How it started

I admit I was uneasy at first, waiting for someone else to work on it but I was surprised no one else was working on it.

Following instructions from the original developers, I tried creating my own solution but with low success because I didn't have exact offset of the first image I had decoded.

After teaming up with RenanSPH, who knows much more about disassembling than I do, he found the instructions that copy the data and pass to the GDI to be paint on screen.

The first sprite he detected with exact precision was the position and length of the raw data representing the publisher's logo. After I discovered its width and height, it was easy to identify that the length was the product of multiplying width by height.
By knowing that, he found the entry on the table as the instruction had as parameter.

Following that, I'm currently working on a decoder. I already have some good results that I will publish later.