Hi Robert,

I've just reviewed your changes, and am bit confused by the code - not
> your changes in particular, but rather than the fact that original
> code claims support for 16bit samples but then the copy_row function
> that just copies unsigned char's across.  You could fixes the 32bit
> case, but doesn't address the 16bit case.
>

That's right, I only read tiff file with 32bit float, and I haven't adress
the 16bit case.

>
> I also feel that the we could probably just get rid of the copy_row
> function and use memcpy instead, and just compute the size of the row
> needing to be copied. i.e.
>
>   memcpy(currPtr, inbuf, format*w);


I also think that It would be a good thing to use memcpy.

Before I merge any changes I'll need to test before and after results
> with some real imagery that tests 8bit, 16bit and float.  Do you have
> some I could test against?


Here is three files :

lena_8.tif --> byte
lena_16.tif --> uint16
lena_32.tif -->  float32

I have send the files only on your personnal mail because it's a little big
for the list.

Regards


-- 
Christophe Loustaunau.
_______________________________________________
osg-submissions mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org

Reply via email to