Nikon ND2 file import issue

[dsudar]

Hi Melissa,

We just installed a Nikon system that uses NIS Elements and is meant to be used for high content screening. One initial acquisition of a screen resulted in a directory of nd2 files for each well (still figuring out how to make Elements treat the whole plate as a single experiment). Anyway, we acquired 4 fields per well and in OMERO only the first field shows up (omero 5.0.2), using the command line importer. I'm uploading the file for Well A1 to the QA system (#9438).
Fiji's BioFormats importer (5.0.2-DEV - 29 July 2014) interprets the file as having 4 time points. Also not correct but better than ignoring the 3 fields. Will 5.0.3 have better handling of multi-location ND2's?

Second problem: for the "Pixels Size (XYZ) (µm)" only the X value is interpreted correctly and it shows 0.0 for Y. Maybe the Y value was missing from the file?? Would it be acceptable kludge for the importer to assume the same pixel size value for Y as for X if that value is not in the ND2 file?

Thanks,
- Damir

[mlinkert]

Hi Damir,

Thank you very much for reporting this problem. 5.0.3 will have improved ND2 support, but unfortunately will not fully resolve this problem. I have filed a ticket for 5.0.4 and CC'd you:

http://trac.openmicroscopy.org.uk/ome/ticket/12508

If you prefer not to receive notifications, please let us know.

-Melissa

[dsudar]

Hi Melissa,

With 5.0.3 the ND2 handling is indeed vastly improved and my multi-position files now show up as time series. And that actually works fine as long as I keep that in mind. Thanks.

Now unfortunately there are a few more issues with ND2 handling and I really don't envy you trying to reverse engineer such a completely opaque format. I (and I hope everyone else on this forum) will really come down hard on Nikon for not making ND2 an open format or at the very least to provide you with the full and complete documentation on the format. After all the files contain OUR DATA so we should have full access to it.

Anyway, the issues I have are with metadata handling:
1) in the file that I uploaded (#9438) and all my other ND2 files the exposure times are not being interpreted. In OMERO.Web it shows "None s" and in Insight it shows: "--". I imagine the exposure times are somewhere in the file ....?
2) A few valuable fields seem to be in the "Original metadata", such as Camera Name, Objective, etc. but the appropriate fields in "Microscope", "Image", and per-channel info do not seem to be populated.

When looking at this file with Nikon's ND2 ImageJ reader it shows some limited metadata as:

pnt=#1 (44868.00, -5126.00, 0.00) um

Wave:470.00nm(BLUE)
Obj:Plan Fluor ELWD 20x Ph1 DM,mag=-1.00,NA=0.45,RI=1.00
Calibration:0.325 microns/pixel

and "Wave: 520.00nm (GREEN)", "Wave:610.00nm(RED)", and "Wave:700.00nm(FAR RED)" for the other channels.

Anything I can do to help improve the metadata extraction?

Thanks,
- Damir

[mlinkert]

Hi Damir,

Thank you for the update - I'm glad to hear that 5.0.3 is a little better.

With 5.0.3 the ND2 handling is indeed vastly improved and my multi-position files now show up as time series. And that actually works fine as long as I keep that in mind.

We are still planning to fix the position count in 5.0.4, we just didn't have an appropriate fix in time for 5.0.3.

1) in the file that I uploaded (#9438) and all my other ND2 files the exposure times are not being interpreted. In OMERO.Web it shows "None s" and in Insight it shows: "--". I imagine the exposure times are somewhere in the file ....?
2) A few valuable fields seem to be in the "Original metadata", such as Camera Name, Objective, etc. but the appropriate fields in "Microscope", "Image", and per-channel info do not seem to be populated.

That's definitely on our radar for 5.0.4, which should be out later this summer. I can see the expected exposure times etc. in the NIS Elements viewer as well, so this shouldn't be too difficult to fix.

-Melissa