Problem opening large stitched CZI tile image

[ebbwdan]

Hi

I am having a problem opening stitched tile CZI images using Bioformats (through FIJI). I made a tile region in Zen Blue which ended up with a file size around 800mb. I stitched and fused the image in Zen but when I try opening the file in FIJI the tiles appear jumbled. This does not happen when the tile region is smaller: I successfully stitched, fused and opened a 300mb file with no problems.

Cheers,

Dan

[rleigh]

Dear Dan,

It's possible that this type of tiling isn't something we have seen before, and so don't yet support. If it would be possible to send us a copy of the file in question, we can investigate this in more detail. You can upload files to us using this web page: https://www.openmicroscopy.org/qa2/qa/upload/


Many thanks,
Roger Leigh

[rleigh]

Many thanks for uploading the file. We'll investigate this in more detail and let you know what we find.

[rleigh]

Dear Dan,

After further investigating the issue, I have been able to reproduce the problem you saw with FiJi and I have opened a ticket to track this issue (https://trac.openmicroscopy.org.uk/ome/ticket/11784). You should be CCd on any changes to the ticket. It looks like there are a few separate problems here:

- The OME-XML metadata in the OME-TIFF exported from Zeiss ZEN Blue is invalid; this is a bug in Zeiss ZEN Blue's OME-TIFF export.
- The original CZI file is fine, but there's a bug in Bio-Formats which is causing the pixel data to be read incorrectly; this above ticket is for fixing this; it's probably relating to the large plane size. From the look of things, it's likely related to miscalculating offsets.


Kind regards,
Roger Leigh

[ebbwdan]

Hi Roger

Many thanks for looking into this for me!

I would like to notify Zeiss about the issues with Zen and was wondering if would be possible to get a technical explanation of the problems with the OME-XML data. Or do you have a mechanism in place for feeding back this kind of information? If not, then would it be possible to get this information offline so that I can approach our Zeiss area manager with this? Although the image opens correctly in Fiji, the broken meta-data is stopping images like this being imported into our Omero server (reported to the Omero team separately) and so I'd really like to get this fixed up.

Also, I just took a look at the ticket but there are no time-scales. How long do you think it will take to fix the Bioformats bug?

Best wishes

Dan.

[ajpatterson]

Hello Dan,

I have looked at the OME-TIFF files produced from three versions of Zen.

Zen 2012 Lite (Free download from Zeiss website)
Version: 1.1.1.0
File version: 1.1.13064.302

Zen 2012 (Dan's version)
Version: 1.1.1.0
File version: 1.1.13064.302
Hotfix: 1.1.1.2

Zen 2011 (Available in Dundee)

The OME-TIFF files produced by these were validated and test imported into OMERO.

Zen 2012 Lite - This file has two problems:
1. There is no schema location defined. This line should appear as an attribute of the OME node

Code: Select all

xsi:schemaLocation="http://www.openmicroscopy.org/Schemas/OME/2010-06
   http://www.openmicroscopy.org/Schemas/OME/2010-06/ome.xsd"


2. The LightSource node has no content. It should contain either an Arc, Filament, Laser or LightEmittingDiode child element. This is what is causing the file to be invalid.

Zen 2012 - This file has two problems:
1. There is no schema location defined. This line should appear as an attribute of the OME node

Code: Select all

xsi:schemaLocation="http://www.openmicroscopy.org/Schemas/OME/2010-06
   http://www.openmicroscopy.org/Schemas/OME/2010-06/ome.xsd"


2. The LightSource node has no content. It should contain either an Arc, Filament, Laser or LightEmittingDiode child element. This is what is causing the file to be invalid.

Zen 2011 - This file has one problem
1. The Pixels nodes use a Type of "uint12", which does not exist. Allowed values are int8, int16, int32, uint8, uint16, uint32, float, bit, double, complex and double-complex.
Note: this version does correctly set xsi:schemaLocation and it does not have any LightSource or LightSourceSettings nodes.

Could pass this on to your local Zeiss contact as the fix will need to be to the Zen software.
We will pass the same information on from here.

If they need further information please give them my contact details.

Thanks,

Andrew