Luts

[will-moore]

This adds LUTs support to OMERO.figure.

To test:

• click on channel color chooser drop-down and select "Lookup Tables".
• This will open a dialog showing available LUTs
• Click a LUT to preview it, then OK
• Channel button AND slider should have LUT background.
• This should also work to set the channel LUT when multiple images are selected, even if they have different LUTs/colors for that channel before setting LUT.
• Export to PDF/TIFF should export images with LUTs applied as in the figure.

[gusferguson]

@will-moore

Tested with https://cowfish.openmicroscopy.org/webmerge/webclient/ user-3 read-only 1 Figure: Test 16-08-04 08:48:09

Selection of colours, LUT and clour picker behave as expected.

Question: @will-moore @jburel

• using the left panel in the test figure I configured the top channel as blue, the second channel as sepia LUT and the third channel as green
• when I select channel 1 and 3 with 2 deselected, the rendering is as expected - screenshot 1
• but as soon as I select channel 2 (the sepia LUT) it seems to obscure channel 1 but channel 3 is visible - screenshot 2
• if I make channel 2 a normal colour and channel 3 a LUT (grey in this screenshot) both channels 1 and 2 are not visible at all - screenshot 3
• if I put the LUT on channel 1 then I can see all 3 channels - screenshot 4

It looks to me as if selecting an LUT for a channel causes all channels “below” it to be obscured - the graphics app metaphor of this would be - normal colour channels have a “transparent background” but the LUTs don’t.

I am getting the same effect with LUTs with the same image in the main Web app - although on some other images there seems to be a small degree of "show through" below LUTs.

Is this expected?

[gusferguson]

@will-moore @jburel - LUTs in Insight also giving this obscuring effect.

[jburel]

same comment here https://github.com/openmicroscopy/openmicroscopy/pull/4764#issuecomment-237485670

[jburel]

Alpha is not taken into account when applying LUT (rendering engine), we may need to review that part

[jburel]

@gusferguson: https://github.com/openmicroscopy/openmicroscopy/pull/4753/commits/d57c42e21cd1978c480c0ab64306613f7ec2480e fixes the "override" issue mentioned above

[will-moore]

To test today's commits:

• Names of LUTs in picker dialog should not have '.lut', underscores should be replaced by space and name should start with capital letter.
• If you Add a new image into OMERO.figure, any LUTs saved on that image from Insight/webclient are shown immediately.
• Layout of channel sliders is fixed (as it was before this PR, instead of start/end labels squashed onto different rows).

[snoopycrimecop]

Conflicting PR. Removed from build FIGURE-merge#214. See the console output for more details. Possible conflicts:

• PR #152 will-moore 'open with images'
  • static/figure/figure.js
  • static/figure/js/views/modal_views.js
• PR #158 dominikl 'Label markdown'
  • static/figure/figure.js
• Upstream changes
  • static/figure/figure.js

[dominikl]

The channel color buttons look a bit strange (black vertical line in the center).

But I don't have a proper local setup (no lookup tables available), might be due to that.

[dominikl]

Ignore the last comment, I guess that's just because I can't get my local server properly setup for lookup tables.

[snoopycrimecop]

Conflicting PR. Removed from build FIGURE-merge#215. See the console output for more details. Possible conflicts:

• PR #152 will-moore 'open with images'
  • static/figure/figure.js
  • static/figure/js/views/modal_views.js
• PR #158 dominikl 'Label markdown'
  • static/figure/figure.js
• Upstream changes
  • static/figure/figure.js

[snoopycrimecop]

Conflicting PR. Removed from build FIGURE-merge#216. See the console output for more details. Possible conflicts:

• PR #152 will-moore 'open with images'
  • static/figure/figure.js
  • static/figure/js/views/modal_views.js
• PR #158 dominikl 'Label markdown'
  • static/figure/figure.js
• Upstream changes
  • static/figure/figure.js

[snoopycrimecop]

Conflicting PR. Removed from build FIGURE-merge#217. See the console output for more details. Possible conflicts:

• PR #152 will-moore 'open with images'
  • static/figure/figure.js
  • static/figure/js/views/modal_views.js
• PR #158 dominikl 'Label markdown'
  • static/figure/figure.js
• Upstream changes
  • static/figure/figure.js

[snoopycrimecop]

Conflicting PR. Removed from build FIGURE-merge#221. See the console output for more details. Possible conflicts:

• PR #158 dominikl 'Label markdown'
  • static/figure/figure.js
• Upstream changes
  • static/figure/figure.js
  • static/figure/js/views/modal_views.js

[jburel]

Having the list of luts hard-coded in 2 different files is far from ideal.

[will-moore]

@jburel: The list of LUTs is only hard-coded in one place. Remember that every commit for figure includes the concatenated JavaScript so every change is duplicated.

[jburel]

true, forget my comment

[gusferguson]

@will-moore

Tested using https://cowfish.openmicroscopy.org/webmerge/figure/new/ Safari user-3

LUTs behaving as expected.

Good to merge.

[jburel]

Thanks merging so we have everything in place