The point I was trying ot make is about how you get to the result. You could probably get the same final, correct rectilnear result using PS Photomerge, but you'd have to; correct lens distortion> merge> correct perspective, or correct lens distortion and perspective> merge. That is, you would have to make cut-and-try attempts at a good result, correcting lens distortion and perspective by way of small proxy images with grids overlayed, re-trying being the only way to refine the results. And do two or three transformations. PT stitchers let you, for rectileaner output for example, lets you pick what should be horizontal, vertical, or straight, and then optimizes lens distortion correction, output projection, and perspective, for correct output, for all imput images all at the same time.That is; the PTAssembler result is from one step (albeit, a longer step.) And it alows you to optimize any creative perspective changes that yould like, in that same one step. For stitching 36 images into one rectilnear pic, the image I posted at the top of the page, that's a really big deal... I mostly do landscapes, and now having mercator output, for tall mosaics, would be a deal maker to me... multi-plane rectilnear I've used or a couple of architecturals too... And morph-to-fit controled morphing for the paralax problems. And to optimize different FOVs for input images, any one or more individually, for different focus settings or different zoom settings images into the same stitched image.