Thank you all VERY much for your time and effort on this.
I wish I had mentioned we were working via remote desktop earlier. I
was going to go back and update to the newest version to see it might
be possible to overcome the original "incorrect color count" error.
After speaking to the lead scientist, she says that they must have a
solution that works remotely. We have scientists from all over the
world that come here to do experiments that generate these .tiffs.
They are usually only here for a short time and return to their home
institutions and do their data reconstruction remotely using Octopus,
Volview and other tools on this data analysis server. She does really
like IV3D but when told she could not utilize it remotely, she said
she would look into another software that would not require OpenGL
(if there is one). If you can think of any way to get IV3D to work
through remote desktop that would be great.
Thanks again for all your time.
-Lee
-----Original Message-----
From: Anderson, Lee [mailto:andersonjl@ornl.gov]
Sent: Thursday, September 12, 2013 4:05 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] RE: Re: RE: Re: RE: Re: RE: Re: RE: Re: RE: Re: RE: Re:
RE: Re: IV3D Question
Version 1.3 works ok in a console session on the server. I tried remote desktop on my
office computer and got the exact same OpenGL error in the debug log. I do some it is
hard to keep it all straight, but as best I recall, I was getting that weird
"incorrect color count" error even during a console session. I will verify
that ASAP, but it will probably be next week before I get a chance.
-Lee
-----Original Message-----
From: tom fogal [mailto:tfogal@sci.utah.edu]
Sent: Thursday, September 12, 2013 3:29 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: RE: Re: RE: Re: RE: Re: RE: Re: RE: Re: RE:
Re: IV3D Question
Aaaahhh, this is all via remote desktop?
Windows (and Linux w/o VirtualGL for that matter) disables advanced rendering
things (like shaders) that we need when accessing a system remotely. You'll
need to be physically sitting at the machine to get the OpenGL support we
need.
Please let me know if it still fails when you're sitting at the machine.
-tom
On 09/12/2013 01:23 PM, Anderson, Lee wrote:
The very latest driver from nVidia is installed. See attached screenshot
showing Add/Remove programs, nVidia Control Panel, and Windows Device Manager
all open. Downloading the driver installer to run it again and see if did not
select something to allow the systray icon. Could be because of remote
connected users. I tried to open the nVidia control panel using the icon is
windows control panel and it said I could not launch it using remote desktop.
I had to go to the server and start a console session before nVidia control
panel would launch. Even then the systray icon was not there though, but the
drivers are for sure nVidia drivers and the most recent.
-Lee
-----Original Message-----
From: tom fogal [mailto:tfogal@sci.utah.edu]
Sent: Thursday, September 12, 2013 1:57 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: RE: Re: RE: Re: RE: Re: RE: Re: RE:
Re: IV3D Question
If the system has a GTX660 in it, this appears to be a driver problem.
I don't see the nvidia control panel in your tray; are you using a driver
from nvidia or somewhere else? That might be the issue.
IIRC the default MS driver cripples OpenGL, so one needs to go to the vendor
for a useful driver.
-tom
For the archives, the error message in the screenshot was:
Insufficient OpenGL support:
OpenGL shader objects not supported (GL_ARB_shader_objects)
OpenGL shading language version 1.0 not supported
(GL_ARB_shading_language_100)
OpenGL 3D textures not supported (GL_EXT_texture3D)
OpenGL framebuffer objects not supported
(GL_EXT_framebuffer_object)
On 09/12/2013 11:25 AM, Anderson, Lee wrote:
Looks like it may be an OpenGL support problem as someone previously
suggested. Debug window screenshot attached.
-Lee
-----Original Message-----
From: tom fogal [mailto:tfogal@sci.utah.edu]
Sent: Thursday, September 12, 2013 12:46 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: RE: Re: RE: Re: RE: Re: RE: Re:
IV3D Question
Hi Lee,
Well, progress at least.. sorry you're still having issues.
I think you forgot to attach the screenshots you referenced. But that said,
a debug log would actually be a lot easier for us to track down what's going
wrong, so if you could provide that instead we'd be grateful.
Cheers,
-tom
On 09/12/2013 10:36 AM, Anderson, Lee wrote:
Changing the temp directory cleared up those errors. It can create the .uvf file now, but is
giving me an error when it tries to open the file after it is created. Screenshot attached and
named "New_Error1". When I try to launch IV3D and open the .uvf I get the message shown
in "NeW_Error2".
This system has a GTX660 with the latest drivers, but it is having rendering
issues for some reason.
Thanks.
-Lee
-----Original Message-----
From: tom fogal [mailto:tfogal@sci.utah.edu]
Sent: Tuesday, September 10, 2013 7:18 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: RE: Re: RE: Re: RE: Re: IV3D
Question
Hi, sorry for the delay.
I tried on a couple windows boxes and my Linux workstation, but could not
reproduce the problem. Fortunately I looked through the email chain and saw
this tiny bit in a screenshot of your debug log:
ERROR (IOManager::ConvertDataset): Could not create temp file
C:\Users\ADMINI~1\AppData\Local\Temp\17\testtttttttttttt.uvf~ aborted
conversion.
This is the more likely reason things are failing; sorry I didn't notice
earlier, it's quite buried.
Try this fix: go into IV3D's "Advanced" settings (see screenshot) and change the temporary
directory to something like "C:/users/administrator/desktop/" (at the bottom; mine is
"/tmp/" in the screenshot). Then try the conversion.
If that doesn't fix things, then please generate a debug log and send it to
us---we would need to see more of what IV3D is doing leading up to the
problem.
If you don't trust the 'Report an Issue' uploader, you can run ImageVis3D from the
command line with "ImageVis3D -log c:\users\administrator\desktop\load-error.log
-loglevel 2" to generate a logfile and send it to us manually.
Thanks,
-tom
This error was originally just in a screenshot, but I wanted to enter it into
the email archives so I could find it again when searching for it.
The original error that started this thread was a libtiff error:
incorrect count for field "ColorMap" (1, expecting 196608);
tag ignored
On 09/09/2013 03:01 PM, Anderson, Lee wrote:
Here are some "open" images I can share. Same as the large data
folder
(1955 sequential images) I was testing with, I still get the same
errors when trying to save the .uvf when using these images as well
on the Server 2008R2 system.
-Lee
-----Original Message-----
From: tom fogal [mailto:tfogal@sci.utah.edu]
Sent: Friday, September 06, 2013 4:04 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: RE: Re: RE: Re: IV3D Question
Hey Lee,
On 09/06/2013 08:03 AM, Anderson, Lee wrote:
I checked with the scientist to be sure and she confirmed that I
cannot share the images we are having difficulty with. She is
putting together some images that I could share, but she said they
would not be consecutive scan like the ones giving us the errors
and they may not cause the same error message. I will touch base
again when I get the open beam images from her.
We don't actually care about the image data themselves, of course. They
don't need to be consecutive images, either. Could your colleague open them
up in an image editing program and bucket-fill them to black or something
else to obscure them? If they still give the IV3D error after doing that,
that's all we need.
By the way, the system giving us the most issues is running Server
2008R2. Is ImageVis3D tested on that OS?
We don't have 2008R2 Server sitting around at the moment, but we've also
never seen issues specific to versions of windows, sans dropping support for
XP semi-recently. There's no code in IV3D that tests for a windows version
(I just checked and the same is true for our libtiff).
-tom
-----Original Message-----
From: tom fogal [mailto:tfogal@sci.utah.edu]
Sent: Tuesday, September 03, 2013 7:44 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: RE: Re: IV3D Question
Hi Lee,
Thanks for digging into this, and sorry you're hitting issues. I dug
into the code && that error message is coming out of libtiff. The difference
in behavior that you see is telling, suggesting that a libtiff upgrade or change in
the way we compile it is to blame.
Could you send us two TIFFs from this data set, so that we could reproduce the issue
locally? That would help us a lot. Section 8.3 of our "Getting Data Into
ImageVis3D" [1] has information on how to send data to us.
Thanks again,
-tom
[1]
http://ci.sci.utah.edu:8011/devbuilds/GettingDataIntoImageVis3D.pd
f
On 08/30/2013 10:55 PM, Anderson, Lee wrote:
I just have tried 3.0.0 and 3.1 on this system and get the error. One of our
scientists said she had a working version on another system. I logged into
her system and verified it would open the same files causing this error
without any issues. Then I installed the version she had (v1.3 x64) on this
system and the error does not occur.
-Lee
-----Original Message-----
From: James Hughes [mailto:jhughes@sci.utah.edu]
Sent: Friday, August 30, 2013 4:42 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: RE: Re: IV3D Question
Just uploaded the new executable to ornl link you gave me. Hopefully that
works.
James
On 08/30, Anderson, Lee wrote:
Still having issues with that link. Probably our firewalls here at the lab.
By the way, I just opened the exact same folder of .tiff files on a system
running version 1.3 64-bit and do not get that error. So I ruled out the
.tiff files themselves as a cause.
-Lee
-----Original Message-----
From: James Hughes [mailto:jhughes@sci.utah.edu]
Sent: Friday, August 30, 2013 4:30 PM
To: iv3d-users@sci.utah.edu
Subject: [IV3D-USERS] Re: IV3D Question
Hey Lee,
Try out this download link for a CI dev-build executable:
http://ci.sci.utah.edu:8011/devbuilds/ImageVis3D-x86_64-3.1.0-r2165-2636-294-2208-release.exe.
We're going to be releasing 3.1 soon and this is the executable we will
likely base it off of. It includes many bug fixes and likely fixes the error
you are getting as I'm currently running ImageVis3D on a GTX 680 on Windows.
Note that this link doesn't use the normal HTTP port 80. Let us know if you
have any trouble downloading it.
James
On 08/30, Anderson, Lee wrote:
Good afternoon.
I have installed IV3D on three systems for some of our scientists. They all
have different video cards, the one I really need to work has an nVidia
GTX660 with latest drivers. Hopefully you have seen this error before and
there is a quick fix.
[cid:image001.jpg@01CEA59B.B4BB97F0]
Thank you for your time.
-Lee Anderson
Information Technology Services Division Oak Ridge National
Laboratory
1 Bethel Valley Road
Bldg. 7964K Room 17 MS6430
Oak Ridge, TN 37831
Office - 865-241-6099
Mobile - 865-617-1416
FAX - 865-241-6770
andersonjl@ornl.gov<mailto:andersonjl@ornl.gov>
Archive powered by MHonArc 2.6.16.