[ome-users] BioFormats CZI-Reader reads wrong sublocks of CZI file when an image pyramid was created during acquistion

[Roger Leigh]

On 06/08/14 16:16, Sebastian Rhode wrote:

> I am not sure if this is actually or bug or a feature request. And here is the problem.
>
> In ZEN exits an option to create an image pyramid during acquisition (--> Image_Pyramid_Option.png), which is usually checked, since the user will only see the acquired images right away already during the acquisition, when this option is activated.
> The resulting file has some additional sublocks which contain scaled images (--> WBa.ImagingTool_IMG_PYR.png)
>
> Those sublocks have a scaling which is not = 1.0
> Scaling is derived from StoredSize / Size (in the example = 0.5. Due the 10% overlap for the tiles the size is not 2x the actual image size (640x640 pixel)
>
> CZI-File: 2014_08_06_S=2_T=1_Z=1_CH=2_IMG_PYR.czi
>
> When one tries to open this using BioFormats (--> Bio-Formats Import Options.png) the result is strangle and contains black images. The OME-XML data also contain the wrong dimensions SizeX and SizeY = 1216 (--> IMG_PYR_OME_XML.png)
>
> If the Image Pyramid option is not activated, the CZI data structure is different (--> WBa.ImagingTool_NO_PYR.png). No additional subblocks for the pyramid are present. All blocks have the scale = 1.0 and a size bof 640x640 pixels.
>
> CZI-File: 2014_08_06_S=2_T=1_Z=1_CH=2_NO_PYR.czi
>
> When this file is opened via BioFormats, the import works as expected. The OME-XML are correct (--> NO_PYR_OME_XML.png)
>
>
> The my question is, if it would be possible, that BioFormats can somehow skip those pyramid blocks, which do not a scaling = 1.0? Those pyramid blocks are not always the first ones (inside the WBa.ImagingTool View). Their order might differ, so one needs to filter out those by some other tricks.
>
> The image data and screenshots can be downloaded here:
>
> https://dl.dropboxusercontent.com/u/623476/CZI_Image_Pyramid_Problem.zip

Dear Sebastian,

Thanks for the example images.  I've checked with the new 5.0.3 release
and the problem is also reproducible with it.  This looks like it's a
problem with the CZI reader, which should be able to handle this.  We
haven't had any CZI samples which use this feature to date as far as I'm
aware.  I've opened a ticket for this and CCd you on it:

   http://trac.openmicroscopy.org.uk/ome/ticket/12512


Many thanks,
Roger

--
Dr Roger Leigh -- Open Microscopy Environment
Wellcome Trust Centre for Gene Regulation and Expression,
College of Life Sciences, University of Dundee, Dow Street,
Dundee DD1 5EH Scotland UK   Tel: (01382) 386364

The University of Dundee is a registered Scottish Charity, No: SC015096