Split colour channels

Four PNGs, one per channel, named after your file with -red, -green, -blue and -alpha appended. Each one is grayscale: that channel's value is copied into all three colour components, and the alpha of the output is set fully opaque so you see the plane rather than a transparent image. All four are the same size as the original.

Because the image has no transparency. Decoding gives every pixel an alpha of 255 when the source has no alpha channel, so the plane comes out uniformly white. That is a correct result, not a failure. A varied alpha image means real transparency data, which is worth a look since it is a place people hide things.

Spotting a difference that is invisible in the combined image. Text pasted into one channel, a region edited in only the blue plane, or a payload written into a channel that barely affects the visible colour all stand out when the channels are viewed on their own. It is a fast first pass before Bit-plane viewer, which goes one bit deeper.

JPG, PNG and WebP, selected by the file extension rather than by the bytes. Anything else is refused with a message naming your file and stating that only JPG, PNG and WebP are supported. Convert other formats first with Convert image. A renamed file fails the same way, so check what it is with Detect file type.

Yes. Each plane is written as PNG, so the channel values you see are exactly the values that were decoded from the input. If the input was a JPEG, those decoded values are already the result of lossy compression, so the plane is a faithful copy of a lossy image rather than of whatever the camera or the editor originally wrote.