Hi Leonard
The current LibTIFF can be easily updated support an additional TIFF field with
ID 52545 for the C2PA manifest data block for an image.
But then, the second use case with multiple images, with the extra IFD having
C2PA field only will still not be supported.
Did you check that an IFD with a single TIFF field is also compatible to DNG
and TIFF/EP specifications?
- C2PA specification says: “The Digital Negative or DNG format
provides camera manufacturers to provide their camera raw formats in a
standardized manner. DNG is based on which is based on TIFF/EP”
- E.g. TIFF/EP (ISO/DIS 12234-2) defines mandatory fields that have to
be within a valid IFD. It says: “All mandatory tags shall be included in the
IFD, since default values are not allowed in TIFF/EP, even in cases where
default values are given in TIFF 6.0.”
My second concern is the requirements about location of the C2PA field and its
manifest data, or update of manifest data:
- C2PA specification says: “To support update manifests, the C2PA
Manifest Store should be located at the end of the file; this ensures that
changes to the size of the C2PA Manifest Store do not impact any of the other
tag offsets.”
- LibTIFF does not have any interface (nor the possibility) to
guarantee the location of IFD data and its associated data within the file.
- For an update of IFD field contents (size or offset of the
associated data) the IFD has to be re-written, which might change its location
in the file.
Regards
Su
Von: Tiff [mailto:[email protected]] Im Auftrag von Kemp Watson via
Tiff
Gesendet: Montag, 16. März 2026 19:27
An: Leonard Rosenthol
Cc: [email protected]
Betreff: Re: [Tiff] Update: Content Credentials and TIFF
Not sure I care for that solution for single-image TIFFs... seems to be
"special-casing" right off the bat...
>From what I can read, maybe the spec needs clarifications such as
- in which IFD TIFFtag (52545) goes relative to the data it points to (must be
in same IFD in my opinion)
- whether data is in an IFD as "raw/image data" or is pure raw data in some
other container format at end of file
W. Kemp Watson, Managing Director
Objective Pathology Services Limited
Toronto, Canada
On Sun, 15 Mar 2026 at 10:31, Leonard Rosenthol via Tiff mailto:[email protected]> > wrote:
Sorry to have taken so long to respond to folks….
We’ve been in discussions in the C2PA about what (if any) changes we should
make to our specification to better align with libTIFF and the TIFF community
in general. Here is what we believe is the correct answer - we would welcome
your feedback.
As previously mentioned by me and also mentioned in our specification, the C2PA
believes that there should only be a single Content Credential in a given TIFF
file - regardless of how many actual images may be present. This is consistent
with all other formats that may contain multiple assets/images - APNG, JPEG XL,
HEIC, etc.
It is for this reason that we put our data into our own IFD, rather than
associating it with some other (arbitrary) IFD. Based on our reading of the
TIFF v6 spec, and in consultation with the maintainers of TIFF at Adobe, we
consider this a valid position.
However, in the case where a TIFF contains only a single image (and thus a
single top-level IFD), we are going to change our spec to have the C2PA
Manifest’s IFD be a sub-IFD to that single image - since that makes it clear
that it is associated with that image. That will also make things more
compatible with libTIFF for the most common scenarios.
Before we finalize this decision for our next revision, we would welcome your
input.
Thanks,
Leonard
___
Tiff mailing list
[email protected] <mailto:[email protected]>
https://lists.osgeo.org/mailman/listinfo/tiff
___
Tiff mailing list
[email protected]
https://lists.osgeo.org/mailman/listinfo/tiff
