Screen Resolution and Pixelchecksum

[ofLight]

When you check more complex arrays with pixelchecksum in a low color mode environment like 256 you recieve speratic results.

As of now the only fix i can think of is to change color mode up to 16bit, then using Pixelchecksum, and then rersetting color mode back down to 256.

I am trying to find a way around this but have been unsuccessfull so far. Any suggestions whould be greatly apreciated.

Edited October 27, 2006 by ofLight

[novatek]

Hi!

I have found this too.

It happens the PixelChecksum returns the same number on two different 256 color surfaces [EDIT: NOT TRUE really , see posts below]:

- 32-bit color mode was Enabled at the time of checksum

- first checksum was made in 256 color DOS window

- second checksum was made over a grey surface near Autoit menu (which is also 256 color, I suppose)

- dimensions of both checksums were the same (the same width and height)

Is the algorythm for PixelChecksum available anywhere?

Edited November 12, 2006 by novatek

[Valik]

How will having the algorithm help anything?

[novatek]

Oh, Just thinking how to avoid identical numbers?

The width and height was 72 x 12 pixles.

Maybe with more pixles...

But should be noticed somwhere at the help pages, that algorythm is not proof as CRC or MD5?

[Valik]

With the astounding proof you are providing, how can we not do something about this?

Seriously, you want anything done? Give me images that are supposed to be different but produce the same checksum.

[novatek]

Hey!

Sorry, it was mistake.

I have done a prove test and found it was mistake.

It happend due to wrong set option "PixelCoordMode".

Therefore checksum from window behind.

Thanks for your patience Valik and sorry again!