I also tried anther task-related fMRI, which has similar parameter, list as follow:

0 1 16 1 move

17 0 16 1 blank

33 2 16 1 static

49 0 16 1 blank

65 1 16 1

81 0 16 1

97 2 16 1

113 0 16 1

129 1 16 1

145 0 16 1

161 2 16 1

177 0 16 1

193 1 16 1

209 0 16 1

225 2 16 1

241 0 16 1

to build the contrast, i use the command

mkanalysis-sess -fsd bold -analysis odd.sm5.fsaverage.lh -surface fsaverage lh -fwhm 5 -paradigm move-vs-static.par -event-related -spmhrf 0 -refeventdur 16 -TR 2   -nconditions 2  -polyfit 2 -per-run

mkcontrast-sess -analysis odd.sm5.fsaverage.lh -contrast MTlocalizer -a 1 -c 2 -nosumconds
the contrast matrix is [1 0; 0 -1], which is correct.  However, the result still show the activation the same with move-vs-blank. Every procedure was finished with out error. I re-run many times but met the same problem. 
Any idea what might be going wrong, or what I can do to fix such a confusing problem?   Thanks!!

Best,
Star

---------- Forwarded message ---------
From: xi star <starxifd@gmail.com>
Date: 2019年1月9日周三下午1:26
Subject: problem with contrast in 1st-level analysis
To: freesurfer <freesurfer@nmr.mgh.harvard.edu>

Hi! FS Experts!

In the fMRI experiment, there are 2 conditions:

1# 3D picture

2# zero disparity picture

interleaved with fixation block (0).

To compare 3D-vs-zero disparity, I used command:

mkcontrast-sess -a 1 -c 2 -nosumconds

but the results showed more activation then expected. To find the problem, I compared 3D-vs-fix (-a 1 -c 0), the result shows the same as 3D-vs-zero disparity.

My question is:

why the two contrast ( [1 0; 0 -1] and [1 0] ) came to the same result? How can I compared the  two conditions?

Thanks a lot !

Yours,

Star 