[MINC-users] mincdiff on world values

[Robert D. Vincent]

Hi Andrew,

Interesting point. Mincdiff is really mostly a tool for comparing headers.
If you want to compare voxel values, minccmp is much more appropriate.

I agree that it might be good to give a slightly more useful message in
mincdiff.

    -bert
On Feb 26, 2016 2:37 PM, "Andrew Wood" <andrew at biospective.com> wrote:

> Hi all,
>
> I came across some interesting behavior with mincdiff:
>
> $ make_phantom ./phantom.mnc
> $ mincmath -add -const 1 ./phantom.mnc ./phantom2.mnc
> $ mincdiff phantom.mnc phantom2.mnc
> ...
> >     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
> 1,
> >     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 ;
> 67c75,80
> <  image-max = 1 ;
> ---
> >  image-max = 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
> 1,
> >     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
> 1,
> >     1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
> 2,
> >     2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
> 1,
> >     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
> 1,
> >     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 ;
> Binary image comparison:
> Images are identical.
>
>
> mincmath didn't change the image block, but rather modified the header to
> change how the voxel values are interpreted, so mincdiff's body comparison
> reported equality.
>
> From a user perspective, it doesn't feel right to see "Images are
> identical" for images that a clearly different. I suppose mincdiff is doing
> what it advertises, but it might be confusing for some users. I was
> wondering if it'd make sense to make mincdiff (optionally) compare *real*
> values, since users are typically more interested in those than they are
> *voxel* values.
>
> Cheers,
> Andrew
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>