Login  Register

Re: I may have found the cause of my problem stitching stacks with 5 planes...

Posted by Pariksheet Nanda on Jan 06, 2014; 10:05pm
URL: http://imagej.273.s1.nabble.com/I-may-have-found-the-cause-of-my-problem-stitching-stacks-with-5-planes-tp5003778p5006045.html

On Fri, Jul 5, 2013 at 5:08 PM, Stephan Preibisch <[hidden email]> wrote:
On Fri, Jul 5, 2013 at 10:16 AM, Gelman, Laurent <[hidden email]> wrote:
>> I tested different numbers (adding empty planes to my
>> stacks), and actually any number equal to or above 8 seem fine. I guess
>> this is because it is now possible to merge 7 channels...??
>>
>> If this is the answer, then I would find nice, provided that image
>> properties are set correctly, if the plugin could not convert Z-planes
>> into channels. It does happen to make Z-stacks with less than 8 planes...
>>
--snip--
> I suspect that something may go wrong when loading the images. Do you use
> TIFFs?
>

Hi Stephan,

What's happening is using "Image output"="Fuse and display", shows the
Z slices as channels if there are less than (8) Z slices.
When using more than (8) slices, the Z appears correctly.
In both cases, the data is written correctly when using "Image
output"="Write to disk" - it saves the filenames with z1, z2, etc. and
importing the data with LOCI Bioformats populates creates Z stacks.

I was able to reproduce this problem with today's continuous build of
FIJI, Stitching_-2.0.0-SNAPSHOT and Bioformats-4.4.9.
Data type = ome.tif files, 1 timepoint, 1 channel, 3 z slices, 3x3
grid, created by Micro-Manager 1.4.15.
You can download the data (11 MB) here:
https://drive.google.com/file/d/0By6i416w_DypOTctbG1vZTNkLWc/edit?usp=sharing
The settings I use to open the data in FIJI are:
1) Open "Grid/Collection Stitching" plugin with:
"Positions from File"
"Defined by image metadata"
2) Set "Browse for Multi_series file" to first image.
3) Uncheck "Compute overlap ..."
4) Set "Increase overlap %" to 0
5) Check "Invert Y coordinates"


Best wishes,

>> Laurent.

> Stephan

Pariksheet

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html