[ImageJ-devel] Serious Bug in 'Plot Profile' coordinates

[Gabriel Landini]

On Thursday 29 Aug 2013 08:11:03 Thiel, Bjoern wrote:
> Analyzing an image of 3.9 x 3.9 microns (130 x 130 pixels) with a profile
> plot I get false coordinates indicating only integer distances (of e.g. 2.0
> microns) while the plots seem to range over fractional distances (of e.g.
> 2.4 microns).

Very odd. Pixel width is 0.0300000... and indeed the plot shows only an 
integer maximum values in live mode. This is not the case if you create a new 
image and assign a calibration to it.

Note that if I remove the scale from your image:

run("Set Scale...", "distance=0 known=0 pixel=1 unit=pixel");

and then add it again:

run("Properties...", "channels=1 slices=1 frames=1 unit=pixel pixel_width=0.03 
pixel_height=0.03 voxel_depth=0.03 frame=[0 sec] origin=0,0");

the scale works as expected.

Also if I create an image, assign that scaling save it and load it again it 
still works fine. So far this seems to happen with that image only. I do not 
know what is exactly going on, for sure Wayne would nail this bug.

Cheers

Gabriel