I thought I commented on this already… Guess not. But anyways that sounds about right. I2G is going to write code for every pixel and PEP5 is going to ignore a very large amount of the picture. I think it is also more efficient in how it writes the code. I have experienced this kinda thing as well.