Importing .nd2 files results in horizontal lines for channels

import
channels
fiji
bio-formats
Tags: #<Tag:0x00007fb879f71c30> #<Tag:0x00007fb879f71af0> #<Tag:0x00007fb879f71988> #<Tag:0x00007fb879f71848>

#1

Hey all,

I’m having an issue with opening some .nd2 files in Fiji. When I import them, the image shows a weird striped pattern where every channel is shown as separate horizontal lines that have the lines from other channels in-between them.

You can see the pattern appearing in this zoom-in. It should be a 5 channel composite (4 fluorescent + brightfield), but here the channels appear as lines. When only looking at one channel, the other lines disappear and there are 4 black lines instead.

These images open as they should in NIS-Elements viewer. Moreover, during the same imaging session I acquired other images that open up in Fiji without any problems. The only differences are that they have only 2 channels and more tiles.
I updated my Fiji to the latest version (v1.51o) and also tried to open the file on another computer, but that didn’t help. I also tried different Bio-Formats importing options but couldn’t find anything that would have resolved the issue.

I would greatly appreciate any help or suggestions you could give.
You can find an image that gives this issue here: https://www.dropbox.com/s/6s4asb46mpduf12/sample_image.nd2?dl=0

Thanks,
Karl


#2

Dear @Karl,

this sounds like a Bio-Formats issue with a (recent?) change in the ND2 format. Thanks for providing a sample file of the issue!

Maybe @s.besson can take it from here?

Best,
Stefan


#3

HI @Karl,

thanks for the bug report and the corresponding sample file. The issue seems to be related with the incorrect detection of the scanline padding when trying to open the individual planes.

We do not have an immediate workaround at the moment and fixing ND2 bugs without affecting the Nikon ND2 samples in our data repository usually turns out to be challenging. In the meantime, we have captured the issue in the following card.

Best,
Sebastien


#4

Thanks for your help and submitting the bug report!

In the meanwhile I have a (cumbersome) workaround by converting the files to .tif in NIS-Elements and then using the .tif files for further analysis.

Best,
Karl


#5

Hi @Karl,

thanks for letting us know of the workaround. Following up on the original problem, following an assessment of our ND2 samples, we are investigating a fix for the scanline issue you reported.

We would like to include the ND2 sample file you have linked to in your initial post to our data repository for non-regression testing. Would this file be suitable to be made publicly available under a CC-BY license to simplify the maintenance of this reader? Or should it remain private?

Best,
Sebastien


#6

Hey!

Sorry for the late reply.
You can make the file available.

And thanks for all the help!

Karl


#7

Hi Karl,

thanks for your input. The recent Bio-Formats 5.5.3 release should have addressed this particular bug.

We have also made your sample image publicly available to the community as mentioned above.

Best,
Sebastien