This is misleading, as it may actually have a message

"seam to have" Of course it doesn't know for certain, everyone would know you're embedding messages if it did.

whether you have to use the same method for decoding as was used to code it.

Yes. I don't know of any desktop applications that decode Pixelknot images which is what I was hoping to find out. I still haven't found any PNG's by Q that appear to be stenanographs. They will have jpeg like artifacts in them. If you zoom in on my image you will see noise in the image, blacks aren't solid, whites aren't solid. That's the stenanography slightly shifting the color of the pixels to store information. It appears to be lossy but a png isn't lossy.

.win seems to be really bad at detecting carriage returns.

This is misleading, as it may actually have a message "seam to have" Of course it doesn't know for certain, everyone would know you're embedding messages if it did.

whether you have to use the same method for decoding as was used to code it. Yes. I don't know of any desktop applications that decode Pixelknot images which is what I was hoping to find out. I still haven't found any PNG's by Q that appear to be stenanographs. They will have jpeg like artifacts in them. If you zoom in on my image you will see noise in the image, blacks aren't solid, whites aren't solid. That's the stenanography slightly shifting the color of the pixels to store information. It appears to be lossy but a png isn't lossy.