Hi Christian,

Sorry for the big delay to answer, my exact need is the following:

I'm retrieving some image information directly from the memory, using the SYS(2600) function. My goal is to manipulate the image directly from the memory, avoiding to make certain GDI+ API calls, specifically on GIFs that are not fully supported bu GDI+.

I'm retrieving byte per byte, converting to ASCII, manipulate each byte separately, reconvert using CHR(), and send back to memory.

In the case of a picture of 1280 x 960, I have 3686400 bytes to manipulate individually.
This task is taking about 3 seconds in my machine.

I was thinking if your libary can convert all the information in an array, and, after I manipulate this VFP array, the library rebuilds the array to send back to the memory...
Do you think using your classes I can have a gain of performance ?

If positive, how can I do that ?

TIA

Cesar Chalom
PS: Hope to meet your team again in the finals ! :-)))