Legacy ImageMagick Discussions Archive

I need to get the depth for all channels. The simplest way is to use identify with the verbose option. But in IM7 this command is very time consuming. I think this belongs to the signature calculation. Is it possible to get this information without signature calculation? I already search for a ...

The output for "-print %n" is 1, but no image was loaded. In version 7.0.7 it works. ImageMagick-7.0.8-12 >magick -script - -print "%n\\n" 1 ImageMagick-7.0.7 >magick -script - -print "%n\\n" 0 see: https://imagemagick.org/discourse-server/viewtopic.php?f=1&t=34778&p=159855#p159855

In you last comment you set a format. My question is now how to reset the format so an identify or verbose identify command will work again? I want to use magick script to reuse the magick process. So the process initialization time can be saved.

Thanks for your explanation.

I test "-print %n" with v7.0.7 and it works, but with v7.0.8 (Portable) it don't work. It is always 1. So maybe it's a bug in v7.0.8.

What possibilities are there to work with image sequences on the command line for "magick -script"? Questions: change the image iterator index (like MagickWand.MagickSetIteratorIndex) get the current iterator index (like MagickWand.MagickGetIteratorIndex) write only the image on the current iterator ...

Yes, you're right the image can be processed. The question is how a program can handle this? Converting this image with magick wand fails with an exception and the convert command fails with error level 1. If the image could be processed then no error should be thrown.

If a tiff file contains the tag 33723 (0x83BB, IPTC) but the count is 0 (zero) then IM report the following exception: convert: test.tif: Null count for "RichTIFFIPTC" (type 4, writecount -3, passcount 1). `_TIFFVSetField' @ error/tiff.c/TIFFErrors/572. A test image can be found here: https://www ...

The method getIteratorIndex return inconsistent values (IM version 6.7.8-6). My Tests - create a new empty MagickWand - load an image - getIteratorIndex -> 1 - create a new empty MagickWand - create a new image - getIteratorIndex -> 0 I test this also with IM version 6.3.2 and both tests return 0. I ...

No one can help or explain this behaviour?

Hello, now I find some time to try the tests again. But my program cannot detect a corrupt jpeg file (premature end of data segment) over the wand api. I test it on Windows XP and Windows 7. My test program was compiled with mingw. The identify command line program can detect the corrupt image. The ...

Forgot my last post. The error is different on a "clean" machine if IM is only copied (no decode delegate for this image format). This means IM can't find the jpg library. I installed IM on some other "clean" machines and I get the same result as on my own computer. The error/warning is not reported ...

I found out why it don't work. If IM is installed during the Windows installer then no error is reported. If only the IM folder is copied to a "clean" (no installed IM) machine then it works fine. I try to rename some registry keys. If the key "HKEY_LOCAL_MACHINE\SOFTWARE\ImageMagick\6.7.3\Q:16" is ...

Thank you for your test. I try it now on three different windows machines. On two it don't work and on one it works like expected. Two machines contains older IM versions in the past and the working node was a clean machine. Have you an explanation for this result?

Thank you

My test images can be downloaded here http://goo.gl/Kg2Bq
You can load a normal jpg image in a hex editor and cut some data from the end of the file.

I test the same with a small C program with the same result. A corrupt jpg file (Premature end of JPEG file) produce no error or warning. This is my test program #include "wand/MagickWand.h" int main() { (void) printf("Copyright: %s\n", GetMagickCopyright()); (void) printf("Features: %s\n ...