ImageJ

Image J - VSI Reader

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

16 messages Options

Crystal

Image J - VSI Reader

Hi all,

I'm trying to do colocalization and/or particle counts from an Olympus file and recently became aware of the VSI reader plugin, which seems perfect for the first step. As instructed, I installed the Action Bar file and then the VSI Reader within this folder. The appearance seems a little different than the manual, mainly with it missing the icons (I also tried installing the icon folder separately but this didn't help). More importantly, the first option I have after opening the VSI Reader is "Open Tile File" instead of "Load .vsi" as described in the manual. The problem I'm having is when I try to open the .vsi file and select the series, I keep getting "Macro error". I've tried installing this plugin on both a 32 and 64 bit computer and get the same error. I don't have any macros created (yet!) so I'm not sure why I get this error when only trying to open the file. Any help would be greatly appreciated!

Crystal

Olivier Burri

Re: Image J - VSI Reader

Dear Crystal,

Hopefully I can help with the two problems you are experiencing (Missing Icons & "Macro Error").

When we had made this ActionBar, we placed the icons in ActionBar/Shared Bars/icons.
This is no longer possible with the latest version of the actionbar plugin.

To fix this, move the 'icons' folder in the 'shared bars' one folder up (So you get "ActionBar/icons"). This should add the icons back.

The "Macro Error" is most likely caused by scifio.jar missing in the "jars" folder (if you look at the error log that is generated, you should see a more 'human readable' message suggesting to get scifio.jar)

You can download it here:
http://www.openmicroscopy.org/site/products/bio-formats/downloads

In case it is not, I suggest you activate the PTBIOP update site and make sure you have the latest version of the VSI Reader.
http://fiji.sc/How_to_follow_a_3rd_party_update_site

In case this does not help, feel free to send me the error log you get, along with a link to your VSI file and I'll help you out from here.

The documentation will be updated as soon as possible.

All the best,

Oli

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

Crystal

Re: Image J - VSI Reader

In reply to this post by Crystal

Dear Olivier,

Thanks so much for taking the time to respond. I tried your suggestions but am still getting the same error. Attached is the log file and one of my test analysis files (ignore the file name... it was after hours of problems with the microscope haha).

error:
Log.txt

vsi file:
yay.vsi

Best,
Crystal

-----------------------
Sep 03, 2013; 4:02am
Re: Image J - VSI Reader
Olivier Burri
50 posts

"To fix this, move the 'icons' folder in the 'shared bars' one folder up (So you get "ActionBar/icons"). This should add the icons back.

The "Macro Error" is most likely caused by scifio.jar missing in the "jars" folder (if you look at the error log that is generated, you should see a more 'human readable' message suggesting to get scifio.jar)

You can download it here:
http://www.openmicroscopy.org/site/products/bio-formats/downloads

In case it is not, I suggest you activate the PTBIOP update site and make sure you have the latest version of the VSI Reader.
http://fiji.sc/How_to_follow_a_3rd_party_update_site

In case this does not help, feel free to send me the error log you get, along with a link to your VSI file and I'll help you out from here.

The documentation will be updated as soon as possible.

All the best,

Oli

Olivier Burri

Re: Image J - VSI Reader

Dear Crystal,

In the error log I read this line:
Caused by: loci.formats.MissingLibraryException: The JAI Image I/O Tools are required to read JPEG-2000 files. Please obtain jai_imageio.jar from http://www.openmicroscopy.org/site/support/bio-formats/developers/java-library.html

You need to go to that page, download jai_imageio.jar and place it in your Fiji "jars" folder (not "plugins");

That should fix the error

On a second note, the .VSI file you annexed is only a fraction of your data. Each .vsi file has a folder with the same name. In your case, there should be a "_yay_" folder in the same location as yay.vsi

The vsi file contains only the metadata and some thumbnails. The actual full resolution image data is on the above-mentioned folder.

Hope this helps!

Best

Oli

Olivier Burri
Engineer - Image Processing
& Software Development
EPFL - SV - PTECH - PTBIOP

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Crystal
Sent: lundi 9 septembre 2013 18:36
To: [hidden email]
Subject: Re: Image J - VSI Reader

Dear Olivier,

Thanks so much for taking the time to respond. I tried your suggestions but am still getting the same error. Attached is the log file and one of my test analysis files (ignore the file name... it was after hours of problems with the microscope haha).

error:
Log.txt <http://imagej.1557.x6.nabble.com/file/n5004731/Log.txt>

vsi file:
yay.vsi <http://imagej.1557.x6.nabble.com/file/n5004731/yay.vsi>

Best,
Crystal

-----------------------
Sep 03, 2013; 4:02am
Re: Image J - VSI Reader
Olivier Burri
50 posts

"To fix this, move the 'icons' folder in the 'shared bars' one folder up (So you get "ActionBar/icons"). This should add the icons back.

The "Macro Error" is most likely caused by scifio.jar missing in the "jars"
folder (if you look at the error log that is generated, you should see a more 'human readable' message suggesting to get scifio.jar)

You can download it here:
http://www.openmicroscopy.org/site/products/bio-formats/downloads

In case it is not, I suggest you activate the PTBIOP update site and make sure you have the latest version of the VSI Reader.
http://fiji.sc/How_to_follow_a_3rd_party_update_site

In case this does not help, feel free to send me the error log you get, along with a link to your VSI file and I'll help you out from here.

The documentation will be updated as soon as possible.

All the best,

Oli

--
View this message in context: http://imagej.1557.x6.nabble.com/Image-J-VSI-Reader-tp5004596p5004731.html
Sent from the ImageJ mailing list archive at Nabble.com.

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

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

ctrueden

Re: Image J - VSI Reader

Hi Crystal,

> loci.formats.MissingLibraryException: The JAI Image I/O Tools are
> required to read JPEG-2000 files

Oli is right that the above error message gives a good hint to the problem.
However, the file Oli mentioned, jai_imageio.jar, is already supposed to
ship with Fiji. The fact that it wasn't being shipped was a bug, which I
have now remedied. You should *not* have to download that file manually in
order to read .vsi files in ImageJ.

In short: I think I fixed it. Can you update your Fiji, try opening your
file again, and let us know whether it works?

Thanks,
Curtis

On Tue, Sep 10, 2013 at 7:45 AM, Burri Olivier <[hidden email]>wrote:

> Dear Crystal,
>
> In the error log I read this line:
> Caused by: loci.formats.MissingLibraryException: The JAI Image I/O Tools
> are required to read JPEG-2000 files. Please obtain jai_imageio.jar from
> http://www.openmicroscopy.org/site/support/bio-formats/developers/java-library.html
>
> You need to go to that page, download jai_imageio.jar and place it in your
> Fiji "jars" folder (not "plugins");
>
> That should fix the error
>
> On a second note, the .VSI file you annexed is only a fraction of your
> data. Each .vsi file has a folder with the same name. In your case, there
> should be a "_yay_" folder in the same location as yay.vsi
>
> The vsi file contains only the metadata and some thumbnails. The actual
> full resolution image data is on the above-mentioned folder.
>
> Hope this helps!
>
> Best
>
> Oli
>
> Olivier Burri
> Engineer - Image Processing
> & Software Development
> EPFL - SV - PTECH - PTBIOP
>
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> Crystal
> Sent: lundi 9 septembre 2013 18:36
> To: [hidden email]
> Subject: Re: Image J - VSI Reader
>
> Dear Olivier,
>
> Thanks so much for taking the time to respond. I tried your suggestions
> but am still getting the same error. Attached is the log file and one of my
> test analysis files (ignore the file name... it was after hours of problems
> with the microscope haha).
>
> error:
> Log.txt <http://imagej.1557.x6.nabble.com/file/n5004731/Log.txt>
>
> vsi file:
> yay.vsi <http://imagej.1557.x6.nabble.com/file/n5004731/yay.vsi>
>
>
> Best,
> Crystal
>
> -----------------------
> Sep 03, 2013; 4:02am
> Re: Image J - VSI Reader
> Olivier Burri
> 50 posts
>
> "To fix this, move the 'icons' folder in the 'shared bars' one folder up
> (So you get "ActionBar/icons"). This should add the icons back.
>
> The "Macro Error" is most likely caused by scifio.jar missing in the "jars"
> folder (if you look at the error log that is generated, you should see a
> more 'human readable' message suggesting to get scifio.jar)
>
> You can download it here:
> http://www.openmicroscopy.org/site/products/bio-formats/downloads
>
>
> In case it is not, I suggest you activate the PTBIOP update site and make
> sure you have the latest version of the VSI Reader.
> http://fiji.sc/How_to_follow_a_3rd_party_update_site
>
> In case this does not help, feel free to send me the error log you get,
> along with a link to your VSI file and I'll help you out from here.
>
>
> The documentation will be updated as soon as possible.
>
> All the best,
>
> Oli
>
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Image-J-VSI-Reader-tp5004596p5004731.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> 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

Vitali Khvatkov-2

Re: Image J - VSI Reader

In reply to this post by Crystal

Crystal,

Does your VSI structure have sub-folders with .ETS files in them?
If it does, you have large multi-scale image structure and you need
specialty software to handle it.

Olympus provides free Windows software for viewing those files called
OlyVIA, you can download it here:

http://www.olympusamerica.com/seg_section/product.asp?product=1089&p=-1&c=10

Beyond that, I don't know any third-party software that can handle large
complex VSI structures correctly. (if image is small than .vsi file
contains actual image pixel array and LOCI can read it )

Our engineers spent last 3 months working with Olympus software team out of
Germany to implement proper support for large VSI files on our Simagis Live
Servers and we are still not quite there yet. I can keep you updated on the
progress if you are interested.

Regards,

*Vitali Khvatkov*
Smart Imaging Technologies Co.
<http://www.simagis.com>http://live.simagis.com

On Wed, Aug 28, 2013 at 10:45 AM, Crystal <[hidden email]> wrote:

> Hi all,
>
> I'm trying to do colocalization and/or particle counts from an Olympus file
> and recently became aware of the VSI reader plugin, which seems perfect for
> the first step. As instructed, I installed the Action Bar file and then the
> VSI Reader within this folder. The appearance seems a little different than
> the manual, mainly with it missing the icons (I also tried installing the
> icon folder separately but this didn't help). More importantly, the first
> option I have after opening the VSI Reader is "Open Tile File" instead of
> "Load .vsi" as described in the manual. The problem I'm having is when I
> try
> to open the .vsi file and select the series, I keep getting "Macro error".
> I've tried installing this plugin on both a 32 and 64 bit computer and get
> the same error. I don't have any macros created (yet!) so I'm not sure why
> I
> get this error when only trying to open the file. Any help would be greatly
> appreciated!
>
>
> Crystal
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Image-J-VSI-Reader-tp5004596.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

Crystal

Re: Image J - VSI Reader

In reply to this post by Crystal

Hi all,

Thank you for the assistance. I'll reply to the last three posts one at a time.

@ Oliver: This forum won't allow me to upload the subfiles, as they seem to be too large.
@ Curtis & Oliver: I removed the duplicate .jar files and updated Image J, which fixed the problem. I was then able to open the .vsi file.
@ Vitali: Yes, I have two subfolders that correspond to each .vsi file. I tried to upload the .ETS files but they are too large. I am acquiring my images and then trying to work with them offline (different lab without access to the original Olympus software). I am familiar with the OlyVIA software but its very basic and I can't do any of my planned analyses. I'm planning to use the Object Counter/Nuclei Counter/OBCOL plugin(s) for my data but needed to figure out how to get the Olympus files into ImageJ. You mention LOCI as a possible reader. Would I need to access this through Plugins --> LOCI --> Bio-Formats Importer or am I essentially doing the same thing by File--> Import--> Bio-Formats? I pretty much have to figure out this situation in the next few weeks but will probably be revisiting this issue for future work so any updates on your software would be welcome.

To all:
Now that the bugs are fixed, I noticed there are two ways to open the .vsi file and resolution seems dependent on which method I use.

Option 1: Action Bar --> Shared Bars --> BIOP VSI Reader --> Open tile file
The resultant image looks blurry

Option 2: File --> Import --> Bio Formats --> open as hyperstack
I decided not to use the "crop on import" feature. I'm imaging blood vessels and each slide has a different size vessel and they are all not mounted in the same place. Not sure how I would handle this problem if the file needs to be cropped but will address that if I need to in the future. This option provided a image that seems to have a close resolution to what I can view using the Olympus software.

Today, I tried to crop my section of interest and save as a separate .vsi file. Another option I plan to try is to extract the 10x layer and save that as a .vsi. That way, my import is smaller (no 2x overview, etc.)

Thanks again for the help and forgive any silly questions. I was used to working with Metamorph and now have had to teach myself ImageJ.

Crystal

----------------------------------------

Olivier Burri
Reply | Threaded | More star
Sep 10, 2013; 8:45am
Re: Image J - VSI Reader
Olivier Burri
51 posts

Dear Crystal,

In the error log I read this line:
Caused by: loci.formats.MissingLibraryException: The JAI Image I/O Tools are required to read JPEG-2000 files. Please obtain jai_imageio.jar from http://www.openmicroscopy.org/site/support/bio-formats/developers/java-library.html

You need to go to that page, download jai_imageio.jar and place it in your Fiji "jars" folder (not "plugins");

That should fix the error

On a second note, the .VSI file you annexed is only a fraction of your data. Each .vsi file has a folder with the same name. In your case, there should be a "_yay_" folder in the same location as yay.vsi

The vsi file contains only the metadata and some thumbnails. The actual full resolution image data is on the above-mentioned folder.

Hope this helps!

Best

Oli

Olivier Burri
Engineer - Image Processing
& Software Development
EPFL - SV - PTECH - PTBIOP

----------------------
Curtis Rueden
Reply | Threaded | More star
Sep 10, 2013; 9:48am
Re: Image J - VSI Reader
Curtis Rueden
735 posts

Hi Crystal,

> loci.formats.MissingLibraryException: The JAI Image I/O Tools are
> required to read JPEG-2000 files

Oli is right that the above error message gives a good hint to the problem.
However, the file Oli mentioned, jai_imageio.jar, is already supposed to
ship with Fiji. The fact that it wasn't being shipped was a bug, which I
have now remedied. You should *not* have to download that file manually in
order to read .vsi files in ImageJ.

In short: I think I fixed it. Can you update your Fiji, try opening your
file again, and let us know whether it works?

Thanks,
Curtis
----------------------

Vitali Khvatkov-2
Reply | Threaded | More star
Sep 11, 2013; 12:25pm
Re: Image J - VSI Reader
Vitali Khvatkov-2
16 posts

In reply to this post by Crystal

Crystal,

Does your VSI structure have sub-folders with .ETS files in them?
If it does, you have large multi-scale image structure and you need
specialty software to handle it.

Olympus provides free Windows software for viewing those files called
OlyVIA, you can download it here:

http://www.olympusamerica.com/seg_section/product.asp?product=1089&p=-1&c=10

Beyond that, I don't know any third-party software that can handle large
complex VSI structures correctly. (if image is small than .vsi file
contains actual image pixel array and LOCI can read it )

Our engineers spent last 3 months working with Olympus software team out of
Germany to implement proper support for large VSI files on our Simagis Live
Servers and we are still not quite there yet. I can keep you updated on the
progress if you are interested.

Regards,

*Vitali Khvatkov*
Smart Imaging Technologies Co.
<http://www.simagis.com>http://live.simagis.com

ctrueden

Re: Image J - VSI Reader

Hi Crystal,

> Would I need to access this through Plugins --> LOCI --> Bio-Formats
> Importer or am I essentially doing the same thing by File--> Import-->
> Bio-Formats?

It doesn't matter; those two commands do exactly the same thing.

Regards,
Curtis

On Wed, Sep 11, 2013 at 2:13 PM, Crystal <[hidden email]> wrote:

> Hi all,
>
> Thank you for the assistance. I'll reply to the last three posts one at a
> time.
>
> @ Oliver: This forum won't allow me to upload the subfiles, as they seem to
> be too large.
> @ Curtis & Oliver: I removed the duplicate .jar files and updated Image J,
> which fixed the problem. I was then able to open the .vsi file.
> @ Vitali: Yes, I have two subfolders that correspond to each .vsi file. I
> tried to upload the .ETS files but they are too large. I am acquiring my
> images and then trying to work with them offline (different lab without
> access to the original Olympus software). I am familiar with the OlyVIA
> software but its very basic and I can't do any of my planned analyses. I'm
> planning to use the Object Counter/Nuclei Counter/OBCOL plugin(s) for my
> data but needed to figure out how to get the Olympus files into ImageJ. You
> mention LOCI as a possible reader. Would I need to access this through
> Plugins --> LOCI --> Bio-Formats Importer or am I essentially doing the
> same
> thing by File--> Import--> Bio-Formats? I pretty much have to figure out
> this situation in the next few weeks but will probably be revisiting this
> issue for future work so any updates on your software would be welcome.
>
>
> To all:
> Now that the bugs are fixed, I noticed there are two ways to open the .vsi
> file and resolution seems dependent on which method I use.
>
> Option 1: Action Bar --> Shared Bars --> BIOP VSI Reader --> Open tile file
> The resultant image looks blurry
>
> Option 2: File --> Import --> Bio Formats --> open as hyperstack
> I decided not to use the "crop on import" feature. I'm imaging blood
> vessels
> and each slide has a different size vessel and they are all not mounted in
> the same place. Not sure how I would handle this problem if the file needs
> to be cropped but will address that if I need to in the future. This option
> provided a image that seems to have a close resolution to what I can view
> using the Olympus software.
>
> Today, I tried to crop my section of interest and save as a separate .vsi
> file. Another option I plan to try is to extract the 10x layer and save
> that
> as a .vsi. That way, my import is smaller (no 2x overview, etc.)
>
> Thanks again for the help and forgive any silly questions. I was used to
> working with Metamorph and now have had to teach myself ImageJ.
>
> Crystal
>
>
> ----------------------------------------
>
> Olivier Burri
> Reply | Threaded | More star
> Sep 10, 2013; 8:45am
> Re: Image J - VSI Reader
> Olivier Burri
> 51 posts
>
> Dear Crystal,
>
> In the error log I read this line:
> Caused by: loci.formats.MissingLibraryException: The JAI Image I/O Tools
> are
> required to read JPEG-2000 files. Please obtain jai_imageio.jar from
>
> http://www.openmicroscopy.org/site/support/bio-formats/developers/java-library.html
>
> You need to go to that page, download jai_imageio.jar and place it in your
> Fiji "jars" folder (not "plugins");
>
> That should fix the error
>
> On a second note, the .VSI file you annexed is only a fraction of your
> data.
> Each .vsi file has a folder with the same name. In your case, there should
> be a "_yay_" folder in the same location as yay.vsi
>
> The vsi file contains only the metadata and some thumbnails. The actual
> full
> resolution image data is on the above-mentioned folder.
>
> Hope this helps!
>
> Best
>
> Oli
>
> Olivier Burri
> Engineer - Image Processing
> & Software Development
> EPFL - SV - PTECH - PTBIOP
>
>
> ----------------------
> Curtis Rueden
> Reply | Threaded | More star
> Sep 10, 2013; 9:48am
> Re: Image J - VSI Reader
> Curtis Rueden
> 735 posts
>
> Hi Crystal,
>
> > loci.formats.MissingLibraryException: The JAI Image I/O Tools are
> > required to read JPEG-2000 files
>
> Oli is right that the above error message gives a good hint to the problem.
> However, the file Oli mentioned, jai_imageio.jar, is already supposed to
> ship with Fiji. The fact that it wasn't being shipped was a bug, which I
> have now remedied. You should *not* have to download that file manually in
> order to read .vsi files in ImageJ.
>
> In short: I think I fixed it. Can you update your Fiji, try opening your
> file again, and let us know whether it works?
>
> Thanks,
> Curtis
> ----------------------
>
> Vitali Khvatkov-2
> Reply | Threaded | More star
> Sep 11, 2013; 12:25pm
> Re: Image J - VSI Reader
> Vitali Khvatkov-2
> 16 posts
>
> In reply to this post by Crystal
>
> Crystal,
>
> Does your VSI structure have sub-folders with .ETS files in them?
> If it does, you have large multi-scale image structure and you need
> specialty software to handle it.
>
> Olympus provides free Windows software for viewing those files called
> OlyVIA, you can download it here:
>
>
> http://www.olympusamerica.com/seg_section/product.asp?product=1089&p=-1&c=10
>
> Beyond that, I don't know any third-party software that can handle large
> complex VSI structures correctly. (if image is small than .vsi file
> contains actual image pixel array and LOCI can read it )
>
> Our engineers spent last 3 months working with Olympus software team out of
> Germany to implement proper support for large VSI files on our Simagis Live
> Servers and we are still not quite there yet. I can keep you updated on the
> progress if you are interested.
>
> Regards,
>
>
> *Vitali Khvatkov*
> Smart Imaging Technologies Co.
> <http://www.simagis.com>http://live.simagis.com
>
>
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Image-J-VSI-Reader-tp5004596p5004757.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

Vitali Khvatkov-2

Re: Image J - VSI Reader

In reply to this post by Crystal

Crystal,

Olympus VSI format is very special.

.vsi file itself contains some, but not all image data. More precisely, it
contains low resolution levels of image pyramid, up to certain pixel size,
the high resolution levels are stored in additional .ets files in sub
folders.

But, if image is small enough, like a single frame from camera, it will fit
inside just .vsi file and you will not see additional sub folders with .ets
files.

LOCI Bio-Formats Library is baked inside ImageJ / Fiji. From my
understanding of LOCI source code, LOCI Library handles just .vsi file
itself - it is not concerned with sub folders with .ets files. (LOCI team,
feel free to correct)

As results if you try to open large VSI structure with LOCI library, you
will see only low resolution version of the large image. But small images
like single camera frames will show fine.

We hope to finish full support for VSI digital slides on Simagis Live
servers within next 10 days, there you will be able to run image analysis
applications including ImageJ macros on full-size VSI images. I will keep
you posted.

Regards,

*Vitali Khvatkov*
Smart Imaging Technologies Co.
<http://www.simagis.com>http://live.simagis.com

On Wed, Sep 11, 2013 at 2:13 PM, Crystal <[hidden email]> wrote:

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

Melissa Linkert-2

Re: Image J - VSI Reader

Hi,

> LOCI Bio-Formats Library is baked inside ImageJ / Fiji. From my
> understanding of LOCI source code, LOCI Library handles just .vsi file
> itself - it is not concerned with sub folders with .ets files. (LOCI team,
> feel free to correct)

Please note that is incorrect; Bio-Formats does indeed recognize and
read .ets files associated with .vsi files. For those who wish to
examine the code, please see:

https://github.com/openmicroscopy/bioformats/blob/v4.4.8/components/bio-formats/src/loci/formats/in/CellSensReader.java

Please also note that Bio-Formats (the proper name for what is often
referred to as the LOCI library) is an independent library that happens
to have an ImageJ plugin, but is not tied to ImageJ in any way. It can
just as easily be used from other software; please see
http://openmicroscopy.org/products/bio-formats for up to date information.

> As results if you try to open large VSI structure with LOCI library, you
> will see only low resolution version of the large image. But small images
> like single camera frames will show fine.

Unless an old version of Bio-Formats without proper .vsi support is
installed, this will not be the case.

Regards,
-Melissa

On Wed, Sep 11, 2013 at 04:38:57PM -0500, Vitali Khvatkov wrote:

> Crystal,
>
> Olympus VSI format is very special.
>
> .vsi file itself contains some, but not all image data. More precisely, it
> contains low resolution levels of image pyramid, up to certain pixel size,
> the high resolution levels are stored in additional .ets files in sub
> folders.
>
> But, if image is small enough, like a single frame from camera, it will fit
> inside just .vsi file and you will not see additional sub folders with .ets
> files.
>
> LOCI Bio-Formats Library is baked inside ImageJ / Fiji. From my
> understanding of LOCI source code, LOCI Library handles just .vsi file
> itself - it is not concerned with sub folders with .ets files. (LOCI team,
> feel free to correct)
>
> As results if you try to open large VSI structure with LOCI library, you
> will see only low resolution version of the large image. But small images
> like single camera frames will show fine.
>
> We hope to finish full support for VSI digital slides on Simagis Live
> servers within next 10 days, there you will be able to run image analysis
> applications including ImageJ macros on full-size VSI images. I will keep
> you posted.
>
> Regards,
>
>
> *Vitali Khvatkov*
> Smart Imaging Technologies Co.
> <http://www.simagis.com>http://live.simagis.com
>
>
>
> On Wed, Sep 11, 2013 at 2:13 PM, Crystal <[hidden email]> wrote:
>
> > Hi all,
> >
> > Thank you for the assistance. I'll reply to the last three posts one at a
> > time.
> >
> > @ Oliver: This forum won't allow me to upload the subfiles, as they seem to
> > be too large.
> > @ Curtis & Oliver: I removed the duplicate .jar files and updated Image J,
> > which fixed the problem. I was then able to open the .vsi file.
> > @ Vitali: Yes, I have two subfolders that correspond to each .vsi file. I
> > tried to upload the .ETS files but they are too large. I am acquiring my
> > images and then trying to work with them offline (different lab without
> > access to the original Olympus software). I am familiar with the OlyVIA
> > software but its very basic and I can't do any of my planned analyses. I'm
> > planning to use the Object Counter/Nuclei Counter/OBCOL plugin(s) for my
> > data but needed to figure out how to get the Olympus files into ImageJ. You
> > mention LOCI as a possible reader. Would I need to access this through
> > Plugins --> LOCI --> Bio-Formats Importer or am I essentially doing the
> > same
> > thing by File--> Import--> Bio-Formats? I pretty much have to figure out
> > this situation in the next few weeks but will probably be revisiting this
> > issue for future work so any updates on your software would be welcome.
> >
> >
> > To all:
> > Now that the bugs are fixed, I noticed there are two ways to open the .vsi
> > file and resolution seems dependent on which method I use.
> >
> > Option 1: Action Bar --> Shared Bars --> BIOP VSI Reader --> Open tile file
> > The resultant image looks blurry
> >
> > Option 2: File --> Import --> Bio Formats --> open as hyperstack
> > I decided not to use the "crop on import" feature. I'm imaging blood
> > vessels
> > and each slide has a different size vessel and they are all not mounted in
> > the same place. Not sure how I would handle this problem if the file needs
> > to be cropped but will address that if I need to in the future. This option
> > provided a image that seems to have a close resolution to what I can view
> > using the Olympus software.
> >
> > Today, I tried to crop my section of interest and save as a separate .vsi
> > file. Another option I plan to try is to extract the 10x layer and save
> > that
> > as a .vsi. That way, my import is smaller (no 2x overview, etc.)
> >
> > Thanks again for the help and forgive any silly questions. I was used to
> > working with Metamorph and now have had to teach myself ImageJ.
> >
> > Crystal
> >
> >
> > ----------------------------------------
> >
> > Olivier Burri
> > Reply | Threaded | More star
> > Sep 10, 2013; 8:45am
> > Re: Image J - VSI Reader
> > Olivier Burri
> > 51 posts
> >
> > Dear Crystal,
> >
> > In the error log I read this line:
> > Caused by: loci.formats.MissingLibraryException: The JAI Image I/O Tools
> > are
> > required to read JPEG-2000 files. Please obtain jai_imageio.jar from
> >
> > http://www.openmicroscopy.org/site/support/bio-formats/developers/java-library.html
> >
> > You need to go to that page, download jai_imageio.jar and place it in your
> > Fiji "jars" folder (not "plugins");
> >
> > That should fix the error
> >
> > On a second note, the .VSI file you annexed is only a fraction of your
> > data.
> > Each .vsi file has a folder with the same name. In your case, there should
> > be a "_yay_" folder in the same location as yay.vsi
> >
> > The vsi file contains only the metadata and some thumbnails. The actual
> > full
> > resolution image data is on the above-mentioned folder.
> >
> > Hope this helps!
> >
> > Best
> >
> > Oli
> >
> > Olivier Burri
> > Engineer - Image Processing
> > & Software Development
> > EPFL - SV - PTECH - PTBIOP
> >
> >
> > ----------------------
> > Curtis Rueden
> > Reply | Threaded | More star
> > Sep 10, 2013; 9:48am
> > Re: Image J - VSI Reader
> > Curtis Rueden
> > 735 posts
> >
> > Hi Crystal,
> >
> > > loci.formats.MissingLibraryException: The JAI Image I/O Tools are
> > > required to read JPEG-2000 files
> >
> > Oli is right that the above error message gives a good hint to the problem.
> > However, the file Oli mentioned, jai_imageio.jar, is already supposed to
> > ship with Fiji. The fact that it wasn't being shipped was a bug, which I
> > have now remedied. You should *not* have to download that file manually in
> > order to read .vsi files in ImageJ.
> >
> > In short: I think I fixed it. Can you update your Fiji, try opening your
> > file again, and let us know whether it works?
> >
> > Thanks,
> > Curtis
> > ----------------------
> >
> > Vitali Khvatkov-2
> > Reply | Threaded | More star
> > Sep 11, 2013; 12:25pm
> > Re: Image J - VSI Reader
> > Vitali Khvatkov-2
> > 16 posts
> >
> > In reply to this post by Crystal
> >
> > Crystal,
> >
> > Does your VSI structure have sub-folders with .ETS files in them?
> > If it does, you have large multi-scale image structure and you need
> > specialty software to handle it.
> >
> > Olympus provides free Windows software for viewing those files called
> > OlyVIA, you can download it here:
> >
> >
> > http://www.olympusamerica.com/seg_section/product.asp?product=1089&p=-1&c=10
> >
> > Beyond that, I don't know any third-party software that can handle large
> > complex VSI structures correctly. (if image is small than .vsi file
> > contains actual image pixel array and LOCI can read it )
> >
> > Our engineers spent last 3 months working with Olympus software team out of
> > Germany to implement proper support for large VSI files on our Simagis Live
> > Servers and we are still not quite there yet. I can keep you updated on the
> > progress if you are interested.
> >
> > Regards,
> >
> >
> > *Vitali Khvatkov*
> > Smart Imaging Technologies Co.
> > <http://www.simagis.com>http://live.simagis.com
> >
> >
> >
> >
> >
> > --
> > View this message in context:
> > http://imagej.1557.x6.nabble.com/Image-J-VSI-Reader-tp5004596p5004757.html
> > Sent from the ImageJ mailing list archive at Nabble.com.
> >
> > --
> > 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

Crystal

Re: Image J - VSI Reader

In reply to this post by Crystal

Thank you Curtis, this helps.

Vitali,

Thank you for explaining the process in such detail. That is exactly what I'm running across and thought the loss of resolution was somehow related to import into ImageJ/ export from Olympus.
I will definitely wait to hear back from you and could wait the two weeks until it's completed.

Re:
> As results if you try to open large VSI structure with LOCI library, you
> will see only low resolution version of the large image. But small images
> like single camera frames will show fine.

I'm working with a tiled image of cross-sectionally sliced blood vessels and am trying to identify colocalized small puncta, so maintaining the resolution would be preferred. Not sure how many frames the tile ends up being on this machine but when I was using a confocal, the final image was anywhere from ~ 5x3 to 8x6 tiles in size, just to give you an idea.

Re:
> .vsi file itself contains some, but not all image data. More precisely, it
> contains low resolution levels of image pyramid, up to certain pixel size,
> the high resolution levels are stored in additional .ets files in sub
> folders.

Do you know what pixel size it maintains? I think at 10x, its ~600nm/pixel.

Lastly, will this improved reader be incorporated into the current VSI Reader or do you plan to create a separate plugin?

Best,

Crystal

-----------------
Vitali Khvatkov-2
Reply | Threaded | More star
Sep 11, 2013; 5:38pm
Re: Image J - VSI Reader
Vitali Khvatkov-2
17 posts

In reply to this post by Crystal
Crystal,

Olympus VSI format is very special.

.vsi file itself contains some, but not all image data. More precisely, it
contains low resolution levels of image pyramid, up to certain pixel size,
the high resolution levels are stored in additional .ets files in sub
folders.

But, if image is small enough, like a single frame from camera, it will fit
inside just .vsi file and you will not see additional sub folders with .ets
files.

LOCI Bio-Formats Library is baked inside ImageJ / Fiji. From my
understanding of LOCI source code, LOCI Library handles just .vsi file
itself - it is not concerned with sub folders with .ets files. (LOCI team,
feel free to correct)

As results if you try to open large VSI structure with LOCI library, you
will see only low resolution version of the large image. But small images
like single camera frames will show fine.

We hope to finish full support for VSI digital slides on Simagis Live
servers within next 10 days, there you will be able to run image analysis
applications including ImageJ macros on full-size VSI images. I will keep
you posted.

Regards,

*Vitali Khvatkov*
Smart Imaging Technologies Co.
<http://www.simagis.com>http://live.simagis.com

Crystal

Re: Image J - VSI Reader

Hi Melissa,

Thanks for the clarification.

> Unless an old version of Bio-Formats without proper .vsi support is
> installed, this will not be the case.

I will look at the install dates for Bio-Formats. What would you consider an old version?

Best,

Crystal

---------------------
Melissa Linkert-2
Reply | Threaded | More star
Sep 11, 2013; 5:57pm
Re: Image J - VSI Reader
Melissa Linkert-2
98 posts

Hi,

> LOCI Bio-Formats Library is baked inside ImageJ / Fiji. From my
> understanding of LOCI source code, LOCI Library handles just .vsi file
> itself - it is not concerned with sub folders with .ets files. (LOCI team,
> feel free to correct)

Please note that is incorrect; Bio-Formats does indeed recognize and
read .ets files associated with .vsi files. For those who wish to
examine the code, please see:

https://github.com/openmicroscopy/bioformats/blob/v4.4.8/components/bio-formats/src/loci/formats/in/CellSensReader.java

Please also note that Bio-Formats (the proper name for what is often
referred to as the LOCI library) is an independent library that happens
to have an ImageJ plugin, but is not tied to ImageJ in any way. It can
just as easily be used from other software; please see
http://openmicroscopy.org/products/bio-formats for up to date information.

> As results if you try to open large VSI structure with LOCI library, you
> will see only low resolution version of the large image. But small images
> like single camera frames will show fine.

Unless an old version of Bio-Formats without proper .vsi support is
installed, this will not be the case.

Regards,
-Melissa

Olivier Burri

Re: Image J - VSI Reader

Dears Melissa, Crystal and all,

The LOCI Bioformats library works perfectly well with the .VSI files since 2 major releases already. (So over a year now).

It reads and associates the .VSI file with the folder containing the .ETS files perfectly. And it's thanks to this that we were able to build a tool that helps to open the .VSI file sin Fiji for our users.

> > @ Oliver: This forum won't allow me to upload the subfiles, as they
> > seem to be too large.
We can either setup a server where you can upload your files or I will message you with a dropbox link to a public folder

In response to:

> > Now that the bugs are fixed, I noticed there are two ways to open
> > the .vsi file and resolution seems dependent on which method I use.
> >
> > Option 1: Action Bar --> Shared Bars --> BIOP VSI Reader --> Open
> > tile file The resultant image looks blurry
> >
> > Option 2: File --> Import --> Bio Formats --> open as hyperstack I
> > decided not to use the "crop on import" feature. I'm imaging blood
> > vessels and each slide has a different size vessel and they are all
> > not mounted in the same place. Not sure how I would handle this
> > problem if the file needs to be cropped but will address that if I
> > need to in the future. This option provided a image that seems to
> > have a close resolution to what I can view using the Olympus
> > software.

If the resulting image looks blurry, I can think of the following. Usually the order of the file series as read from LOCI follows a certain order in our system (Series 1=label, Series 2=First acquisition, Series 3=...) And we've managed to sort of guess the order of the files depending on whether there is a label or not, and other variants.

So blurriness can be that the VSI reader cannot associate the thumbnail preview to the right image series, which means the structure of the file output by your machine is different than that of ours. I can't help you unless I have the full dataset (.vsi file + "_folder_" ).

Seeing as you can open the image using option 2, this means they are not so big that you would need the VSI reader. We had to use it because of slide scans that were over 100'000 pixels in both directions, and thus impossible to open directly.

Again the best would be for you to send me your file and we'll open it / modify the VSI reader to cope with your format.

Best

Oli

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

Crystal

Re: Image J - VSI Reader

In reply to this post by Crystal

Dear Oli,

>We can either setup a server where you can upload your files or I will message you with a dropbox link to >a public folder

Either option works for me.

> If the resulting image looks blurry, I can think of the following. Usually the order of the file series as read > from LOCI follows a certain order in our system (Series 1=label, Series 2=First acquisition, Series 3=...)
> And we've managed to sort of guess the order of the files depending on whether there is a label or not,
> and other variants.

Would it help if I first extracted the layer of interest and saved it as a separate .vsi file. That way there is no confusion between the 2x overview scan and 10x scan of interest? On another note, I was thinking about "cropping" the section of interest from the 10x scan and saving that as a separate file anyway.

> So blurriness can be that the VSI reader cannot associate the thumbnail preview to the right image >series, which means the structure of the file output by your machine is different than that of ours. I >can't help you unless I have the full dataset (.vsi file + "_folder_" ).

Extracting and saving the isolated 10x scan should result in a corresponding subfolder, which may solve this problem (?)

>Seeing as you can open the image using option 2, this means they are not so big that you would need >the VSI reader. We had to use it because of slide scans that were over 100'000 pixels in both directions, >and thus impossible to open directly.

The "yay" scan I'm referencing was for simplicity, as it's only a partial scan of a vessel. The real images I'll be working with should be much larger.

>Again the best would be for you to send me your file and we'll open it / modify the VSI reader to cope >with your format.

Let me know the easiest way for me to send you the file(s).

Best,

Crystal

Olivier Burri

Re: Image J - VSI Reader

Dear Crystal,

> Would it help if I first extracted the layer of interest and
> saved it as a separate .vsi file.
> That way there is no confusion between the 2x overview
> scan and 10x scan of interest?
> On another note, I was thinking about "cropping" the
> section of interest from the 10x scan and saving that as a separate file anyway.

The idea of the VSI reader is to try and locate these relationships by using the least tampered-with files. Ideally I'd like to complete acquisition as it came out of the system. You can provide me with both if you'd like.

What size are the files?

Best

Oli

Olivier Burri
Engineer - Image Processing
& Software Development
EPFL - SV - PTECH - PTBIOP

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Crystal
Sent: vendredi 13 septembre 2013 00:59
To: [hidden email]
Subject: Re: Image J - VSI Reader

Dear Oli,

>We can either setup a server where you can upload your files or I will
message you with a dropbox link to >a public folder

Either option works for me.

> If the resulting image looks blurry, I can think of the following.
> Usually the order of the file series as read > from LOCI follows a
> certain order in our system (Series 1=label, Series 2=First
> acquisition, Series 3=...) And we've managed to sort of guess the
> order of the files depending on whether there is a label or not, and
> other variants.

Would it help if I first extracted the layer of interest and saved it as a separate .vsi file. That way there is no confusion between the 2x overview scan and 10x scan of interest? On another note, I was thinking about "cropping" the section of interest from the 10x scan and saving that as a separate file anyway.

> So blurriness can be that the VSI reader cannot associate the
> thumbnail preview to the right image >series, which means the
> structure of the file output by your machine is different than that of
> ours. I >can't help you unless I have the full dataset (.vsi file + "_folder_" ).

Extracting and saving the isolated 10x scan should result in a corresponding subfolder, which may solve this problem (?)

>Seeing as you can open the image using option 2, this means they are
>not so
big that you would need >the VSI reader. We had to use it because of slide scans that were over 100'000 pixels in both directions, >and thus impossible to open directly.

The "yay" scan I'm referencing was for simplicity, as it's only a partial scan of a vessel. The real images I'll be working with should be much larger.

>Again the best would be for you to send me your file and we'll open it
>/
modify the VSI reader to cope >with your format.

Let me know the easiest way for me to send you the file(s).

Best,

Crystal

--
View this message in context: http://imagej.1557.x6.nabble.com/Image-J-VSI-Reader-tp5004596p5004770.html
Sent from the ImageJ mailing list archive at Nabble.com.

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

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

Crystal

Re: Image J - VSI Reader

Dear Oli,

>The idea of the VSI reader is to try and locate these relationships by using the least tampered-with files. >Ideally I'd like to complete acquisition as it came out of the system. You can provide me with both if >you'd like.
I spoke to an engineer from Olympus this morning. He suggested extracting the layer of interest and "save as" a separate .vsi file. However, he is familiar with the Bio-Formats import option and seems to think this step might not be necessary if we can somehow get the system to identify the file with the largest pixel size.
His second suggestion was to 1) extract the layer, then 2) crop the image to the region of interest, and 3) save as. The "save as" is necessary because I have several hundred slides and want corresponding file name/slide names. I know there is an option to designate files names in a batch but we couldn't get it to work, so I'm left with "Image 01, 02, etc.).

>What size are the files?
Sizes seem to range from 600-700 KB for the .vsi files and ~ 49,000 KB for the .ets files.

Best,

Crystal
----------------------------------------------
Olivier Burri
Reply | Threaded | More star
Sep 13, 2013; 6:40am
Re: Image J - VSI Reader
Olivier Burri
53 posts

Dear Crystal,

> Would it help if I first extracted the layer of interest and
> saved it as a separate .vsi file.
> That way there is no confusion between the 2x overview
> scan and 10x scan of interest?
> On another note, I was thinking about "cropping" the
> section of interest from the 10x scan and saving that as a separate file anyway.

The idea of the VSI reader is to try and locate these relationships by using the least tampered-with files. Ideally I'd like to complete acquisition as it came out of the system. You can provide me with both if you'd like.

What size are the files?

Best

Oli

Olivier Burri
Engineer - Image Processing
& Software Development
EPFL - SV - PTECH - PTBIOP