1

I am trying to set a metadata copyright notice using exiftool with

exiftool -copyright="©2016 Jim Berry" *.JPG

but it comes out in the images as "Copyright=©2016 Jim Berry" Why the "Â" and how can I avoid it?

jimberry
  • 61
  • 3
  • For some reason you're having a character set encoding issue. I believe that © is how © would look when converted from UTF-8 to ascii. I've seen this problem using Exiftool under Windows, but from what I understand it shouldn't happen in Ubuntu. You might also check [Exiftool FAQ 10](http://www.sno.phy.queensu.ca/~phil/exiftool/faq.html#Q10) to see if you can find a solution. What program are you using to view the copyright? – StarGeek Jul 31 '16 at 06:39
  • I am using a Beanshell tool in the java program jAlbum to view the metadata. I believe this uses Drew Noakes metadata extractor library. I have discovered that if I place the copyright notice in a XMP rights field it shows up OK - xmp.dc:rights[1]=©2016 Jim Berry - so this might be a better option as XMP seems to be considered "the way of the future" – jimberry Aug 01 '16 at 09:07

1 Answers1

1

Here's what I believe is happening, based upon Exiftool FAQ 10.

Metadata tags in the EXIF group are often stored in ASCII but for most (including the Copyright tag) there is no standard for how the characters are encoded. The Metadata Working Group recommends that the data be encoded in UTF-8 but not all tools do this. Exiftool is very flexible and has options to encode it however you like. But it's default is to encode it how it receives it. In this case, your shell character set is probably UTF-8 based character set and that's how the tag is encoded when written.

The Beanshell tool, however, is interpreting the tag as ASCII, which is why the copyright symbol is showing up as ©.

If you still want to use the Copyright tag, then I think you can use this command:
exiftool -L -copyright="©2016 Jim Berry" File
The -L option is a shortcut to make Exiftool use Latin, Latin1, and cp1252 character sets. I think you should be able to set the XMP:Rights tag in the same command without problems, as XMP tags are supposed to be written as UTF-8 anyway. But test it out to be sure.

Just for completeness, there is also the IPTC:CopyrightNotice tag. If you write this tag, it's best if you add -codedcharacterset=utf8 as well to force the IPTC:CopyrightNotice to be written as UTF-8.

StarGeek
  • 463
  • 3
  • 7
  • Thanks for that . I guessed that was the case, but didn't follow up since I was able to use XMP for the purpose. But on reflection, I might follow your advice and write to Exif and IPTC copyright as well, with your suggested options to ensure UTF compliance. – jimberry Aug 04 '16 at 08:27
  • You might also look into the `-use MWG` option. This makes it easier to keep EXIF, IPTC, and XMP tags in sync, as you only have to write one tag name and it will write to the appropriate tag in each group. See [MWG Tags](http://www.sno.phy.queensu.ca/~phil/exiftool/TagNames/MWG.html). In this case, you could use `exiftool -use mwg -Copyright='©2016 Jim Berry' -L -codedcharacterset=utf8 FILE` to write all three tags and they should be properly encoded, though test to make sure. – StarGeek Aug 04 '16 at 23:15
  • I got the result I needed with exiftool -xmp:rights="©2016 - Jim Berry" -ext jpg metadatatest but after trying the "3in1" option there were odd extra characters in all three tags This might be an issue with the "metadata extractor" library being used to display the metadata, since listing the tags with exiftool displays the tags as expected. Until I can sort out the problem with"metadata extractor" I can get by without syncing the Exif and IPTC tags. – jimberry Aug 06 '16 at 14:16