----- Original Message ----- From: "Qkano" : On further thought, I'm a bit surprised by this test as a test. : "If I were gonna do it" ... : : Firstly, the test images would be available as TIFFs for download. : : Second, the image fragments would be identically sized and cropped. Since jpeg is essential, random choices of which bits to show could have visible differences. : other bits clipped. as you say, speed is oftern the deciding factor but given 'genuine' fractals *costs* and irfanview doesn't, and given the results and the fact that differences are difficult to ascertain - it'd be very hard to justify spending $$ on gf when everything suggests it's not any better than irfanview! I have o agree with you too - that 'test' is an abomination! nice for a rough overview but a test - I think not. random crops seem more than a little messy k
