Posted by
ctrueden on
Oct 03, 2013; 5:30pm
URL: http://imagej.273.s1.nabble.com/Memory-issue-tp5005018p5005031.html
Hi Andrew,
> The format is thousands of individual TIFF files compiled into a
> folder. I collected the images using a Prairie scope (PrairieView
> software), so I'm using the .XML file to open everything.
OK, good to know.
> I'm trying to open a fairly large file (3 GB) which is a Z series over
> time with multiple stage locations totaling about 2800 images.
We have lots of such data here at LOCI. We use the Fiji Stitching plugin to
create mosaics. Do you need to do that, or is it sufficient for you to
analyze each stage position individually? If you need a mosaic, have you
tried following the instructions at:
http://loci.wisc.edu/software/fiji-image-stitching?
Depending how you laid out your stage positions, you may also want to check
the "Ignore Z stage position" option to control exactly how Z planes are
stitched. (If this needs further explanation, I can improve the
documentation to illustrate.)
We also recently added a "Add tiles as ROIs" feature so that you can see
where the original image plane positions fall within the mosaic (and track
down which TIFF file each one came from).
Lastly, if your mosaic is extremely large you can use the "Downsample
tiles" feature to scale down each image plane as they are read in before
they get stitched together. This can save a large amount of RAM.
> I should also note that the file is on an external harddrive. At home
> I was loading into a USB 3.0 port, but at work I only have USB 2.0
> ports.
This shouldn't matter much (at most 2-3x difference) [1].
> I should note that I have never tried to open the files on a Windows
> operating system
Windows can have performance problems when there are many thousands of
small files (as is the case for Prairie datasets). But the only way to know
for certain would be to try it. I would suggest getting it working with the
same configuration on a different OS first, though, if possible.
> We have XP installed on our lab computers, but XP only recognizes
> around 3GB of RAM, so it's not sufficient for opening these files.
Yes, of course if you are running a 32-bit OS you will be limited in RAM.
But e.g. Windows 7 64-bit should work.
And even on 32-bit XP, with the "Use virtual stack" option checked, it
should be possible to open and roam through your data since it is read on
demand from disk in that mode.
> that's why I was trying to confirm if there wasn't another setting I
> needed to change in Windows as opposed to Mac in order to change the
> memory settings.
As far as ImageJ is concerned, there is no OS-specific memory setting
anywhere. But again, as you point out, with 32-bit OSes you will be limited
to ~1.5GB of RAM (as low as ~1.3GB on Windows XP), or ~3GB if you use the
/3GB switch [2].
Regards,
Curtis
> P.S. Since we're at the same place, would it be easier to try to meet
> with you in person to talk about this issue?
Sure, if the Bio-Formats Importer's "Use virtual stack" option does not
solve your problem, and/or you cannot get the Fiji Stitching plugin to work
satisfactorily, we could meet to troubleshoot. Just keep me posted.
[1]
http://blogs.howstuffworks.com/2011/04/13/is-usb-3-0-really-faster-than-usb-2-0/[2]
http://msdn.microsoft.com/en-us/library/windows/hardware/ff556232(v=vs.85).aspxOn Thu, Oct 3, 2013 at 12:09 PM, Andrew Riching <
[hidden email]> wrote:
> Hi Curtis,
>
> The format is thousands of individual TIFF files compiled into a folder. I
> collected the images using a Prairie scope (PrairieView software), so I'm
> using the .XML file to open everything. I have not tried a virtual stack,
> but I can try that when I get home tonight. At work, I successfully opened
> the file on a Mac 10.6.8 machine with 8 GB of RAM and a 2 GHz dual core
> processor in 20 minutes. I should also note that the file is on an external
> harddrive. At home I was loading into a USB 3.0 port, but at work I only
> have USB 2.0 ports. In theory, every aspect of my home computer I would
> expect to outperform the lab computer as far as hardware. I should note
> that
> I have never tried to open the files on a Windows operating system (We have
> XP installed on our lab computers, but XP only recognizes around 3GB of
> RAM,
> so it's not sufficient for opening these files). I wouldn't think changing
> the operating system should make a difference, so that's why I was trying
> to
> confirm if there wasn't another setting I needed to change in Windows as
> opposed to Mac in order to change the memory settings.
>
> Thanks,
> Andrew
>
> P.S. Since we're at the same place, would it be easier to try to meet with
> you in person to talk about this issue?
>
>
>
> --
> View this message in context:
>
http://imagej.1557.x6.nabble.com/Memory-issue-tp5005018p5005028.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