Lookup tables are not correctly applied if there is one LUT per channel

[melissa]

See for example the first series of data/leica/joel/sequential scan 2/sequential scan 2.lei. There are two channels; the first should have a red LUT, the second should have a grey LUT (you can see this using 'showinf').

However, the ImageJ import plugin applies the red LUT to both channels when 'Color mode' is set to 'Default'. Previous stable releases correctly applied the grey LUT to the second channel.

[curtis]

The "sequential scan 2" dataset reports that isIndexed=false, but has an 8-bit color lookup table:

	Indexed = false (false color, 8-bit LUT: 3 x 256)

Is this by design, or is this a bug?

[curtis]

Replying to curtis:

Indexed = false (false color, 8-bit LUT: 3 x 256)

Strangely, on another one of my systems, with r6481, the isIndexed flag returns true:

	Indexed = true (false color, 8-bit LUT: 3 x 256)

I'll double-check the first system's result tomorrow, just to be sure.

[curtis]

It may be that isIndexed=false when using the "-nometa" flag with showinf. This is related to #496.

[melissa]

(In [6500]) Apply one LUT to each channel if the original file contains its own LUTs. Closes #495.

[melissa]

This bug was reintroduced in r6554 (trunk) and r6555 (4.2 branch).

[curtis]

(In [6598]) Preserve channel-specific LUTs. Fixes #495 (again).

[melissa]

Milestone bio-formats-4.2 deleted