Checking the validity of a .dng writer I'm working on and I tried to check as much as I could but I still can't figure out what's triggering this:
[ExifTool] - Warning : Bad IFD1 directory
Here's a link to the dng if you want to take a look:
https://mega.nz/file/ehBwzJ7S#2raYjJNqrk2uOdQfC0DYvmYAjSK4jcz4EDJHnwKlZn8
And is there a command I can put into exiftool to get more detail about warnings like this?
You can use this to see all the warnings and problems
exiftool -g1 -a -s -warning -validate badifd.dng
You can add -v3 (-verbose3) option (https://exiftool.org/exiftool_pod.html#v-NUM--verbose) to get a breakdown of the structure. The -htmlDump option (https://exiftool.org/exiftool_pod.html#htmlDump-OFFSET) will give you an interactive HTML file. See this post (https://exiftool.org/forum/index.php?topic=13920.msg74866#msg74866) for examples.
Phil will probably have to look at it closer but odds are that the IFD1 is mangled such that exiftool can't find any data. Or maybe there's the start of an IFD1 but nothing else.
Yup. That pretty much tells me what I need to know.
Thank you!
You forgot to write an IFD1 pointer at the end of IFD0. If there is no IFD1, then you should write four zero bytes immediately after the last IFD0 directory entry.
- Phil
Copy that sir!
Thank you very much!
Now I just have minor warnings:
[minor] Odd offset for IFD0 tag 0x011a XResolution
but the tiff spec says to align with word boundaries which I don't understand because word lengths are dependent on the architecture, unless judging by your output, do they actually mean even numbered offsets?
"Word" boundary means 2-byte boundary (yes, offsets must be even)
- Phil