Bug: channel names mixed up deconvolved Deltavision files

[mlinkert]

See ​https://www.openmicroscopy.org/community/posting.php?mode=reply&f=13&t=7772&sid=4f8165e1388271b6772b276a9c604596#pr15538 and QA 10945/10946.

If both .dv files are put in the same directory with a file named HCC1143_BEZ_LaminACB1_1_R3D.dv.log that contains:

EM filter: channel 1
Stage coordinates: 0
EM filter: channel 2
Stage coordinates: 0
EM filter: channel 3
Stage coordinates: 0
EM filter: channel 4

'inf -omexml' on the *R3D.dv file should show that each channel has a name set, and that the names are in the order defined by the log file. 'inf -omemxl' on the *R3D_D3D.dv file, however, shows only the first channel having a name, and that it's set to 'channel 4'.

[dlindner]

Did some debugging... The problem is, that bioformats determines the order "XYCZT" for the *R3D.dv image, but "XYZTC" for the *R3D_D3D.dv image. Therefore getZCTCoords() (DeltavisionReader?.java:1133) always returns zero for the channel index (i.e. on each iteration the name of the first channel is set, finally to name 'channel 4'). Unfortunately I don't know if the order is wrongly detected or if the code in getZCTCoords() method is wrong.

[mlinkert]

Fixed with: ​https://github.com/openmicroscopy/bioformats/pull/1735

[Melissa Linkert <melissa@…>]

(In [e670a5f3de86b35e9e7797f7be50c2fcba03cb5d/bioformats.git] on branch develop) Deltavision: fix channel name order for deconvolved data

Fixes #12807.