I have been experiencing some strange behavior when opening .zvi phase z-stacks with ImageJ 1.43b on a Vista Machine. The individual frames of the stack have scrambled squares that each represent a small fraction of the normal image, but with no particular order to the squares. The squares repeat maybe 4 times and don't together represent the entire image. Scrolling through the stack causes the frames to progress through the z-plane focus, but each frame is made of scrambled squares.
Strangely, the (mis)behavior is not 100% consistent. I can occasionaly get the stacks to open like normal stacks. I have superstiously attributed this "normalcy" to running ImageJ update, opening as a hyperstack, rebooting the computer, but in the end, none of thes seem to produce a consistent fix. The version of ImageJ that is doing this was downloaded from MacBiophotonics site. I have installed and reinstalled this to no avail. Interesting, I have an older imageJ on the same machine that will open the same Image Stack completely normally even as the MacBiophotonics 1.43b version is stuggling. Of course this could be a pure Vista or Video driver problem, but I am at a loss as to how to debug the problem. Any suggestions or ideas why this is happening? Thanks Rob --------------------- Robert W. Baer, Ph.D. Professor of Physiology Kirksville College of Osteopathic Medicne A. T. Still University of Health Sciences 800 W. Jefferson St. Kirksville, MO 63501 |
I tried Wayne's suggestion of updating LOCI using both today's daily build
and the 2008 stable build to no avail. I have the following behavior that may provide the more knowledge with insight. If I load a .zvi image stack into ImageJ with load into Image5D selected instead of standard ImageJ selected I still get the scambled image displayed. If I now switch from grayscale to ovl on the right the first frame of the stack is displayed completely properly. If I now move the z-slider to a subsequent frame, I get a scrambled view of that frame. Moving to color or grayscale and back to ovl make this subsequent frame appear normally. Finally, moving the slider back to frame1 produces a scrambled view of Frame1. After using the ovl trick to get a normal image, performing other ImageJ operations such as Image->Adjust->Brightness and contrast causes the frame to scramble. Any additional insights. Thanks, Rob ----- Original Message ----- From: "Robert W. Baer, Ph.D." <[hidden email]> To: <[hidden email]> Sent: Wednesday, July 08, 2009 11:48 AM Subject: Stack Problem on Vista/Bioimageformats I have been experiencing some strange behavior when opening .zvi phase z-stacks with ImageJ 1.43b on a Vista Machine. The individual frames of the stack have scrambled squares that each represent a small fraction of the normal image, but with no particular order to the squares. The squares repeat maybe 4 times and don't together represent the entire image. Scrolling through the stack causes the frames to progress through the z-plane focus, but each frame is made of scrambled squares. Strangely, the (mis)behavior is not 100% consistent. I can occasionaly get the stacks to open like normal stacks. I have superstiously attributed this "normalcy" to running ImageJ update, opening as a hyperstack, rebooting the computer, but in the end, none of thes seem to produce a consistent fix. The version of ImageJ that is doing this was downloaded from MacBiophotonics site. I have installed and reinstalled this to no avail. Interesting, I have an older imageJ on the same machine that will open the same Image Stack completely normally even as the MacBiophotonics 1.43b version is stuggling. Of course this could be a pure Vista or Video driver problem, but I am at a loss as to how to debug the problem. Any suggestions or ideas why this is happening? Thanks Rob --------------------- Robert W. Baer, Ph.D. Professor of Physiology Kirksville College of Osteopathic Medicne A. T. Still University of Health Sciences 800 W. Jefferson St. Kirksville, MO 63501 |
Hi Rob,
Does the behavior happen with only ZVI files, or with all images in ImageJ? (Try File > Open Samples) If all images, I am inclined to agree that it might be a bug with your video hardware. You can turn down the video's hardware acceleration ( http://www.pixelmetrics.com/Tips/VidBlank/GlobalAcceleration.php) to see if that helps. Another thing to try: if you double-click on loci_tools.jar, it should launch an ImageViewer application. Choose File > Open from the menu, and open your ZVI. Does it look scrambled there? If so, it is probably a bug in Bio-Formats, in which case we would like to obtain a sample nonworking dataset from you, so that we can fix the problem. If you are willing to send us a file but need somewhere to upload, let us know and we will pass along our FTP server information. -Curtis On Wed, Jul 8, 2009 at 5:46 PM, Robert W. Baer, Ph.D. <[hidden email]>wrote: > I tried Wayne's suggestion of updating LOCI using both today's daily build > and the 2008 stable build to no avail. > > I have the following behavior that may provide the more knowledge with > insight. If I load a .zvi image stack into ImageJ with load into Image5D > selected instead of standard ImageJ selected I still get the scambled image > displayed. If I now switch from grayscale to ovl on the right the first > frame of the stack is displayed completely properly. If I now move the > z-slider to a subsequent frame, I get a scrambled view of that frame. Moving > to color or grayscale and back to ovl make this subsequent frame appear > normally. Finally, moving the slider back to frame1 produces a scrambled > view of Frame1. After using the ovl trick to get a normal image, performing > other ImageJ operations such as Image->Adjust->Brightness and contrast > causes the frame to scramble. > > Any additional insights. > > Thanks, > Rob > > ----- Original Message ----- From: "Robert W. Baer, Ph.D." <[hidden email] > > > To: <[hidden email]> > Sent: Wednesday, July 08, 2009 11:48 AM > Subject: Stack Problem on Vista/Bioimageformats > > > I have been experiencing some strange behavior when opening .zvi phase > z-stacks with ImageJ 1.43b on a Vista Machine. The individual frames of the > stack have scrambled squares that each represent a small fraction of the > normal image, but with no particular order to the squares. The squares > repeat maybe 4 times and don't together represent the entire image. > Scrolling through the stack causes the frames to progress through the > z-plane focus, but each frame is made of scrambled squares. > > Strangely, the (mis)behavior is not 100% consistent. I can occasionaly get > the stacks to open like normal stacks. I have superstiously attributed > this "normalcy" to running ImageJ update, opening as a hyperstack, rebooting > the computer, but in the end, none of thes seem to produce a consistent fix. > > The version of ImageJ that is doing this was downloaded from > MacBiophotonics site. I have installed and reinstalled this to no avail. > Interesting, I have an older imageJ on the same machine that will open the > same Image Stack completely normally even as the MacBiophotonics 1.43b > version is stuggling. > > Of course this could be a pure Vista or Video driver problem, but I am at a > loss as to how to debug the problem. Any suggestions or ideas why this is > happening? > > Thanks > > Rob > > --------------------- > Robert W. Baer, Ph.D. > Professor of Physiology > Kirksville College of Osteopathic Medicne > A. T. Still University of Health Sciences > 800 W. Jefferson St. > Kirksville, MO 63501 > |
Thanks Curtis. At Wayne's suggestion, I traced the problem to Java
malfunction and the Jave distributed with MBF. Here is the resolution summary I sent Wayne. I have an NVidia GEForce 6150SE video adaptor. The driver is 7.15.11.7516 datated 5-2-2008 and according to an "update check" is current. This leaves Java. Summary: The MBF ImageJ was running version 1.43c with Java 1.6.0_10-rc (32-bit) (from the JRE file in ImageJ directory) and produces scrambled slices on the z-stack. This Java is presumably part of the MBF Windows installation package. The older ImageJ is 1.40g with Java 1.6.0_05 again installed in the JRE directory of ImageJ. The plugin folder was a copy from the MBF folder above. It opens .zvi z-stacks fine. The 32-bit windows Vista Machine had a native JAVA installation of 1.6.0.11. Upon removing the JRE files from the MBF ImageJ folder, MBF ImageJ reverted, as expedted, to using the Java 1.6.0_11 (32-bit) installed on the system. The .zvi stack files are now apparently opening correctly with properly rendered slices. ----- Original Message ----- From: "Curtis Rueden" <[hidden email]> To: <[hidden email]> Sent: Thursday, July 09, 2009 5:45 PM Subject: Re: more info - Vista/Bioimageformats > Hi Rob, > > Does the behavior happen with only ZVI files, or with all images in > ImageJ? > (Try File > Open Samples) If all images, I am inclined to agree that it > might be a bug with your video hardware. You can turn down the video's > hardware acceleration ( > http://www.pixelmetrics.com/Tips/VidBlank/GlobalAcceleration.php) to see > if > that helps. > > Another thing to try: if you double-click on loci_tools.jar, it should > launch an ImageViewer application. Choose File > Open from the menu, and > open your ZVI. Does it look scrambled there? If so, it is probably a bug > in > Bio-Formats, in which case we would like to obtain a sample nonworking > dataset from you, so that we can fix the problem. If you are willing to > send > us a file but need somewhere to upload, let us know and we will pass along > our FTP server information. > > -Curtis > > On Wed, Jul 8, 2009 at 5:46 PM, Robert W. Baer, Ph.D. > <[hidden email]>wrote: > >> I tried Wayne's suggestion of updating LOCI using both today's daily >> build >> and the 2008 stable build to no avail. >> >> I have the following behavior that may provide the more knowledge with >> insight. If I load a .zvi image stack into ImageJ with load into >> Image5D >> selected instead of standard ImageJ selected I still get the scambled >> image >> displayed. If I now switch from grayscale to ovl on the right the first >> frame of the stack is displayed completely properly. If I now move the >> z-slider to a subsequent frame, I get a scrambled view of that frame. >> Moving >> to color or grayscale and back to ovl make this subsequent frame appear >> normally. Finally, moving the slider back to frame1 produces a scrambled >> view of Frame1. After using the ovl trick to get a normal image, >> performing >> other ImageJ operations such as Image->Adjust->Brightness and contrast >> causes the frame to scramble. >> >> Any additional insights. >> >> Thanks, >> Rob >> >> ----- Original Message ----- From: "Robert W. Baer, Ph.D." >> <[hidden email] >> > >> To: <[hidden email]> >> Sent: Wednesday, July 08, 2009 11:48 AM >> Subject: Stack Problem on Vista/Bioimageformats >> >> >> I have been experiencing some strange behavior when opening .zvi phase >> z-stacks with ImageJ 1.43b on a Vista Machine. The individual frames of >> the >> stack have scrambled squares that each represent a small fraction of the >> normal image, but with no particular order to the squares. The squares >> repeat maybe 4 times and don't together represent the entire image. >> Scrolling through the stack causes the frames to progress through the >> z-plane focus, but each frame is made of scrambled squares. >> >> Strangely, the (mis)behavior is not 100% consistent. I can occasionaly >> get >> the stacks to open like normal stacks. I have superstiously attributed >> this "normalcy" to running ImageJ update, opening as a hyperstack, >> rebooting >> the computer, but in the end, none of thes seem to produce a consistent >> fix. >> >> The version of ImageJ that is doing this was downloaded from >> MacBiophotonics site. I have installed and reinstalled this to no avail. >> Interesting, I have an older imageJ on the same machine that will open >> the >> same Image Stack completely normally even as the MacBiophotonics 1.43b >> version is stuggling. >> >> Of course this could be a pure Vista or Video driver problem, but I am at >> a >> loss as to how to debug the problem. Any suggestions or ideas why this >> is >> happening? >> >> Thanks >> >> Rob >> >> --------------------- >> Robert W. Baer, Ph.D. >> Professor of Physiology >> Kirksville College of Osteopathic Medicne >> A. T. Still University of Health Sciences >> 800 W. Jefferson St. >> Kirksville, MO 63501 >> > |
Free forum by Nabble | Edit this page |