There's been a lot of threads about this with LOTS of specific information. I'd suggest you do a search for "panorama stitching" or "stitching" here and also on google for other articles and sites.
Briefly, you can do perfectly good hand held shots for stitching if there's no close subject matter. If there's close subject matter you have to use a tripod and an accurately calibrated panorama head. Otherwise you'll get parallax resulting in double imaging of the close subject matter. This can not be corrected in any manner except (generally inefficiently and imperfectly) manually with the cloning tool in Photoshop.
There are many claims about easy ways to do panorama shooting and easy quick software that works just fine. Due to my own exhaustive pano stitching experience and testing almost all the software there is for both PC and Mac I don't believe these claims. You may luck out now and then with just the right lens and just the right subject matter, but to get consistently good panoramas stitches with a variety of lenses takes the right software and very careful technique. Some people have also quite obviously not looked carefully at their stitch lines at high enough magnification to detect misalignments. This is OK of you don't print large, but otherwise you'll see the misalignments in prints. You absolutely can't judge stitch alignment on a web picture, no matter how nice that picture looks.
I now do pano stitching on a routine daily basis. Almost everything I shoot is for stitching. I've concluded that PTMac is the best program, and of everything I've tested, the only program that works consistently. Since I only tested PT Assembler briefly quite a while ago, I assume this is now also a good program, being based on the same foundation of Panorama Tools. Both these programs have a considerable learning curve, but I think there's just no way around that if you want consistently perfect stitches with various lenses.
Good luck. It's not an easy road either for hardware or software considerations, but it's worth it.