[Chimera-users] byte order? mac.

Thomas Goddard goddard at cgl.ucsf.edu
Thu Jun 3 10:52:05 PDT 2010

Hi Ray,

   I think I know your problem.  IMOD writes unsigned bytes for data 
values but the MRC format does not support unsigned bytes (0-255), only 
signed bytes (-128 - 127).  So the IMOD .rec file header which is MRC 
format says it is signed bytes which messes up the data display.  The 
data values from 128-255 get mapped to -128 to -1.  When Chimera reads a 
file with suffix ".rec" it assumes signed byte MRC mode means unsigned 
byte.  So I would expect your file to be shown correctly if you kept the 
IMOD .rec suffix on the file name.

   This has nothing to do with the big-endian versus little-endian byte 
order of 4-byte numbers in the file.  Chimera recognizes both byte 
orders and handles that appropriately.

   I wouldn't expect this to effect TIFF image stacks.

   Since it sounds like you see the same problem with TIFF stacks maybe 
you are just noting that the high density gets the smallest data values. 
  That is conventional for tomographic reconstructions I think because 
the signal at the detector is smaller when passing through the higher 
density material.  So large values in the reconstruction corresponds to 
high transparency.  I often invert the scale using the Chimera Volume 
Filter dialog or the vop (volume operation) command.

	vop scale #0 factor -1

Then save the result to a new file (with volume dialog File / Save Map 




Raymond Wightman wrote:
> Hi,
> I'm trying to open several files that were prepared on imod on a unix system - an imod .rec file, tif stacks and an .mrc file. When I open any of these file formats on my mac the histogram seems to be inverted. Inverting the LUT on the tifs using imagej does not work and using the mrcx command of imod to toggle between big- and little-endian of the .mrc file has no effect - Chimera still displays the same inverted histogram. Is there a way that I can invert everything using Chimera?
> The mac is intel.
> Much appreciated,
> Ray.

More information about the Chimera-users mailing list