methinks johnkiv is talking about masking instead of cropping too.
Dude, crop tool in photoshop is to change the size/dimensions/content of your canvas. if you want multiple pics sitting inside there with their own 'cropped' dimensions, you're actually talking about masking.
"Dude", you're not necessarily talking about masking. You could be talking about compositing. The canvas is the largest size needed to accommodate all the images. You adjust and crop each image in the set to your requirements, and small enough so that all the images going onto one background canvas will fit with appropriate borders between them if that is what is wanted. Then you drag them one on top of another on the background canvas, and lay them out as you wish. It is only necessary that the bit depth and the resolution of the background and the layered images be the same. You temporarily flatten each image before dragging it onto the background canvas so that all the adjustments are conveyed. Once transferred, the original images can be saved with their layers intact so that cropped pixels are not lost in the originals. The composite can be saved with the individual image layers intact, and/or flattened into one flat composite for onward use if that is necessary. As long as the original images have been cropped preserving the pixels and saved with that layer intact, and the composite saved with its layers intact, the whole process is reversible by going back to these images for making changes and replacements. This is something I do routinely, and works quickly. It is even more efficiently done in Lightroom 4.1 than in Photoshop, because in LR you don't need to think about coherence of resolution and bit depth, and the whole history of each image is preserved and reversible automatically.