[ome-users] Problem loading lambda-scans from .oir files

[David Gault (Staff)]

Hi Eric,

Thank you very much for providing the detailed report and sample file.
I have been able to reproduce the issue exactly as described and have been able to do some further debugging on the problem.
This is a bug in the Bio-Formats OIRReader and I have opened a Trello card on the Bio-Formats inbox with details to track the issue: https://trello.com/c/D6wZog99/252-oir-duplicate-planes

With Thanks,
David Gault

On 22 Jun 2018, at 18:25, Eric Kercher <kercher.e at husky.neu.edu<mailto:kercher.e at husky.neu.edu>> wrote:

I'm trying to load some lambda scans taken with an Olympus microscope (.oir file) and ran into a potential bug.

Bio-formats recognizes the correct number of channels, but loads the first channel of the image for each subsequent plane. This occurs in both imageJ and Matlab. In Matlab, I traced the issue to line 78 in the bfGetPlane.m file:

plane = r.openBytes(...
    ip.Results.iPlane - 1, ip.Results.x - 1, ip.Results.y - 1, ...
    ip.Results.width, ip.Results.height);

r.openBytes loads identical vectors for 'plane' through each iteration of the loop in bfopen.m.

I've reproduced this problem using the BF matlab toolbox and imageJ plugins on both windows (Windows server 2016) and Mac (macOS High Sierra 10.13.5); each system has the latest versions of BF (5.8.2) and Java (10.0.1).

This script reproduces the bug:

% Read data
impath = 'path/to/lambda/scan.oir';
data = bfOpen3DVolume(impath);
im = data{1,1}{1,1};
% Compare channels
isequal(im(:,:,1), im(:,:,2))
plot(squeeze(mean(mean(im))));

isequal returns true for any 2 channels you pick; the plot displays the full z-axis profile which is a flat line. In ImageJ, Image>Stacks>Plot Z-axis Profile also produces the same flat line. I also tried loading the image using each stack configuration option (XYCTZ, XYZCT, etc...) with no luck.

The OlympusViewer plugin for ImageJ (not the BF one) loads the lambda scans correctly on the Mac & Windows systems described above. I also downloaded the three z-stacks from the OME repository and those load correctly as well. I searched through the metadata (comparing z-scan and lambda-scan) but found nothing out of the ordinary; so I don't know if its an issue with my data files or if its an actual bug.

I've uploaded a sample .oir lambda scan file (exampleLambdaScan_AF546soln.oir) for reproduction through the OME QA uploader. In the meantime, please let me know if I missed anything. Thank you and happy to provide more information as needed.

Regards,
Eric Kercher
_______________________________________________
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/20180625/6c87c33b/attachment.html>