BUG: Wrong scaling of big images

[pwalczysko]

During testing of PR ​https://github.com/openmicroscopy/openmicroscopy/pull/634 on Gretzky, user-8 group read-only, project "big images to test PR634".

Image

/ome/data_repo/test_images_good/cellsens/anda/_Image_01_V76 K2 11_lung_/stack1/frame_t.ets [frame_t.ets]

Problems:

1. Web: Being at Scale 100% and clciking once at minus sign in Web goes to Scale:523.53%. Another click on minus brings goes to Scale: 3.03%.

2. Tiles are not corresponding correctly with the birds eye red square position (both Web and Insight). (See screenshots)

3. Insight: First click on minus goes to x0.8 - all the tiles are black. Second click on minus goes to x0.64 - tiles never show up (because the whole image is perched in the topmost leftmost tile, see screenshot). The placement of red square in birds eye is okay when panning is used. When the red square itself is moved, it becomes displaced.

Screenshots: squig- team/screenshots/pwalczysko/08-02-13/PR634

[wmoore]

With the frame_t.ets image above, with this code

        levels = re.getResolutionDescriptions()
        for l in levels:
            print l.sizeX, l.sizeY

I get:

8704    16896
45568   55296
264     512
133     258
710     1377
354     431

• Rendering level 0 gives me 512 X 512 tiles at 100% as expected. ../webclient/render_image_region/1/0/0/?tile=0,0,0,512,512
• Level 1 I get black tiles, with no obvious errors in web or Blitz logs .../render_image_region/1/0/0/?tile=1,0,0,512,512
• Level 2 I get the whole image in a 264 x 512 tile. .../render_image_region/1/0/0/?tile=2,0,0,512,512
• Level 3 I get the whole image at 133 x 258.
• Level 4 I get this error:

  serverStackTrace = ome.conditions.ApiUsageException: Invalid C index: 1/1
  	at loci.formats.FormatTools.getIndex(FormatTools.java:325)
  	at loci.formats.FormatTools.getIndex(FormatTools.java:278)
  	at loci.formats.ChannelSeparator.getIndex(ChannelSeparator.java:258)
  	at ome.io.bioformats.BfPixelsWrapper.getTile(BfPixelsWrapper.java:356)
  

• Level 5 I get the whole image as a mosaic at 354 x 431 pixels

[wmoore]

Not sure if this is a BF issue or something else?

[rleigh]

Using current bioformats develop:

Note that this image does *not* use sub-resolutions, and CellSensReader?.java definitely does not set resolutionCount, so these should all be present as separate series as far as I can tell. Big images should be a result of normal pyramid generation.

% showinf -nopix -noflat "/Users/rleigh/images/test_images_good/cellsens/anda/_Image_01_V76 K2 11_lung_/stack1/frame_t.ets"
SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/Users/rleigh/bioformats/artifacts/slf4j-log4j12-1.5.10.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [file:/Users/rleigh/bioformats/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.
Checking file format [CellSens VSI]
Initializing reader
CellSensReader initializing /Users/rleigh/images/test_images_good/cellsens/anda/_Image_01_V76 K2 11_lung_/stack1/frame_t.ets
Initialization took 0.156s

Reading core metadata
Filename = /Users/rleigh/images/test_images_good/cellsens/anda/_Image_01_V76 K2 11_lung_/stack1/frame_t.ets
Used files:
	/Users/rleigh/images/test_images_good/cellsens/anda/_Image_01_V76 K2 11_lung_/stack1/frame_t.ets
	/Users/rleigh/images/test_images_good/cellsens/anda/_Image_01_V76 K2 11_lung_/stack10001/frame_t.ets
	/Users/rleigh/images/test_images_good/cellsens/anda/Image_01_V76 K2 11_lung.vsi
Series count = 7
Series #0 :
	Image count = 1
	RGB = true (3) 
	Interleaved = true
	Indexed = false (false color)
	Width = 8704
	Height = 16896
	SizeZ = 1
	SizeT = 1
	SizeC = 3 (effectively 1)
	Thumbnail size = 65 x 128
	Endianness = motorola (big)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = false
	-----
	Plane #0 <=> Z 0, C 0, T 0

Series #1 :
	Image count = 1
	RGB = true (3) 
	Interleaved = true
	Indexed = false (false color)
	Width = 45568
	Height = 55296
	SizeZ = 1
	SizeT = 1
	SizeC = 3 (effectively 1)
	Thumbnail size = 105 x 128
	Endianness = motorola (big)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = false
	-----
	Plane #0 <=> Z 0, C 0, T 0

Series #2 :
	Image count = 1
	RGB = true (3) 
	Interleaved = false
	Indexed = false (true color)
	Width = 264
	Height = 512
	SizeZ = 1
	SizeT = 1
	SizeC = 3 (effectively 1)
	Thumbnail size = 66 x 128
	Endianness = intel (little)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = true
	-----
	Plane #0 <=> Z 0, C 0, T 0

Series #3 :
	Image count = 1
	RGB = true (3) 
	Interleaved = false
	Indexed = false (true color)
	Width = 133
	Height = 258
	SizeZ = 1
	SizeT = 1
	SizeC = 3 (effectively 1)
	Thumbnail size = 65 x 128
	Endianness = intel (little)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = true
	-----
	Plane #0 <=> Z 0, C 0, T 0

Series #4 :
	Image count = 1
	RGB = false (1) 
	Interleaved = false
	Indexed = false (true color)
	Width = 710
	Height = 1377
	SizeZ = 1
	SizeT = 1
	SizeC = 1
	Thumbnail size = 65 x 128
	Endianness = intel (little)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = true
	-----
	Plane #0 <=> Z 0, C 0, T 0

Series #5 :
	Image count = 1
	RGB = true (3) 
	Interleaved = false
	Indexed = false (true color)
	Width = 354
	Height = 431
	SizeZ = 1
	SizeT = 1
	SizeC = 3 (effectively 1)
	Thumbnail size = 105 x 128
	Endianness = intel (little)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = true
	-----
	Plane #0 <=> Z 0, C 0, T 0

Series #6 :
	Image count = 1
	RGB = false (1) 
	Interleaved = false
	Indexed = false (true color)
	Width = 37
	Height = 60
	SizeZ = 1
	SizeT = 1
	SizeC = 1
	Thumbnail size = 37 x 60
	Endianness = intel (little)
	Dimension order = XYCZT (uncertain)
	Pixel type = uint8
	Valid bits per pixel = 8
	Metadata complete = true
	Thumbnail series = true
	-----
	Plane #0 <=> Z 0, C 0, T 0


Reading global metadata

Reading series #0 metadata

[mlinkert]

This should have been fixed with ​https://github.com/openmicroscopy/bioformats/pull/369 as Roger noted above. However, if you imported CellSens? (.vsi or .ets) files with a build that did not include that PR and are now trying to view with a current build, you will run into trouble. Re-importing and viewing the new images should work, and result in 7 Images each with 1 resolution.

[pwalczysko]

Seems fine after testing on 15-02-13. Re-test spotted smaller, unrelated issue, reported in ​http://trac.openmicroscopy.org.uk/ome/ticket/10406