[ome-users] Bio-Formats Import OME-XML metadata timestamp problem since 5.2.0

[Sebastien Besson (Staff)]

Dear George,

thank you very much for bringing this issue to our attention. We were able to reproduce the problem using the test data you uploaded.

This change has been introduced as part of a 5.2.0 PR [1] and was aiming at reading and importing invalid OME-TIFF data. It turns out the assumptions upon which these modifications were relying  have undesired effects for valid OME-TIFF.

As this constitutes a regression from a previously functional behavior, we will be working on a fix imminently and hope to release a Bio-Formats 5.2.4 next week to address the problem.

Best regards,
Sebastien

[1] https://github.com/openmicroscopy/bioformats/pull/2343
[2] https://trello.com/c/aTV56qEX/6-fix-ome-tiff-regression-plane-population

On 12 Oct 2016, at 21:23, George Patterson <geopascway at gmail.com<mailto:geopascway at gmail.com>> wrote:

Hi all,
After a recent FIJI update, we have encountered a problem with the metadata from our experiments using Micromanager.  In these experiments, we collect three channel data, but for two channels the images are collected only every 10th time point.  We've been running these type of experiments for some time now and only recently found that the timestamps do not match the image times.  If we display the OME-XML metadata when importing using Bio-Formats and look at the Plane information located at the bottom, the metadata shows all three channels taken sequentially for the whole experiment.   Because the two skipped channels are listed as if imaged at every time point, the DeltaT values are correlated with the wrong channels.
I can reproduce this with a small data set collected using the demo mode of Micromanager.  I'll upload it to the QA system shortly.  For this example, Channel 0 (1 sec exp), Channel 1(0.1 sec exp), and Channel 2 (0.01 sec exp) are imaged, then Channels 0 and 1 are skipped for 9 time points.  This is repeated for 3 cycles for a total of 3 images each of Channels 0 and 1 and 30 images of Channel 2.
My colleague has traced this change to occur with the release of Bio-Formats 5.2.0.  I hesitate to refer to this as a bug since it looks like numerous changes occurred between 5.1.10 and 5.2.0 and this could just be a feature change that we are not using properly.
We did our tests of the different versions of Bio-Formats using ImageJ 1.51a, but as I mentioned earlier we initially discovered it using an updated FIJI.
Computers used are a Mac Pro running 10.9.5 and Macbook Pro running 10.10.5.
My apologies if I missed the discussion of this issue on the listServs.  If you could direct me there I'd appreciate it.
Thanks for any clarification if it's a feature and any help if it's a bug.
And as always, thanks to OME and Bio-Formats teams for sharing.
Best,
George
_______________________________________________
ome-users mailing list
ome-users at lists.openmicroscopy.org.uk<mailto:ome-users at lists.openmicroscopy.org.uk>
http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-users


The University of Dundee is a registered Scottish Charity, No: SC015096
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openmicroscopy.org.uk/pipermail/ome-users/attachments/20161013/fcae3ef5/attachment.html>