Page 1 of 1

more issues with czi files

PostPosted: Wed Jun 04, 2014 4:15 am
by ebbwdan
Hi

Sometime ago I was having an issue opening large, tiled images acquired and stitched in Zen and saved as CZI which bioformats 5.0.2 seems to have addressed. In the time that these fixes were being incorporated into 5.0.2 Zen has been updated and this seems to have broken something else. Quite difficult to describe, but files generated by the newest Zen Blue essentially don't open at all with bioformats 5.0.2. Bioformats opens a series options dialog where the displayed list is completely messed up. I gave it a tiled image with 4 channels but the series options shows 6 series with 2 planes and 2 channels.

I have uploaded a file from this latest Zen Blue version, a snap-shot of the Zen version numbers and an example from a slightly older version of Zen Blue which now works.

Cheers,

Dan.

Re: more issues with czi files

PostPosted: Wed Jun 04, 2014 4:53 am
by jmoore
Thanks, Dan, & sorry for your troubles. Most of the team is traveling this week to the annual users' meeting in Paris. We'll take a look at the file and get back to you asap.

Cheers,
~Josh.

Re: more issues with czi files

PostPosted: Mon Jun 09, 2014 11:22 pm
by mlinkert
Hi Dan,

Thank you for reporting this problem. Unfortunately, it looks like only the working file was uploaded successfully (IP-Stitching-01.czi, corresponding to QA #9267). That file opens as 6 tiles with 2 channels using 5.0.2, which matches what I see in Zen 2012.

If you could please try to upload the file that doesn't work again, that would be much appreciated. I have sent alternate upload instructions in a private message, in case the QA system can't handle the file for whatever reason.

Regards,
-Melissa

Re: more issues with czi files

PostPosted: Tue Jun 10, 2014 1:56 am
by ebbwdan
Hi Melissa,

Thanks for taking a look at this. My sincere apologies, the bug fix does in fact work for stitched, fused datasets. My mistake was not choosing 'fuse' in Zen before trying to import in Bioformats. In the case where the data is not fused, Bioformats gets the number of series and channels correct. However, when opening these unfused datasets some of the series do not open correctly. I have uploaded as TIF's the stacks generated by trying to open the unfused series: only series 3 and 6 open correctly. Also, trying to use the 'stitch tiles' option in the Bioformats importer causes an exception.

Thanks again!

Dan.