| Re: FLOAT images instead of BYTE ones from IDL Object graphics ? [message #64839 is a reply to message #64838] |
Fri, 23 January 2009 04:47   |
Gianluca Li Causi
Messages: 21
Registered: August 2005
|
Junior Member |
|
|
On Jan 22, 3:59 pm, David Fanning <n...@dfanning.com> wrote:
> Gianluca Li Causi writes:
>> I'm working with 3D object rendering and I see that the oWindow ->
>> GetProperty, IMAGE_DATA=Img always returns a BYTE-type image, while I
>> need a FLOAT-type one, not quantized in the 256 levels.
>
>> How can I do?
>
>> I need this because I always need a 0_to_255 grayscale image of a
>> volumetric data, while, for any values of the Opacity, I cannot
>> produce a final image with a maximum gray greater than 100 (in fact
>> the final gray levels depends on both the volume data, the opacity and
>> the number of elements of the volume array).
>> If I scale up the final byte image I get a very bad image with
>> quantized grayscale...
>
>> Someone can help?
>
> It seems to me you are confusing data *display* with the
> actual data. The IMAGE_DATA keyword doesn't so much return
> a BYTE-type image as it returns a true-color rendition
> of what you displayed in the graphics window. What you displayed
> is NOT your data, it is a representation of your data, and
> that is exactly what you are getting back.
No David, I'm not confusing: I just need a 16bit/channel display
instead of an 8bit/channel display, i.e. I want a float-type image
representation of my 3d data.
If this is not possible, how can I always get a well-visible image of
my volume?
In fact the final grayscale depends on the volume sampling, volume
data ancd volume opacity: it is the same problem of visualizing an
image with TV, that is solved with TVSCL: how can I get a TVSCL-like
view of an emission volume, or automatically set the OPACITY_TABLE0 in
order to always span the full grayscale range?
Thanks
Gianluca
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
