that I added --with-quantum-depth=32. PKGBUILDS shown below.
Both compile using makepkg and install via pacman but when octave compiles I get the following warning:
configure: WARNING: GraphicsMagick++ library not found. The function for reading image files will not be fully functional.
And when I fire up the installation of octave and attempt to use imread I get
error: imread: invalid image file: imread: image reading capabilities were disabled when Octave was compiled
Any ideas?
PKGBUILD for GraphicsMagick
# $Id$
# Maintainer: Ronald van Haren <ronald.archlinux.org>
# Maintainer: Gaetan Bisson <[email protected]>
# Contributor: Jaroslav Lichtblau <[email protected]>
# Contributor: Anton Leontiev <[email protected]>
pkgname=graphicsmagick
pkgver=1.3.13
pkgrel=1
pkgdesc='Image processing system'
url='http://www.graphicsmagick.org/'
arch=('i686' 'x86_64')
license=('MIT')
makedepends=('perl')
depends=('bzip2' 'freetype2' 'ghostscript' 'jasper' 'lcms2' 'xz'
'libtiff' 'libwmf' 'libsm' 'libxml2' 'libltdl')
options=('!libtool')
source=("http://downloads.sourceforge.net/project/${pkgname}/${pkgname}/${pkgver}/GraphicsMagick-${pkgver}.tar.xz")
sha1sums=('b85b21785bb072fd795f7fc6ac902cffae4e96e6')
build() {
cd "${srcdir}/GraphicsMagick-$pkgver"
./configure \
--prefix=/usr \
--with-perl \
--enable-shared \
--disable-static \
--with-gs-font-dir=/usr/share/fonts/Type1 \
--with-quantum-depth=32 \
make
package() {
cd "${srcdir}/GraphicsMagick-$pkgver"
make DESTDIR="${pkgdir}" install
# Install MIT license
install -Dm644 "Copyright.txt" "${pkgdir}/usr/share/licenses/$pkgname/Copyright.txt"
# Install perl bindings
# The patching was introduced in order to build perl module without installing package itself and
# not to introduce unnecessary path into LD_RUN_PATH
cd PerlMagick
sed -i -e "s:'LDDLFLAGS' => \"\(.*\)\":'LDDLFLAGS' => \"-L${pkgdir}/usr/lib \1\":" Makefile.PL
perl Makefile.PL INSTALLDIRS=vendor PREFIX=/usr DESTDIR="${pkgdir}"
sed -i -e "s/LDLOADLIBS =/LDLOADLIBS = -lGraphicsMagick/" Makefile
make
make install
# Remove perllocal.pod and .packlist
rm -rf "${pkgdir}/usr/lib/perl5/core_perl"
rm "${pkgdir}/usr/lib/perl5/vendor_perl/auto/Graphics/Magick/.packlist"
PKGBUILD for Octave
# $Id$
# Maintainer: Ronald van Haren <ronald.archlinux.org>
# Contributor : shining <shiningxc.at.gmail.com>
# Contributor : cyberdune <[email protected]>
pkgname=octave
pkgver=3.4.3
pkgrel=1
pkgdesc="A high-level language, primarily intended for numerical computations."
arch=('i686' 'x86_64')
url="http://www.octave.org"
license=('GPL')
depends=('fftw' 'pcre' 'curl' 'lapack' 'libx11' 'graphicsmagick' 'glpk' 'hdf5' 'gcc-libs' 'qhull')
makedepends=('texinfo' 'graphicsmagick' 'gcc-fortran' 'umfpack' 'texlive-core' 'fltk')
optdepends=('texinfo: for help-support in octave'
'gnuplot: alternative plotting'
'umfpack: LU decomposition of some large sparse matrices'
'fltk: alternative plotting')
source=("ftp://ftp.gnu.org/gnu/octave/octave-$pkgver.tar.bz2")
options=('!emptydirs')
install=octave.install
sha1sums=('fe622c28a38f8730c59e46211bc7b18e7f51a679')
build() {
cd ${srcdir}/${pkgname}-${pkgver}
# http://www.nabble.com/Random-rounding-errors-td16010966.html
FFLAGS="-O -ffloat-store" \
./configure --prefix=/usr --libexecdir=/usr/lib \
--enable-shared --disable-static --with-quantum-depth=32 \
--with-magick=GraphicsMagick
LANG=C make
package(){
cd ${srcdir}/${pkgname}-${pkgver}
make DESTDIR="${pkgdir}" install
# add octave library path to ld.so.conf.d
install -d ${pkgdir}/etc/ld.so.conf.d
echo "/usr/lib/${pkgname}-${pkgver}" > ${pkgdir}/etc/ld.so.conf.d/${pkgname}.conf
Last edited by scarypezsanta (2012-01-06 19:13:40)

Answer 1

Hello,
I think for greyscale, the RBG colors are all equal, and vary between 0 and 255.  That is, 0,0,0 corresponds to black and 255, 255, 255 correponds to black... where 127, 127, 127 would be some intermediate shade of grey.  You can use RGB to Color.vi to build an array of colors which correspond to these greyscale colors to use for your color table input.  Attached is a screenshot of the code which you can build very quickly in your version of LabVIEW!
Best Regards,
JLS
Best,
JLS
Sixclear
Attachments:
Generate Greyscale Color Table Array.bmp ‏114 KB