Login  Register

Re: Saving and reloading FFT

Posted by Zack Gainsforth on Sep 23, 2014; 5:50am
URL: http://imagej.273.s1.nabble.com/Saving-and-reloading-FFT-tp5009594p5009720.html

Hi All,

Thanks for looking into this!

I am currently using Fiji based on ImageJ 1.49g.  

File>Save As>Tiff does not record that the FFT is in reciprocal space.  The units are saved as direct space coordinates since when I measure a distance or move the mouse around, I get direct space distances.  I can send a sample Tiff offline.

These are not RGB tiffs, they are 16 or 32 bit tiffs.  (Occasionally I use 8 bit, but that's rare.)  Typically they are an FFT directly produced from an HRTEM dm3 file written by digital micrograph on a Zeiss Libra or an FEI Titan.

I would love a solution to this as much as Divakar!

Zack

On Sep 22, 2014, at 7:02 PM, Divakar Ramachandran <[hidden email]> wrote:

On Tuesday 23 September 2014 06:23 AM, Rasband, Wayne (NIH/NIMH) [E] wrote:

> On Sep 22, 2014, at 8:12 PM, Divakar Ramachandran wrote:
>
>> On Friday 12 September 2014 02:35 AM, Zack Gainsforth wrote:
>>> Hi All,
>>>
>>> When I take an FFT of an image, it is obviously using reciprocal space coordinates.  However, if I save that FFT to a tif and then reload it later, it loses the reciprocal space and reports the units incorrectly as direct space.  Is there a way to set an image back into reciprocal space?  I tried, for example, Analyze->Set Scale... Unit of length = nm/c, or changing the distance in pixels from 17.6 to 1/17.6.  Better yet, it should retain that info when I save the image.
>>>
>>> Any ideas?
>> I have been eagerly awaiting a response to this. A few years ago I had
>> asked a related question about how electron diffraction patterns could
>> be scaled in reciprocal units. The response at that time was that it was
>> required to set a property, that was not accessible through the macro
>> language. Is there a solution now? If the info could be retained when
>> saved as TIFF, it would help a lot.
> What version of ImageJ are you using? With version 1.48g and later, File>Save As>Tiff and File>Save As>ZIP save the FFT data of FFT windows, except for FFTs of RGB images.
>
> -wayne
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
I am using 1.48k installed from the Ubuntu 14.04 LTS 64-bit repsitory.
When I open a TIA ser file (HRSTEM image recorded on a Tecnai TEM) with
the TIA plug-in, scale info is correctly available. Process -> FFT gives
a spectrum in which the correct reciprocal distances are displayed in
the status bar in nm units, as the mouse cursor is moved across the
image. When I use Save As TIFF, close the FFT spectrum and re-read it
in, the info at the top of the image is shown with "26.75x26.75 nm
(512x512)" text suggesting that scale info was read from the file;
however, only pixel units are displayed in the status bar as I move
mouse across the FFT spectrum.

Divakar

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

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