Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

imgCIF / CBF (fwd)

Dear Colleagues,

	I am circulating a report on the imgCIF/CBF project that I have
received from Andy Hammersley.  This project started as an attempt to
write a dictionary for storing images from image plates.  It ran afoul of
Comcifs rules because the working group was unanimous that the image
had to be stored in a binary file which contravenes the basic rule of
STAR.

	The compromise that has been worked out is that the standard will
define two files, a Crystallographic Binary File (CBF) and a cif compliant
imgCIF (imageCIF).  The two standards are identical except that the image
itself will be written in binary in CBF and in ASCII in imgCIF.  Thus
translation of a file between these formats will be relatively
straightforward.  Herb Bernstein has been doing noble work in ensuring the
interconvertability of the two files and ensuring cif compliance.  The
binary form was deemed necessary for efficiency in transferring files
(expected to be its primary function) but imgCIF would be used for long
term archiving.  Although designed with image plates in mind, the standard
can be used to transfer or archive any images. 

	As you can see from the report below, we can expect this
dictionary to come for tentative approval during the next few months.  I
will remind you that the msCIF (modulated structures) dictionary is also
expected to be presented for tentative approval at about the same time. 

			Best wishes

				David


More below:
*****************************************************
Dr.I.David Brown,  Professor Emeritus
Brockhouse Institute for Materials Research, 
McMaster University, Hamilton, Ontario, Canada
Tel: 1-(905)-525-9140 ext 24710
Fax: 1-(905)-521-2773
idbrown@mcmaster.ca
*****************************************************

Dear David,

   I asked other members of our informal group for comments on the 
report on the status of imgCIF/CBF. Herb provided an excellent response,
which I'm sure we're all happy with. I include this in its entirety.
Perhaps, I'd just like to emphasise that "imageNCIF is dead, long live
imgCIF". Herb has done an excellent job in providing software to write
images in pure ASCII CIF form. In fact he allows a whole range of encoding
and translation possibilities. Such a file has (I believe) every right 
to be considered a CIF. Thus I suggest that "imgCIF" is presented as a 
fully compliant CIF dictionary.

Best Regards,

            Andy

----------------------------------------------------------------------------
Herb's suggestion:
------------------

Dear David,
  We are pleased to report excellent progress on CBF/imgCIF (the new name
for imgNCIF).  In the course of our discussions, we have come to the
conclusion that there is no reason not to consider our efforts part of
the CIF effort.  On the contrary, if the IUCR and COMCIFS are agreeable,
we would very much like to have CBF/imgCIF be considered part of CIF.  We
are working on our dictionary with an eye to submitting it for COMCIFS
approval in the near future, so that it could be discussed in Glasgow.  We
understand that the binary aspects of our work (CBF) cannot be considered
part of a "true" CIF, but the ascii side (imgCIF) is intended to be fully
CIF compliant, and both sides share exactly the same dictionary of tags,
with the ability for faithful translation in both directions.  We hope
that what we are doing may prove useful for other CIF-based efforts in
need of a reliable way to translate to and from binary representations.
 
We have some more work to do before we submit what we have done to
COMCIFS, but here is where things now stand:

1.  The dictionary has been augmented to cover the tags necessary for
representation of image data.  We still need to add tags to represent
diffraction geometry.

2.  A test-version of an API has been written by Paul Ellis and Herbert
Bernstein.  There are some efficiency improvements needed, and there
are two more compression formats that need to be coded.  Paul is still
looking at some changes recently made by Herbert, so the next version is
not quite ready for distribution, but it can be seen in its current state
(along with the current state of the dictionary) at

           http://www.bernstein-plus-sons.com/software/CBF

3.  We hope to resolve most of the open questions over the next few weeks,
and be ready to submit a dictionary to COMCIFS in February.