Bio-formats *.vsi Olympus cellSens bug

Dear OME-team,

we recently bought a new Olympus spinning disk system which stores its data in the *.vsi file format as written by the cellSens Dimension 2.3 software.

These files readily open with Bio-formats, but I noticed that a few metadata fields were not or incorrectly populated. I have hence uploaded an example file including an export of the original (aka “correct”) metadata as exported directly from the cellSens software itself for comparison.

  1. When you open the image in Fiji with Bio-formats, the voxel depth is reported as 1, or maybe in other words it was not found. It should be 0.11 um.
  • “PositionZ” in the ome-xml for each plane is correct, but a field for the z interval is missing.
  • In the “Original metadata”-file as opened by Bio-formats, I did however find the correct value in line 489:
    405, 561 Z incrementValue #1 0.11
  1. The exposure time for C1 is not correct (80ms written, but should be 100ms), but for C2 it is correct (80ms).
  • In the “Original metadata”-file as opened by Bio-formats, I did however find the correct values in lines 369 & 370:
    405, 561 Microscope Exposure time (microseconds) #1 100000
    405, 561 Microscope Exposure time (microseconds) #2 80000

I hope this helps and that the issue can be fixed :slight_smile:.

Thank you for your help and cheers,


Hi @CellKai, thank you for providing the report and sample file. I have opened a Trello card to track the issue. Hopefully we should be able to address this as the correct values are already being parsed as part of the Original Metadata.

1 Like

Hi @dgault,
thanks for your reply and the Trello card!
We really appreciate the hard work on Bio-Formats :slight_smile:


I actually have an issue opening vsi zstack files using bio-formats package 6.5.0 with ImageJ 1.53b. On both stations CellSens is installed (2 different microscopes), with Bio-Formats, the properties give me a Voxel depth remains at 1, regardless of the original z step value. In the Metadata, “Z incrementValue #1” is at the right value : 0.2
To finish, OlympusViewer plugin open zstack with the right value.

Is that a know issue or is there any solution ?

Thank you for your impressive work !


Hi Thomas, this sounds very much like the same issue reported in the original post. Unfortunately we don’t yet have a fix released to address it, I will update the existing Trello card to note your bug report. If you have any sample files showing the issue which can be used for testing then please upload them to