I am looking for advice on best practices.  I am coding against a
digital camera SDK.  One of the callback functions returns the
large, multi megabyte images in smaller chunks.  Some of the
images will be JPG, others will be RAW.  

GUI is C#
Middle layer:  MC++
Digital Camera SDK: UMC SDK

The MC++ layer will gather all the different pieces of the images
and put them together in one _______.  The question is:  How best
to gather all those pieces as to send them up to the C# GUI?  When
the pieces are a JPG, C# is going to want to be able to turn it
into an Image object, when the pieces are a RAW image, the C# will
simply save the image to the hard drive.  

Sam