Problem with split channels

[PEARSON Matthew]

Hi all,

I also have a problem with split channels.  I'm using it within a macro and i have two images, one is a single 16bit image the other is a 3 colour composite.  I'm using the print function to troubleshoot.  Basically i want the merged image to be split into its RGB components.  Firstly i use selectImage to activate the composite image to be split.  The print command confirms that the composite image is being activated prior to split.  The image is split and produces the 3 component images.  I then have a getTitle function to see which image is active post split channels and its the 16bit image that i mentioned earlier, not the last component image of the composite.  

Perhaps this is normally behaviour that the last newly created window when using split channels does not automatically become activated?  I've just never had an issue with this before and i've used split channels in a macro many times.

Thanks a lot,

Matt



--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

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

[Rasband, Wayne (NIH/NIMH) [E]]

On Sep 21, 2014, at 9:41 AM, Matt Pearson wrote:

> Hi all,
>
> I also have a problem with split channels.  I'm using it within a macro and i have two images, one is a single 16bit image the other is a 3 colour composite.  I'm using the print function to troubleshoot.  Basically i want the merged image to be split into its RGB components.  Firstly i use selectImage to activate the composite image to be split.  The print command confirms that the composite image is being activated prior to split.  The image is split and produces the 3 component images.  I then have a getTitle function to see which image is active post split channels and its the 16bit image that i mentioned earlier, not the last component image of the composite.  
>
> Perhaps this is normally behaviour that the last newly created window when using split channels does not automatically become activated?  I've just never had an issue with this before and i've used split channels in a macro many times.

Are you using batch mode? These types of problems are less likely in batch mode because images are not displayed and activated on a separate thread. For example, the following macro always outputs "C3-Composite" when run in batch mode but it sometimes displays "Overlays" when not run it batch mode. The macro also runs *much* faster in batch mode.

  setBatchMode(true);
  newImage("Overlays", "16-bit black", 400, 400, 1);
  newImage("Composite", "RGB random", 400, 400, 1);
  run("Make Composite");
  run("Split Channels");
  print(getTitle);

-wayne

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

[PEARSON Matthew]

Hi wayne,

I disabled batch mode to troubleshoot the overlays issue i posted previously and in doing so solved that issue but then created the split channels problem.  The macro code you posted demonstrated the issue and when that is ran in batch mode it always selects the right image post split channels.  However in my actual macro code it still activates the wrong image.  Below is a selection of the code where the error occurs, i might add that batch mode is enabled prior to this segment but i then have to exit to allow the user to add ROI's to an image.  I still don't understand where the issue lies, the image "merge" is definitely the image to be split because the print function suggests so, but then when i query the active image post split, the wrong image is active.  Is it a problem to enable/disable batch mode multiple times?  I know the recommendation is only to use it once, but here i need access to the image.

Thanks,

Matt


run("Z Project...", "projection=[Max Intensity]");
run("Specify...", "width=30 height=30 x=148 y=116 slice=1 oval");
run("ROI Manager...");
setTool("oval");
setBatchMode("exit and display");
waitForUser("Add spots", "Move ROI over spots\n"+"Press t to add to ROI Manager");
setBatchMode(true);
close();//close the projection image
selectImage(merge);
what1=getTitle;
print(what1);
run("Split Channels");
what=getTitle;
print(what);
ch3=getImageID(); ch2=ch3+1; ch1=ch3+2;
print(ch3); print(ch2); print(ch1);
ch3 is DAPI channel
selectImage(ch3);
close();
selectImage(ch2);
 
On 22 Sep 2014, at 16:24, "Rasband, Wayne (NIH/NIMH) [E]" <[hidden email]> wrote:

> On Sep 21, 2014, at 9:41 AM, Matt Pearson wrote:
>
>> Hi all,
>>
>> I also have a problem with split channels.  I'm using it within a macro and i have two images, one is a single 16bit image the other is a 3 colour composite.  I'm using the print function to troubleshoot.  Basically i want the merged image to be split into its RGB components.  Firstly i use selectImage to activate the composite image to be split.  The print command confirms that the composite image is being activated prior to split.  The image is split and produces the 3 component images.  I then have a getTitle function to see which image is active post split channels and its the 16bit image that i mentioned earlier, not the last component image of the composite.  
>>
>> Perhaps this is normally behaviour that the last newly created window when using split channels does not automatically become activated?  I've just never had an issue with this before and i've used split channels in a macro many times.
>
> Are you using batch mode? These types of problems are less likely in batch mode because images are not displayed and activated on a separate thread. For example, the following macro always outputs "C3-Composite" when run in batch mode but it sometimes displays "Overlays" when not run it batch mode. The macro also runs *much* faster in batch mode.
>
>  setBatchMode(true);
>  newImage("Overlays", "16-bit black", 400, 400, 1);
>  newImage("Composite", "RGB random", 400, 400, 1);
>  run("Make Composite");
>  run("Split Channels");
>  print(getTitle);
>
> -wayne
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

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

[Herbie-4]

Dear Matt,

it looks as if it is a kind of race condition. ImageJ uses threads to
distribute routines to them and if you have several processors, to
execute them in parallel or at least nearly in parallel. The outcome of
such distribution is often difficult to predict and of course it depends
on the routines and how they can be executed. Displaying images is time
consuming and, as Wayne has pointed out, is handled in a separate thread...

Of course there are ways to synchronize threads, but it often makes code
clumsy or slow and even may lead to other problems.

You may have a look at the available macro functions:

        <http://rsb.info.nih.gov/ij/developer/macro/functions.html>

I'm sure you will find a way to interrupt batchmode.

HTH

Herbie

::::::::::::::::::::::::::::::::::::::
On 22.09.14 20:56, Matt Pearson wrote:

> Hi wayne,
>
> I disabled batch mode to troubleshoot the overlays issue i posted
> previously and in doing so solved that issue but then created the
> split channels problem.  The macro code you posted demonstrated the
> issue and when that is ran in batch mode it always selects the right
> image post split channels.  However in my actual macro code it still
> activates the wrong image.  Below is a selection of the code where
> the error occurs, i might add that batch mode is enabled prior to
> this segment but i then have to exit to allow the user to add ROI's
> to an image.  I still don't understand where the issue lies, the
> image "merge" is definitely the image to be split because the print
> function suggests so, but then when i query the active image post
> split, the wrong image is active.  Is it a problem to enable/disable
> batch mode multiple times?  I know the recommendation is only to use
> it once, but here i need access to the image.
>
> Thanks,
>
> Matt
>
>
> run("Z Project...", "projection=[Max Intensity]"); run("Specify...",
> "width=30 height=30 x=148 y=116 slice=1 oval"); run("ROI
> Manager..."); setTool("oval"); setBatchMode("exit and display");
> waitForUser("Add spots", "Move ROI over spots\n"+"Press t to add to
> ROI Manager"); setBatchMode(true); close();//close the projection
> image selectImage(merge); what1=getTitle; print(what1); run("Split
> Channels"); what=getTitle; print(what); ch3=getImageID(); ch2=ch3+1;
> ch1=ch3+2; print(ch3); print(ch2); print(ch1); ch3 is DAPI channel
> selectImage(ch3); close(); selectImage(ch2);
>
> On 22 Sep 2014, at 16:24, "Rasband, Wayne (NIH/NIMH) [E]"
> <[hidden email]> wrote:
>
>> On Sep 21, 2014, at 9:41 AM, Matt Pearson wrote:
>>
>>> Hi all,
>>>
>>> I also have a problem with split channels.  I'm using it within a
>>> macro and i have two images, one is a single 16bit image the
>>> other is a 3 colour composite.  I'm using the print function to
>>> troubleshoot.  Basically i want the merged image to be split into
>>> its RGB components.  Firstly i use selectImage to activate the
>>> composite image to be split.  The print command confirms that the
>>> composite image is being activated prior to split.  The image is
>>> split and produces the 3 component images.  I then have a
>>> getTitle function to see which image is active post split
>>> channels and its the 16bit image that i mentioned earlier, not
>>> the last component image of the composite.
>>>
>>> Perhaps this is normally behaviour that the last newly created
>>> window when using split channels does not automatically become
>>> activated?  I've just never had an issue with this before and
>>> i've used split channels in a macro many times.
>>
>> Are you using batch mode? These types of problems are less likely
>> in batch mode because images are not displayed and activated on a
>> separate thread. For example, the following macro always outputs
>> "C3-Composite" when run in batch mode but it sometimes displays
>> "Overlays" when not run it batch mode. The macro also runs *much*
>> faster in batch mode.
>>
>> setBatchMode(true); newImage("Overlays", "16-bit black", 400, 400,
>> 1); newImage("Composite", "RGB random", 400, 400, 1); run("Make
>> Composite"); run("Split Channels"); print(getTitle);
>>
>> -wayne
>>
>> -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html

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

[PEARSON Matthew]

Hi Wayne and Herbie,

Thanks for the info.  I can confirm that the daily build has fixed the issue.

Thanks again for your help on this.

Matt

On 23 Sep 2014, at 01:34, Rasband, Wayne (NIH/NIMH) [E] wrote:

> Hi Matt,
>
> Try the latest ImageJ daily build (1.49i2). The Split Channels command now closes the composite image before opening the channels, which seems to fix the problem you reported. You can upgrade by using the Help>Update ImageJ command and selecting "daily build" from the drop down menu.
>
> Best regards,
>
> -wayne
>
>
> On Sep 22, 2014, at 2:56 PM, Matt Pearson wrote:
>
>> Hi wayne,
>>
>> I disabled batch mode to troubleshoot the overlays issue i posted previously and in doing so solved that issue but then created the split channels problem.  The macro code you posted demonstrated the issue and when that is ran in batch mode it always selects the right image post split channels.  However in my actual macro code it still activates the wrong image.  Below is a selection of the code where the error occurs, i might add that batch mode is enabled prior to this segment but i then have to exit to allow the user to add ROI's to an image.  I still don't understand where the issue lies, the image "merge" is definitely the image to be split because the print function suggests so, but then when i query the active image post split, the wrong image is active.  Is it a problem to enable/disable batch mode multiple times?  I know the recommendation is only to use it once, but here i need access to the image.
>>
>> Thanks,
>>
>> Matt
>>
>>
>> run("Z Project...", "projection=[Max Intensity]");
>> run("Specify...", "width=30 height=30 x=148 y=116 slice=1 oval");
>> run("ROI Manager...");
>> setTool("oval");
>> setBatchMode("exit and display");
>> waitForUser("Add spots", "Move ROI over spots\n"+"Press t to add to ROI Manager");
>> setBatchMode(true);
>> close();//close the projection image
>> selectImage(merge);
>> what1=getTitle;
>> print(what1);
>> run("Split Channels");
>> what=getTitle;
>> print(what);
>> ch3=getImageID(); ch2=ch3+1; ch1=ch3+2;
>> print(ch3); print(ch2); print(ch1);
>> ch3 is DAPI channel
>> selectImage(ch3);
>> close();
>> selectImage(ch2);
>>
>> On 22 Sep 2014, at 16:24, "Rasband, Wayne (NIH/NIMH) [E]" <[hidden email]> wrote:
>>
>>> On Sep 21, 2014, at 9:41 AM, Matt Pearson wrote:
>>>
>>>> Hi all,
>>>>
>>>> I also have a problem with split channels.  I'm using it within a macro and i have two images, one is a single 16bit image the other is a 3 colour composite.  I'm using the print function to troubleshoot.  Basically i want the merged image to be split into its RGB components.  Firstly i use selectImage to activate the composite image to be split.  The print command confirms that the composite image is being activated prior to split.  The image is split and produces the 3 component images.  I then have a getTitle function to see which image is active post split channels and its the 16bit image that i mentioned earlier, not the last component image of the composite.  
>>>>
>>>> Perhaps this is normally behaviour that the last newly created window when using split channels does not automatically become activated?  I've just never had an issue with this before and i've used split channels in a macro many times.
>>>
>>> Are you using batch mode? These types of problems are less likely in batch mode because images are not displayed and activated on a separate thread. For example, the following macro always outputs "C3-Composite" when run in batch mode but it sometimes displays "Overlays" when not run it batch mode. The macro also runs *much* faster in batch mode.
>>>
>>> setBatchMode(true);
>>> newImage("Overlays", "16-bit black", 400, 400, 1);
>>> newImage("Composite", "RGB random", 400, 400, 1);
>>> run("Make Composite");
>>> run("Split Channels");
>>> print(getTitle);
>>>
>>> -wayne
>>>
>>> --
>>> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>>>
>>
>>
>> --
>> The University of Edinburgh is a charitable body, registered in
>> Scotland, with registration number SC005336.
>>
>> --
>> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
>

--
Matt Pearson
Microscopy Facility
MRC Human Genetics Unit
IGMM
University of Edinburgh
Crewe Road
EH4 2XU


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

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

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