Messed composite display of large two 16-bit channel image - possible bug

[Schebique]

Dear experts.

I have a problem displaying two big channels (2GB each) together in
composite mode. Despite each image open and display correctly even
separately or in color mode of the composite image, the composite (overlay)
image is messed up. See attached sreenshot
(https://owncloud.cesnet.cz/index.php/s/SRFI1fCvWb8sXeW/download). My user
told me, that this happen to him only in case that two files doesnt have
exactly the same size in bytes. I did not have check this info, but could be
useful to you.

Here is the macro demonstrating my problem:

showMessage("Download first large image (2GB). Click OK");
open("https://owncloud.cesnet.cz/index.php/s/DNagS4AtywsBqpw/download");
showMessage("Download second large image (2GB). Click OK");
open("https://owncloud.cesnet.cz/index.php/s/Hkp07p2XtcvT0ZA/download");
dir=getDirectory("Choose a folder where you have downloaded images");
open(dir+"A_overview_C1_max.tif");
open(dir+"A_overview_C2_Z4.tif");
print("note both channels looks ok.");
run("Merge Channels...", "c2=A_overview_C1_max.tif c4=A_overview_C2_Z4.tif
create");
print("composite image of green and gray channel looks messed up");
run("Channels Tool...");
print("color display mode seems OK as well. Play around.");
/*all files includin metadata files of the original Zeiss LSM880 confocal
microscope
 * can be found here: https://owncloud.cesnet.cz/index.php/s/1gowkYOeQazeswH
 * OME-XML: https://owncloud.cesnet.cz/index.php/s/AOFawGuKyHHSgv9/download
 * original metadata extracted by bioformats: https://owncloud.cesnet.cz/
index.php/s/UBtQCtjdm6QiVAx/download
 * my FIJI info: https://owncloud.cesnet.cz/index.php/s/c5c73wb1l5qHAAX/
download
 * For original file (30 GB) pleas contact me at [hidden email].
 * Thank you.
 */

Thank you in advanced for any tips, hints, debug or repair.
Sincerely Ondřej Šebesta

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

[Stoyan Pavlov]

Dear Ondřej ,
Iam not one of the experts but i observed similar problem with very large
RGB images -ImageJopens the image but doesn't display it correctly. I think
(let the experts correct me if I am wrong) that this is due to the internal
limitation of JAva for the maximum number of elements in an array of 231−1.
Two 2G channels will surely exceed this limit. Since than I have found a
way around this - I am doing stitching of the separate images with MIST
plugin and output a list with the image position, which I use in a second
step in the trakem2 plugin to display and blend the images . You can try
stitching directly in trakem2  but doesnt support image grids acquired in
meander fashion. However it supports loading image positions from a file,
so if your acquisition software can produce such file you may skip the
first stitching in the external plugin.

Best luck!
Stoyan


---
Dr. Stoyan P. Pavlov, MD, PhD
Departament of Anatomy, Histology and Embryology
Medical University "Prof. Dr. Paraskev Stoyanov", Varna
Prof. Marin Drinov Str.55
9002 Varna
 Bulgaria
Tel: +359 (0) 52 - 677 - 052
e-mail: [hidden email]
           [hidden email]

2017-09-04 12:49 GMT+03:00 Ondřej Šebesta <[hidden email]>:

> Dear experts.
>
> I have a problem displaying two big channels (2GB each) together in
> composite mode. Despite each image open and display correctly even
> separately or in color mode of the composite image, the composite (overlay)
> image is messed up. See attached sreenshot
> (https://owncloud.cesnet.cz/index.php/s/SRFI1fCvWb8sXeW/download). My user
> told me, that this happen to him only in case that two files doesnt have
> exactly the same size in bytes. I did not have check this info, but could
> be
> useful to you.
>
> Here is the macro demonstrating my problem:
>
> showMessage("Download first large image (2GB). Click OK");
> open("https://owncloud.cesnet.cz/index.php/s/DNagS4AtywsBqpw/download");
> showMessage("Download second large image (2GB). Click OK");
> open("https://owncloud.cesnet.cz/index.php/s/Hkp07p2XtcvT0ZA/download");
> dir=getDirectory("Choose a folder where you have downloaded images");
> open(dir+"A_overview_C1_max.tif");
> open(dir+"A_overview_C2_Z4.tif");
> print("note both channels looks ok.");
> run("Merge Channels...", "c2=A_overview_C1_max.tif c4=A_overview_C2_Z4.tif
> create");
> print("composite image of green and gray channel looks messed up");
> run("Channels Tool...");
> print("color display mode seems OK as well. Play around.");
> /*all files includin metadata files of the original Zeiss LSM880 confocal
> microscope
>  * can be found here: https://owncloud.cesnet.cz/
> index.php/s/1gowkYOeQazeswH
>  * OME-XML: https://owncloud.cesnet.cz/index.php/s/AOFawGuKyHHSgv9/
> download
>  * original metadata extracted by bioformats: https://owncloud.cesnet.cz/
> index.php/s/UBtQCtjdm6QiVAx/download
>  * my FIJI info: https://owncloud.cesnet.cz/index.php/s/c5c73wb1l5qHAAX/
> download
>  * For original file (30 GB) pleas contact me at [hidden email].
>  * Thank you.
>  */
>
> Thank you in advanced for any tips, hints, debug or repair.
> Sincerely Ondřej Šebesta
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

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

[Olivier Burri]

Dear Ondřej and Stoyan,

Tested just now on ImageJ 1.51n and Java 1.8.0_144 [64 bit] and it works OK. Just OK in the sense that response time is very slow but the display is correct.

Are you running a 64 bit of 32 bit version of Java? That could explain perhaps the behavior of your display?

All the best

Oli


-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Stoyan Pavlov
Sent: Wednesday, September 6, 2017 12:31 PM
To: [hidden email]
Subject: Re: Messed composite display of large two 16-bit channel image - possible bug

Dear Ondřej ,
Iam not one of the experts but i observed similar problem with very large RGB images -ImageJopens the image but doesn't display it correctly. I think (let the experts correct me if I am wrong) that this is due to the internal limitation of JAva for the maximum number of elements in an array of 231−1.
Two 2G channels will surely exceed this limit. Since than I have found a way around this - I am doing stitching of the separate images with MIST plugin and output a list with the image position, which I use in a second step in the trakem2 plugin to display and blend the images . You can try stitching directly in trakem2  but doesnt support image grids acquired in meander fashion. However it supports loading image positions from a file, so if your acquisition software can produce such file you may skip the first stitching in the external plugin.

Best luck!
Stoyan


---
Dr. Stoyan P. Pavlov, MD, PhD
Departament of Anatomy, Histology and Embryology Medical University "Prof. Dr. Paraskev Stoyanov", Varna Prof. Marin Drinov Str.55
9002 Varna
 Bulgaria
Tel: +359 (0) 52 - 677 - 052
e-mail: [hidden email]
           [hidden email]

2017-09-04 12:49 GMT+03:00 Ondřej Šebesta <[hidden email]>:

> Dear experts.
>
> I have a problem displaying two big channels (2GB each) together in
> composite mode. Despite each image open and display correctly even
> separately or in color mode of the composite image, the composite
> (overlay) image is messed up. See attached sreenshot
> (https://owncloud.cesnet.cz/index.php/s/SRFI1fCvWb8sXeW/download). My
> user told me, that this happen to him only in case that two files
> doesnt have exactly the same size in bytes. I did not have check this
> info, but could be useful to you.
>
> Here is the macro demonstrating my problem:
>
> showMessage("Download first large image (2GB). Click OK");
> open("https://owncloud.cesnet.cz/index.php/s/DNagS4AtywsBqpw/download"
> ); showMessage("Download second large image (2GB). Click OK");
> open("https://owncloud.cesnet.cz/index.php/s/Hkp07p2XtcvT0ZA/download"
> ); dir=getDirectory("Choose a folder where you have downloaded
> images"); open(dir+"A_overview_C1_max.tif");
> open(dir+"A_overview_C2_Z4.tif");
> print("note both channels looks ok."); run("Merge Channels...",
> "c2=A_overview_C1_max.tif c4=A_overview_C2_Z4.tif create");
> print("composite image of green and gray channel looks messed up");
> run("Channels Tool..."); print("color display mode seems OK as well.
> Play around."); /*all files includin metadata files of the original
> Zeiss LSM880 confocal microscope
>  * can be found here: https://owncloud.cesnet.cz/ 
> index.php/s/1gowkYOeQazeswH
>  * OME-XML: https://owncloud.cesnet.cz/index.php/s/AOFawGuKyHHSgv9/
> download
>  * original metadata extracted by bioformats:
> https://owncloud.cesnet.cz/ index.php/s/UBtQCtjdm6QiVAx/download
>  * my FIJI info:
> https://owncloud.cesnet.cz/index.php/s/c5c73wb1l5qHAAX/
> download
>  * For original file (30 GB) pleas contact me at [hidden email].
>  * Thank you.
>  */
>
> Thank you in advanced for any tips, hints, debug or repair.
> Sincerely Ondřej Šebesta
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

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

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

[Stoyan Pavlov]

Dear Oli,
Thank you for the good news. I am running 64bit version, but haven't tried
for long time ever since I found the workaround.
Best wishes
Stoyan

На 7.09.2017 г. 10:08 пр.об. "Burri Olivier" <[hidden email]> написа:

> Dear Ondřej and Stoyan,
>
> Tested just now on ImageJ 1.51n and Java 1.8.0_144 [64 bit] and it works
> OK. Just OK in the sense that response time is very slow but the display is
> correct.
>
> Are you running a 64 bit of 32 bit version of Java? That could explain
> perhaps the behavior of your display?
>
> All the best
>
> Oli
>
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> Stoyan Pavlov
> Sent: Wednesday, September 6, 2017 12:31 PM
> To: [hidden email]
> Subject: Re: Messed composite display of large two 16-bit channel image -
> possible bug
>
> Dear Ondřej ,
> Iam not one of the experts but i observed similar problem with very large
> RGB images -ImageJopens the image but doesn't display it correctly. I think
> (let the experts correct me if I am wrong) that this is due to the internal
> limitation of JAva for the maximum number of elements in an array of 231−1.
> Two 2G channels will surely exceed this limit. Since than I have found a
> way around this - I am doing stitching of the separate images with MIST
> plugin and output a list with the image position, which I use in a second
> step in the trakem2 plugin to display and blend the images . You can try
> stitching directly in trakem2  but doesnt support image grids acquired in
> meander fashion. However it supports loading image positions from a file,
> so if your acquisition software can produce such file you may skip the
> first stitching in the external plugin.
>
> Best luck!
> Stoyan
>
>
> ---
> Dr. Stoyan P. Pavlov, MD, PhD
> Departament of Anatomy, Histology and Embryology Medical University "Prof.
> Dr. Paraskev Stoyanov", Varna Prof. Marin Drinov Str.55
> 9002 Varna
>  Bulgaria
> Tel: +359 (0) 52 - 677 - 052
> e-mail: [hidden email]
>            [hidden email]
>
> 2017-09-04 12:49 GMT+03:00 Ondřej Šebesta <[hidden email]>:
>
> > Dear experts.
> >
> > I have a problem displaying two big channels (2GB each) together in
> > composite mode. Despite each image open and display correctly even
> > separately or in color mode of the composite image, the composite
> > (overlay) image is messed up. See attached sreenshot
> > (https://owncloud.cesnet.cz/index.php/s/SRFI1fCvWb8sXeW/download). My
> > user told me, that this happen to him only in case that two files
> > doesnt have exactly the same size in bytes. I did not have check this
> > info, but could be useful to you.
> >
> > Here is the macro demonstrating my problem:
> >
> > showMessage("Download first large image (2GB). Click OK");
> > open("https://owncloud.cesnet.cz/index.php/s/DNagS4AtywsBqpw/download"
> > ); showMessage("Download second large image (2GB). Click OK");
> > open("https://owncloud.cesnet.cz/index.php/s/Hkp07p2XtcvT0ZA/download"
> > ); dir=getDirectory("Choose a folder where you have downloaded
> > images"); open(dir+"A_overview_C1_max.tif");
> > open(dir+"A_overview_C2_Z4.tif");
> > print("note both channels looks ok."); run("Merge Channels...",
> > "c2=A_overview_C1_max.tif c4=A_overview_C2_Z4.tif create");
> > print("composite image of green and gray channel looks messed up");
> > run("Channels Tool..."); print("color display mode seems OK as well.
> > Play around."); /*all files includin metadata files of the original
> > Zeiss LSM880 confocal microscope
> >  * can be found here: https://owncloud.cesnet.cz/
> > index.php/s/1gowkYOeQazeswH
> >  * OME-XML: https://owncloud.cesnet.cz/index.php/s/AOFawGuKyHHSgv9/
> > download
> >  * original metadata extracted by bioformats:
> > https://owncloud.cesnet.cz/ index.php/s/UBtQCtjdm6QiVAx/download
> >  * my FIJI info:
> > https://owncloud.cesnet.cz/index.php/s/c5c73wb1l5qHAAX/
> > download
> >  * For original file (30 GB) pleas contact me at [hidden email].
> >  * Thank you.
> >  */
> >
> > Thank you in advanced for any tips, hints, debug or repair.
> > Sincerely Ondřej Šebesta
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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