[ome-users] VMS file format: incorrect image dimension

[Melissa Linkert]

Hi Matthias,

> The *.vms files do not contain the image dimensions in pixels.
> However, the overall image dimensions can be obtained by looking at
> the dimensions of all JPEG "tile" files (the number of tiles depends
> on the overall image size and is not always 2x2) - e.g. the
> dimensions of CMU-1 are 102400x76288 (wxh). Those JPEG files are -
> in contrast to the NDPI JPEG-images - valid as they are close to,
> but not exceeding the maximum JPEG dimensions. As the scanned area
> is most likely rectangular, not all tiles have a square shape: the
> tiles in edge regions (right & bottom) are also rectangular.
> 
> Going back to the *.vms file: I may found a solution to get the
> overall image dimensions in pixels without having to look at all
> tiles. It seems that the "PhysicalHeight" and "PhysicalWidth"
> metadata elements contain the physical size of the image in
> nanometer. I came to this assumption by dividing the physical width
> and height by the overall width and height in pixels -> The
> width-quotient is 228.198242187 and the height-quotient is
> 227.53125. I think those numbers represent the size of one pixel in
> nanometer in X and Y direction, because they are pretty close to the
> 40x scanning resolution of 0.23 microns/pixel (as stated by the
> manufacturer) - and the CMU test images were acquired at 40x
> magnification. I am not sure if the quotients are equal for every
> scanner, but I am pretty sure that the physical size unit is
> nanometer as converting the "PhysicalMacroWidth" and
> "PhysicalMacroHeight" from nanometers to millimeters gives the size
> of the microscopy slide (76x26 mm).

Thank you for the additional information.  I have opened a ticket for
this on our issue tracking system:

http://trac.openmicroscopy.org.uk/ome/ticket/11535

You have been CC'd, and so will receive an automated email each time the
status of that ticket is updated.

Regards,
-Melissa

On Thu, Oct 10, 2013 at 01:46:06PM +0200, Matthias Baldauf wrote:
> Hello Melissa,
> 
> Thank you for your response!
> 
> > Thank you for the bug report.  I assume that you see the same problem
> > with 5.0.0-beta1 (from
> http://www.openmicroscopy.org/site/products/ome5/bio-formats)?
> 
> I tried to read the CMU files with LOCI Tools 5.0.0-beta1, but the
> problem remains.
> 
> > What would you expect the overall image dimensions to be for CMU-1 (or
> > CMU-2 or CMU-3)?  My understanding for each of those is that there are
> > 2x2 tiles, so 122880x122880 seems to make sense; if that's incorrect
> > though please let us know.
> 
> The *.vms files do not contain the image dimensions in pixels.
> However, the overall image dimensions can be obtained by looking at
> the dimensions of all JPEG "tile" files (the number of tiles depends
> on the overall image size and is not always 2x2) - e.g. the
> dimensions of CMU-1 are 102400x76288 (wxh). Those JPEG files are -
> in contrast to the NDPI JPEG-images - valid as they are close to,
> but not exceeding the maximum JPEG dimensions. As the scanned area
> is most likely rectangular, not all tiles have a square shape: the
> tiles in edge regions (right & bottom) are also rectangular.
> 
> Going back to the *.vms file: I may found a solution to get the
> overall image dimensions in pixels without having to look at all
> tiles. It seems that the "PhysicalHeight" and "PhysicalWidth"
> metadata elements contain the physical size of the image in
> nanometer. I came to this assumption by dividing the physical width
> and height by the overall width and height in pixels -> The
> width-quotient is 228.198242187 and the height-quotient is
> 227.53125. I think those numbers represent the size of one pixel in
> nanometer in X and Y direction, because they are pretty close to the
> 40x scanning resolution of 0.23 microns/pixel (as stated by the
> manufacturer) - and the CMU test images were acquired at 40x
> magnification. I am not sure if the quotients are equal for every
> scanner, but I am pretty sure that the physical size unit is
> nanometer as converting the "PhysicalMacroWidth" and
> "PhysicalMacroHeight" from nanometers to millimeters gives the size
> of the microscopy slide (76x26 mm).
> 
> I hope this information is helpful for you!
> 
> Regards,
> Matthias