[ITK] [slicer-devel] BigTIFF
Gib Bogle
g.bogle at auckland.ac.nz
Wed Sep 17 03:35:18 EDT 2014
Hi Bradley, Matt and Steve,
Now for the embarrassing confession. In External_ITK4.cmake I turned ITKV3_COMPATIBILITY off and ITK_USE_64BITS_IDS on as suggested, but when I finally checked what the code that is being built was using (with #if defined ...) I discovered that both ITKV3_COMPATIBILITY and ITK_USE_64BITS_IDS were off. How can this be?
It's easy, all it takes is a typo:
-DITK_USE_64_BITS_IDS:BOOL=ON
That typo has cost me a couple of days, and caused annoyance to a few people, I suspect.
Unfortunately... now Slicer's vtkITK does not build. There are very many errors like this:
3>------ Build started: Project: vtkITK, Configuration: Release x64 ------
3>Build started 17/09/2014 7:25:52 p.m..
3>InitializeBuildStatus:
3> Touching "vtkITK.dir\Release\vtkITK.unsuccessfulbuild".
3>CustomBuild:
3> Building Custom Rule C:/Slicer/Libs/vtkITK/CMakeLists.txt
3> CMake does not need to re-run because C:\s\b\Slicer-build\Libs\vtkITK\CMakeFiles\generate.stamp is up-to-date.
3>ClCompile:
3> vtkITKNumericTraits.cxx
3> vtkITKArchetypeDiffusionTensorImageReaderFile.cxx
3> vtkITKArchetypeImageSeriesReader.cxx
3> vtkITKArchetypeImageSeriesScalarReader.cxx
3> vtkITKArchetypeImageSeriesVectorReaderFile.cxx
3> vtkITKArchetypeImageSeriesVectorReaderSeries.cxx
3> vtkITKImageWriter.cxx
3> vtkITKGradientAnisotropicDiffusionImageFilter.cxx
3> vtkITKDistanceTransform.cxx
3>C:\Slicer\Libs\vtkITK\vtkITKArchetypeImageSeriesScalarReader.cxx(179): error C2679: binary '=' : no operator found which takes a right-hand operand of type 'itk::ImportImageContainer<TElementIdentifier,TElement> *' (or there is no acceptable conversion)
3> with
3> [
3> TElementIdentifier=unsigned __int64,
3> TElement=double
3> ]
3> c:\s\b\itkv4\modules\core\common\include\itkSmartPointer.h(124): could be 'itk::SmartPointer<TObjectType> &itk::SmartPointer<TObjectType>::operator =(const itk::SmartPointer<TObjectType> &)'
3> with
3> [
3> TObjectType=itk::ImportImageContainer<unsigned long,double>
3> ]
3> c:\s\b\itkv4\modules\core\common\include\itkSmartPointer.h(128): or 'itk::SmartPointer<TObjectType> &itk::SmartPointer<TObjectType>::operator =(itk::ImportImageContainer<TElementIdentifier,TElement> *)'
3> with
3> [
3> TObjectType=itk::ImportImageContainer<unsigned long,double>,
3> TElementIdentifier=unsigned long,
3> TElement=double
3> ]
3> while trying to match the argument list '(itk::SmartPointer<TObjectType>, itk::ImportImageContainer<TElementIdentifier,TElement> *)'
3> with
3> [
3> TObjectType=itk::ImportImageContainer<unsigned long,double>
3> ]
3> and
3> [
3> TElementIdentifier=unsigned __int64,
3> TElement=double
3> ]
I presume these are related to the use of 64-bit integers.
Over to you guys :)
Gib
________________________________
From: Bradley Lowekamp [blowekamp at mail.nih.gov]
Sent: Tuesday, 16 September 2014 12:45 a.m.
To: Steve Pieper
Cc: Gib Bogle; slicer-devel at bwh.harvard.edu
Subject: Re: [slicer-devel] BigTIFF
Hello,
For you slicer build did you also remove the ITKV3_COMPATIBILITY flag here:
https://github.com/Slicer/Slicer/blob/master/SuperBuild/External_ITKv4.cmake#L48
To test for ITK big tiff compatibility, I'd recommend getting python 2.7 64-bits then installing SimpleITK (i.e pip install SimpleITK), and try to read and write files like these there. More information about getting started is here[1].
You should be a able to generate a image with:
import SimpleITK as sitk
img = sitk.Image( [1250,1250,1250], sitk.sitkUInt8)
img += 10
stats = sitk.StatisticsImageFilter()
stats.Execute(img)
print stats
This should give you basic access to ITK in python to interactively experiment with options with these large images. I have been able to load large images with the binaries of SimpleITK on windows 64-bits. I don't know if I have tried big tiff though.
A simular test could be done with C++ and ITK directly, but I think SimpleITK in python may be quicker for experimentation.
Brad
[1] http://www.itk.org/Wiki/SimpleITK/GettingStarted
[2] http://www.itk.org/SimpleITKDoxygen/html/namespaceitk_1_1simple.html#a838017f46860e8a7bc8a9ebc7f3dbfee
On Sep 15, 2014, at 7:37 AM, Steve Pieper <pieper at isomics.com<mailto:pieper at isomics.com>> wrote:
Hi Gib -
Are you testing only on windows? What we would often try to do for this kind of debugging is to create a small example that other people can test on their configurations. Slicer has a mechanism called a Self Test [1] to allow easy replication and debugging across multiple platforms. It's ideal if we can hone it down to something like what is described as a 'sscce' [2]. The self tests will typically download some data from a public web site, run some operations, and then print something out saying if it passed or failed (or it may crash). Since it's possible to make a small data file that expands into a giant tiff image it should be pretty tractable to pose this issue in the form of a self test.
Best,
Steve
[1] http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/Tutorials/SelfTestModule
[2] http://sscce.org/
On Mon, Sep 15, 2014 at 2:21 AM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
Hi J-C,
Some more info about Slicer with big tiff files.
1250x1250x1250 = 1,953,125,000 bytes loads and displays correctly
1300x1300x1300 = 2,197,000,000 bytes loads but does not display anything
1650x1650x1650 = 4,492,125,000 bytes fails to load (crashes reading page 72)
All suggestions will be gratefully received.
Cheers
Gib
________________________________
From: Jean-Christophe Fillion-Robin [jchris.fillionr at kitware.com<mailto:jchris.fillionr at kitware.com>]
Sent: Sunday, 14 September 2014 8:32 p.m.
To: Gib Bogle
Cc: Steve Pieper; slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Hi Gib,
What do you mean by "reinstalled" ? You should not have to install Slicer to be able to execute it from the build directory.
After building with "Slicer_BUILD_BRAINSTOOLS" turned off the option "ITK_USE_64_BITS_IDS" added to the ITKv4 external project, it should work out.
When done building, make sure to start the Slicer launcher:
C:/path/to/project/Slicer-Superbuild/Slicer-build/Slicer.exe
and not the Slicer executable:
C:/path/to/project/Slicer-Superbuild/Slicer-build/bin/Release/SlicerApp-real.exe
By using the launcher, you ensure that the PATH are properly setup and the expected libraries are loaded.
Hth
Jc
On Sat, Sep 13, 2014 at 11:49 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
I hope I understand correctly. I added the line as shown to C:\Slicer\SuperBuild\External_ITKv4.cmake:
...
# ZLIB
-DITK_USE_SYSTEM_ZLIB:BOOL=ON
-DZLIB_ROOT:PATH=${ZLIB_ROOT}
-DZLIB_INCLUDE_DIR:PATH=${ZLIB_INCLUDE_DIR}
-DZLIB_LIBRARY:FILEPATH=${ZLIB_LIBRARY}
# Big TIFF
-DITK_USE_64_BITS_IDS:BOOL=ON
...
I then did ALL_BUILD in the root of the build directory (C:\s\b). I presume this is the "superbuild". I then redid ALL_BUILD in C:\s\b\Slicer-build, and ITKv4-build, and reinstalled. The program still crashes with the big tiff. Is there somewhere I can check in the project configuration files to see if setting ITK_USE_64_BITS_IDS actually took effect?
Thanks
Gib
________________________________
From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
Sent: Sunday, 14 September 2014 1:36 p.m.
To: Gib Bogle
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Hey Gib -
Sounds like progress - the ITK_USE_64_BITS_IDS should be set to ON in the ITK build process [1] during the superbuild. I don't think much else was required except turning off the IKTv3 parts.
[1] https://github.com/Slicer/Slicer/blob/master/SuperBuild/External_ITKv4.cmake#L32-L68
-Steve
p.s. I hear the ABs beat the Boks recently in a close match (we don't get it on TV). So the Jo'berg game should be a great rematch!
On Sat, Sep 13, 2014 at 9:11 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
You did warn me.
Here is the list of options not ticked:
Slicer_BUILD_BRAINSTOOLS
_LEGACY_CLI
_LandmarkRegistration
_TESTING_EXTENSIONS
_UPDATE_TRANSLATION
_USE_GIT_PROTOCOL
_OpenIGTLink
_PYTHONQT_WITH_OPENSSL
_PYTHONQT_WITH_TCL
_QtTesting
_SimpleITK
_VTK_DEBUG_LINKS
Slicer_WITH_LIBRARY_VERSION
USE_BRAINS*
USE_ConvertBetweenFileFormats
USE_DWIConvert
USE_DebugImageViewer
USE_GTRACT
USE_ICCDEF
USE_ImageCalculator
USE_ReferenceAtlas
-------
Progress! Process Monitor showed me that the exit code was C0000139, which means a DLL was not found. I then realized that I have different occurrences of many of the DLLs on my disk (QT, ITK and VTK) and this prompted me to look more carefully at the PATH environment variable. I just tested with all the new paths at the start of the PATH string, and Slicer starts successfully! I can load a tiff! But ... it still crashes when I attempt to load a 4 GB tiff.
I am now wondering (belatedly!) about the way I tried to build with big tiff capability. I guessed, and added the entry ITK_USE_64BITS_IDS as 'bool'. It appears ticked in Ungrouped Entries. I really should have asked for directions before doing this - now I'm wondering if that switch actually had any effect. Is that the right thing to do, and if not, what is?
Thanks
Gib
________________________________
From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
Sent: Sunday, 14 September 2014 10:02 a.m.
To: Gib Bogle
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Yes, well, don't say I didn't warn you! ; )
You do need loadable modules and you probably do want python (a lot of core stuff is in python, like the editor).
BRAINSTools should be off for now, since it needs ITKv3.
I'm not sure either what to do next, but I suggest maybe trying the Process Monitor from microsoft (sysinternals) and see what it was doing when it exited. But that may be a longshot.
-Steve
On Sat, Sep 13, 2014 at 4:39 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
Hi Steve,
That was an interesting exercise. I managed to build Slicer without Qtloadablemodules. As you might expect, it was not much use. The program started up but without any capabilities, just an empty shell. (I'm a bit handicapped by not knowing anything about how Slicer is constructed.) So I built it again, this time selecting most things except BRAINTools. The build took a long time, partly because I turned on the Python option, which I don't need. This time there were no errors, except for a couple of occasions where the attempt to create a directory failed - this was because the directory already existed, and was easily solved by deleting the directory and building again. I installed everything that wanted to be installed, and modified the environment variables to suit.
My sense of satisfaction was short-lived, however. When Slicer is invoked the Acknowledgements screen flashes up briefly then the program exits. There are no error messages. Dependency Walker shows that all the DLLs are being found. I am assuming that BRAINTools is not needed (not knowing what it is).
I sense that I must be close to getting it working, but I'm not sure what to do next.
Cheers
Gib
________________________________
From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
Sent: Sunday, 14 September 2014 12:36 a.m.
To: Gib Bogle
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Hi Gib -
A lot of the building goes on in parallel so you need to review the whole build log to see what is going on. I'm guessing ITK didn't build correctly. Typically I do a Control-A Control-C in the visual studio output window and then paste it into a text editor to look for issues. Definitely you don't want to be editing any of the visual studio files, since they are generated by cmake every configure -- much better to sort out what is going on at the cmake flag level.
Best,
Steve
On Fri, Sep 12, 2014 at 9:27 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
Looking at the Project Properties for qSlicerBaseQTCore I notice something odd. Here is a snippet from the list of Linker > Input > Additional Dependencies:
...
C:\s\b\ITKv4-build\lib\Release\itktiff-4.6.lib
C:\s\b\ITKv4-build\lib\Release\itkjpeg-4.6.lib
C:\s\b\ITKv4-build\lib\Release\ITKIOMRC-4.6.lib
C:\s\b\ITKv4-build\lib\Release\ITKIOImageBase-4.6.lib
C:\s\b\VTKv5-build\bin\Release\vtkInfovis.lib
C:\s\b\VTKv5-build\bin\Release\vtkWidgets.lib
C:\s\b\VTKv5-build\bin\Release\vtkVolumeRendering.lib
..\..\lib\Slicer-4.3\Release\FreeSurfer.lib
ITKCommon.lib
C:\Qt64\4.8.1\lib\QtOpenGL4.lib
...
For some reason ITKCommon.lib is treated differently from all the other (at first sight) ITK libraries. Any ideas about how I might fix this? If I edit the linker configuration for qSlicerBaseQTCore to point correctly to the ITKCommon library it builds OK, but I'd rather not have to do this 40+ times.
Thanks again
Gib
________________________________
From: slicer-devel-bounces at bwh.harvard.edu<mailto:slicer-devel-bounces at bwh.harvard.edu> [slicer-devel-bounces at bwh.harvard.edu<mailto:slicer-devel-bounces at bwh.harvard.edu>] on behalf of Gib Bogle [g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>]
Sent: Saturday, 13 September 2014 12:59 p.m.
To: Steve Pieper
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Hi Steve,
All goes well until the point where the qSlicer*.lib get made. The first error, while building qSlicerBaseQTCore.lib, is:
LINK : fatal error LNK1181: cannot open input file 'ITKCommon.lib'
I don't know how to interpret this. I am guessing that it wants to link the ITK library that was just built, i.e.:
\s\b\ITKv4-build\lib\Release\ITKCommon-4.6.lib
This is the first of about 40 such link errors.
I'm not sure if ITK really needed to be built, since it is already installed (64-bit) on my system, but I didn't see any option for skipping that step (same for VTK). Is an INSTALL step for ITK needed? Should I set ITK_DIR appropriately?
Thanks
Gib
________________________________
From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
Sent: Saturday, 13 September 2014 9:02 a.m.
To: Gib Bogle
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Hi Gib -
I'd give a qualified 'yes' answer. Technically what you propose is quite reasonable, and would be workable. But if you haven't built a big package on windows before there are quite a few prerequites (qt, cmake, git, visual studio...) and odd things can go wrong if it isn't all done just right. That said, slicer is actually quite well supported on windows, better, IMHO, than many cross-platform systems. But windows is really the trickiest platform in my experience.
-Steve
On Fri, Sep 12, 2014 at 4:32 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
Hi Steve,
Do you think I would be able to build Slicer myself, with ITK_USE_64BITS_IDS and leaving out BRAINSTools and anything else that depends on either ITKDeprecated or ITKV3Compatibility? My need at this point is simply for a 3D viewer that can handle big images.
Thanks
Gib
________________________________
From: Gib Bogle
Sent: Wednesday, 10 September 2014 10:34 a.m.
To: Jean-Christophe Fillion-Robin; Steve Pieper
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: RE: [slicer-devel] BigTIFF
Hi Jean-Christophe,
Any idea when 4.5 will enter the world?
Cheers
Gib
________________________________
From: slicer-devel-bounces at bwh.harvard.edu<mailto:slicer-devel-bounces at bwh.harvard.edu> [slicer-devel-bounces at bwh.harvard.edu<mailto:slicer-devel-bounces at bwh.harvard.edu>] on behalf of Jean-Christophe Fillion-Robin [jchris.fillionr at kitware.com<mailto:jchris.fillionr at kitware.com>]
Sent: Wednesday, 10 September 2014 10:31 a.m.
To: Steve Pieper
Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
Subject: Re: [slicer-devel] BigTIFF
Hi Folks,
Thanks to Brad Lowekamp, Slicer core (r23663) does NOT depend on either ITKDeprecated or ITKV3Compatibility.
That said, BRAINSTools and some extensions are still depending on these module. All details are captured here:
http://na-mic.org/Mantis/view.php?id=3825#c12486
Note also that this will NOT be addressed for Slicer 4.4 release. The enabling of ITK_USE_64BITS_IDS that will allow BigTiff to be loaded is planned for 4.5.
Thanks
Jc
On Fri, Sep 5, 2014 at 10:26 AM, Steve Pieper <pieper at isomics.com<mailto:pieper at isomics.com>> wrote:
Hi Matt -
Thanks, that's helpful.
I suspect that for slicer, the only real issues will be in vtkITK, which includes all the file IO adapters. But I think it will be pretty straightforward.
But there's no reason to start that until we know that BRAINSTools will work - right now at least some of them depend on ITKDeprecated.
-Steve
On Fri, Sep 5, 2014 at 10:06 AM, Matt McCormick <matt.mccormick at kitware.com<mailto:matt.mccormick at kitware.com>> wrote:
Here is what the process would look like:
0) Start with a clean software quality dashboard.
1) Set ITKV3_COMPATIBILITY to OFF in Superbuild/External_ITKv4.cmake
2) Rebuild and encounter errors and test failures.
3) Fix the errors with the help of the ITK migration guide [1] and the
ITK mailing list [2]. They usually fairly strait forward changes
(minor modifications to the API, etc).
4) Set ITK_USE_64BITS_IDS to ON.
5) Profit.
HTH,
Matt
[1] http://itk.org/migrationv4/
[2] http://www.itk.org/ITK/help/mailing.html
On Thu, Sep 4, 2014 at 4:52 PM, Steve Pieper <pieper at isomics.com<mailto:pieper at isomics.com>> wrote:
> Yes, agreed it's a good goal and I believe the expertise exists, but it's
> looking like there's a fair amount of code to retrofit.
>
> Maybe Hans, Brad, Matt or others could provide some idea just what is
> involved?
>
> -Steve
>
>
>
> On Thu, Sep 4, 2014 at 4:50 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
>>
>> :( I imagine that people are increasingly working with very large
>> datasets, so the number of people who would like to have this capability
>> will grow. I, alas, am not an ITK expert.
>> ________________________________
>> From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
>> Sent: Friday, 5 September 2014 8:42 a.m.
>>
>> To: Gib Bogle
>> Cc: Bradley Lowekamp; slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>> Subject: Re: [slicer-devel] BigTIFF
>>
>> Alas the build where I enabled 64 bit id's failed to even configure
>> properly. Lots of key code depends on the ITKDeprecated module, like vtkITK
>> and various components of BRAINSTools. So I would say some ITK experts are
>> going to need to dig in before this mode can be enabled.
>>
>> -Steve
>>
>>
>> On Thu, Sep 4, 2014 at 4:08 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>> wrote:
>>>
>>> I await the results with interest...
>>> ________________________________
>>> From: slicer-devel-bounces at bwh.harvard.edu<mailto:slicer-devel-bounces at bwh.harvard.edu>
>>> [slicer-devel-bounces at bwh.harvard.edu<mailto:slicer-devel-bounces at bwh.harvard.edu>] on behalf of Steve Pieper
>>> [pieper at isomics.com<mailto:pieper at isomics.com>]
>>> Sent: Friday, 5 September 2014 1:32 a.m.
>>> To: Bradley Lowekamp
>>>
>>> Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>> Subject: Re: [slicer-devel] BigTIFF
>>>
>>> Makes sense, thanks for clarifying Brad. Jc recently moved out a lot of
>>> legacy CLI code - perhaps those were the main places ITKV3_COMPATIBILITY was
>>> required? I kicked off a build to see what will happen.
>>>
>>> -Steve
>>>
>>>
>>> On Thu, Sep 4, 2014 at 8:36 AM, Bradley Lowekamp <blowekamp at mail.nih.gov<mailto:blowekamp at mail.nih.gov>>
>>> wrote:
>>>>
>>>> Steve,
>>>>
>>>> Yes, this is certainly an issue with how ITK has been configured in
>>>> Slicer.
>>>>
>>>> To turn on the 64-bit IDs for windows 64, ITKV3_COMPATIBILITY needs to
>>>> be disabled, and then the ITKDeprecated module can no longer be used. I have
>>>> brought these issue up before and offered assistance but no one wanted to
>>>> champion changing these configuration flags.
>>>>
>>>> Brad
>>>>
>>>> On Sep 3, 2014, at 6:30 PM, Steve Pieper <pieper at isomics.com<mailto:pieper at isomics.com>> wrote:
>>>>
>>>> Hi Brad -
>>>>
>>>> Right, this is not tiff specific but is ITK specific. I was able to
>>>> create a 1650^3 volume (of doubles) in VTK, save with slicer's vtkNRRDWriter
>>>> and reload with vtkNRRDReader. Could not read the same file with ITK's NRRD
>>>> reader. This was all on a slicer current trunk build with default options
>>>> on win7 64bit.
>>>>
>>>> So are we out of luck enabling 64bit IDs or is it just something nobody
>>>> has tried?
>>>>
>>>> -Steve
>>>>
>>>>
>>>> On Wed, Sep 3, 2014 at 11:17 AM, Bradley Lowekamp
>>>> <blowekamp at mail.nih.gov<mailto:blowekamp at mail.nih.gov>> wrote:
>>>>>
>>>>> Steve,
>>>>>
>>>>> Can you usr/load any image greater that 2GB into Slicer on windows 64?
>>>>>
>>>>> There should be some image source filters in the SimpleFilters modules
>>>>> to generate images. I don't think this issue is tiff specific.
>>>>>
>>>>> Bread
>>>>>
>>>>> On Sep 3, 2014, at 11:03 AM, Steve Pieper <pieper at isomics.com<mailto:pieper at isomics.com>> wrote:
>>>>>
>>>>> I was able to replicate the crash - Thanks for the data, Gib!
>>>>>
>>>>> I happens in the itk tiff reader. I created a bug report and pasted in
>>>>> a stack trace and put a link to the data.
>>>>>
>>>>> http://na-mic.org/Bug/view.php?id=3825
>>>>>
>>>>> -Steve
>>>>>
>>>>>
>>>>> On Tue, Sep 2, 2014 at 5:24 PM, Bradley Lowekamp
>>>>> <blowekamp at mail.nih.gov<mailto:blowekamp at mail.nih.gov>> wrote:
>>>>>>
>>>>>> Matt, I was just about to point that out.
>>>>>>
>>>>>> The other things is the compatibility flag:
>>>>>>
>>>>>> https://github.com/Slicer/Slicer/blob/master/SuperBuild/External_ITKv4.cmake#L48
>>>>>>
>>>>>> I think the prevents the 64-bit IDs being used. That is it's still
>>>>>> long not long long, so that it's "compatible" with ITKv3.
>>>>>>
>>>>>> https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/Core/Common/include/itkIntTypes.h#L122
>>>>>>
>>>>>> Brad
>>>>>>
>>>>>> On Sep 2, 2014, at 5:15 PM, Steve Pieper <pieper at isomics.com<mailto:pieper at isomics.com>> wrote:
>>>>>>
>>>>>> Thanks Matt - that looks logical! Let's see if we can replicate the
>>>>>> issue and then we can see if that has any side effects.
>>>>>>
>>>>>>
>>>>>> On Tue, Sep 2, 2014 at 4:44 PM, Matt McCormick
>>>>>> <matt.mccormick at kitware.com<mailto:matt.mccormick at kitware.com>> wrote:
>>>>>>>
>>>>>>> Here is an untested patch that may fix the problem:
>>>>>>>
>>>>>>>
>>>>>>> https://github.com/thewtex/Slicer/commit/24d571c970d012015a95ac6063e3b0a3fede96fb
>>>>>>>
>>>>>>> On Tue, Sep 2, 2014 at 4:42 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>>
>>>>>>> wrote:
>>>>>>> > I have to do something else for a while now. I'll get back to this
>>>>>>> > later.
>>>>>>> > ________________________________
>>>>>>> > From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
>>>>>>> > Sent: Wednesday, 3 September 2014 8:41 a.m.
>>>>>>> >
>>>>>>> > To: Gib Bogle
>>>>>>> > Cc: Bradley Lowekamp; slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>>> > Subject: Re: [slicer-devel] BigTIFF
>>>>>>> >
>>>>>>> > Sure, if that works that's fine. Or maybe Brad knows how to
>>>>>>> > generate a big
>>>>>>> > file with SimpleITK. Or a statically linked executable for any
>>>>>>> > standard
>>>>>>> > platform should be fine too.
>>>>>>> >
>>>>>>> >
>>>>>>> > On Tue, Sep 2, 2014 at 4:40 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>>
>>>>>>> > wrote:
>>>>>>> >>
>>>>>>> >> On second thoughts a very simple image will compress to a
>>>>>>> >> reasonable size.
>>>>>>> >> Slicer seems to handle the compression modes used by ITK without
>>>>>>> >> problems.
>>>>>>> >> ________________________________
>>>>>>> >> From: Steve Pieper [pieper at isomics.com<mailto:pieper at isomics.com>]
>>>>>>> >> Sent: Wednesday, 3 September 2014 8:31 a.m.
>>>>>>> >> To: Gib Bogle
>>>>>>> >> Cc: Bradley Lowekamp; slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>>> >> Subject: Re: [slicer-devel] BigTIFF
>>>>>>> >>
>>>>>>> >> Hi -
>>>>>>> >>
>>>>>>> >> Is there an easy way to generate a big tiff like this, for example
>>>>>>> >> with a
>>>>>>> >> few lines of SimpleITK code? (Easier than trying to transfer a
>>>>>>> >> 4.5GB file
>>>>>>> >> from NZ).
>>>>>>> >>
>>>>>>> >> -Steve
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> On Tue, Sep 2, 2014 at 4:26 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>>
>>>>>>> >> wrote:
>>>>>>> >>>
>>>>>>> >>> Hi Brad,
>>>>>>> >>>
>>>>>>> >>> I installed Slicer-4.3.1 64-bit version on Windows 7. I can load
>>>>>>> >>> a tiff
>>>>>>> >>> file that is about 1 GB (grey-scale 1000x1000x1000), but a couple
>>>>>>> >>> of bigger
>>>>>>> >>> files fail. When I try to open a file that is about 4.5 GB
>>>>>>> >>> (actually when I
>>>>>>> >>> try to add data to the scene) the program crashes immediately,
>>>>>>> >>> exception
>>>>>>> >>> code c0000005. The result with an image of about half this size
>>>>>>> >>> is the
>>>>>>> >>> same. Sometimes the behaviour is slightly different - I see the
>>>>>>> >>> "Loading..." popup, but it's stuck on 15% and the windows greys
>>>>>>> >>> out.
>>>>>>> >>> Nothing happens after that, until I click the x button on the
>>>>>>> >>> popup, and the
>>>>>>> >>> program crashes.
>>>>>>> >>>
>>>>>>> >>> These files were all written with ITK. The reason I installed
>>>>>>> >>> Slicer is
>>>>>>> >>> simply to be able to view big files - Irfanview is very good but
>>>>>>> >>> BigTIFFs
>>>>>>> >>> are not enabled.
>>>>>>> >>>
>>>>>>> >>> Is it a case of having to build Slicer myself, with different
>>>>>>> >>> build
>>>>>>> >>> options?
>>>>>>> >>>
>>>>>>> >>> Thanks
>>>>>>> >>> Gib
>>>>>>> >>> ________________________________
>>>>>>> >>> From: Bradley Lowekamp [blowekamp at mail.nih.gov<mailto:blowekamp at mail.nih.gov>]
>>>>>>> >>> Sent: Wednesday, 3 September 2014 1:00 a.m.
>>>>>>> >>> To: Gib Bogle
>>>>>>> >>> Cc: slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>>> >>> Subject: Re: [slicer-devel] BigTIFF
>>>>>>> >>>
>>>>>>> >>> Hello,
>>>>>>> >>>
>>>>>>> >>> I believe Slicer is using ITK's libtiff library 4.0.3, and the
>>>>>>> >>> associated
>>>>>>> >>> ITK ImageIO. This should support BigTIFFs. Could you please
>>>>>>> >>> provide more
>>>>>>> >>> detail about your issue?
>>>>>>> >>>
>>>>>>> >>> Brad
>>>>>>> >>>
>>>>>>> >>> On Sep 1, 2014, at 5:14 PM, Gib Bogle <g.bogle at auckland.ac.nz<mailto:g.bogle at auckland.ac.nz>>
>>>>>>> >>> wrote:
>>>>>>> >>>
>>>>>>> >>> Hi,
>>>>>>> >>> It seems that Slicer (the Windows version, anyway) does not
>>>>>>> >>> handle
>>>>>>> >>> BigTIFFs. Is there a plan to add this capability?
>>>>>>> >>> Thanks
>>>>>>> >>> Gib
>>>>>>> >>> _______________________________________________
>>>>>>> >>> slicer-devel mailing list
>>>>>>> >>> slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>>> >>> http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
>>>>>>> >>> To unsubscribe: send email to
>>>>>>> >>> slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu> with unsubscribe as
>>>>>>> >>> the
>>>>>>> >>> subject
>>>>>>> >>>
>>>>>>> >>>
>>>>>>> >>> http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
>>>>>>> >>>
>>>>>>> >>>
>>>>>>> >>> The information in this e-mail is intended only for the person to
>>>>>>> >>> whom it
>>>>>>> >>> is
>>>>>>> >>> addressed. If you believe this e-mail was sent to you in error
>>>>>>> >>> and the
>>>>>>> >>> e-mail
>>>>>>> >>> contains patient information, please contact the Partners
>>>>>>> >>> Compliance
>>>>>>> >>> HelpLine at
>>>>>>> >>> http://www.partners.org/complianceline . If the e-mail was sent
>>>>>>> >>> to you in
>>>>>>> >>> error
>>>>>>> >>> but does not contain patient information, please contact the
>>>>>>> >>> sender and
>>>>>>> >>> properly
>>>>>>> >>> dispose of the e-mail.
>>>>>>> >>>
>>>>>>> >>>
>>>>>>> >>>
>>>>>>> >>> _______________________________________________
>>>>>>> >>> slicer-devel mailing list
>>>>>>> >>> slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>>> >>> http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
>>>>>>> >>> To unsubscribe: send email to
>>>>>>> >>> slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu> with unsubscribe as
>>>>>>> >>> the
>>>>>>> >>> subject
>>>>>>> >>>
>>>>>>> >>>
>>>>>>> >>> http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
>>>>>>> >>>
>>>>>>> >>>
>>>>>>> >>> The information in this e-mail is intended only for the person to
>>>>>>> >>> whom it
>>>>>>> >>> is
>>>>>>> >>> addressed. If you believe this e-mail was sent to you in error
>>>>>>> >>> and the
>>>>>>> >>> e-mail
>>>>>>> >>> contains patient information, please contact the Partners
>>>>>>> >>> Compliance
>>>>>>> >>> HelpLine at
>>>>>>> >>> http://www.partners.org/complianceline . If the e-mail was sent
>>>>>>> >>> to you in
>>>>>>> >>> error
>>>>>>> >>> but does not contain patient information, please contact the
>>>>>>> >>> sender and
>>>>>>> >>> properly
>>>>>>> >>> dispose of the e-mail.
>>>>>>> >>>
>>>>>>> >>
>>>>>>> >
>>>>>>> >
>>>>>>> > _______________________________________________
>>>>>>> > slicer-devel mailing list
>>>>>>> > slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>>> > http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
>>>>>>> > To unsubscribe: send email to
>>>>>>> > slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu>
>>>>>>> > with unsubscribe as the subject
>>>>>>> >
>>>>>>> > http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
>>>>>>> >
>>>>>>> >
>>>>>>> > The information in this e-mail is intended only for the person to
>>>>>>> > whom it is
>>>>>>> > addressed. If you believe this e-mail was sent to you in error and
>>>>>>> > the
>>>>>>> > e-mail
>>>>>>> > contains patient information, please contact the Partners
>>>>>>> > Compliance
>>>>>>> > HelpLine at
>>>>>>> > http://www.partners.org/complianceline . If the e-mail was sent to
>>>>>>> > you in
>>>>>>> > error
>>>>>>> > but does not contain patient information, please contact the sender
>>>>>>> > and
>>>>>>> > properly
>>>>>>> > dispose of the e-mail.
>>>>>>> >
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> slicer-devel mailing list
>>>>>> slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
>>>>>> http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
>>>>>> To unsubscribe: send email to
>>>>>> slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu> with unsubscribe as the
>>>>>> subject
>>>>>>
>>>>>> http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
>>>>>>
>>>>>>
>>>>>> The information in this e-mail is intended only for the person to whom
>>>>>> it is
>>>>>> addressed. If you believe this e-mail was sent to you in error and the
>>>>>> e-mail
>>>>>> contains patient information, please contact the Partners Compliance
>>>>>> HelpLine at
>>>>>> http://www.partners.org/complianceline . If the e-mail was sent to you
>>>>>> in error
>>>>>> but does not contain patient information, please contact the sender
>>>>>> and properly
>>>>>> dispose of the e-mail.
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>
>
> _______________________________________________
> slicer-devel mailing list
> slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
> http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
> To unsubscribe: send email to slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu>
> with unsubscribe as the subject
> http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
>
>
> The information in this e-mail is intended only for the person to whom it is
> addressed. If you believe this e-mail was sent to you in error and the
> e-mail
> contains patient information, please contact the Partners Compliance
> HelpLine at
> http://www.partners.org/complianceline . If the e-mail was sent to you in
> error
> but does not contain patient information, please contact the sender and
> properly
> dispose of the e-mail.
>
_______________________________________________
slicer-devel mailing list
slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
To unsubscribe: send email to slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu> with unsubscribe as the subject
http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.
--
+1 919 869 8849<tel:%2B1%20919%20869%208849>
_______________________________________________
slicer-devel mailing list
slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
To unsubscribe: send email to slicer-devel-request at massmail.spl.harvard.edu<mailto:slicer-devel-request at massmail.spl.harvard.edu> with unsubscribe as the subject
http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.
--
+1 919 869 8849<tel:%2B1%20919%20869%208849>
_______________________________________________
slicer-devel mailing list
slicer-devel at bwh.harvard.edu<mailto:slicer-devel at bwh.harvard.edu>
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-devel
To unsubscribe: send email to slicer-devel-request at massmail.spl.harvard.edu with unsubscribe as the subject
http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Developers/FAQ
The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/community/attachments/20140917/33341cea/attachment-0002.html>
More information about the Community
mailing list