The problem you are having is that there is a lot of DNA in your samples and you are not getting complete digest. Or there are two plasmids in the preparation. This is obvious when you see the lower band in lanes 2-4 for example is much brighter than most of the higher bands. But if there was complete digest the larger bands should be brighter if in the same molar concentration.
The firs thing I would try is to use much less DNA in your digestion (maybe 3-fold less) and see if that helps.
Really can't interpret this without answers to Sofiane Benyamina questions. Also what is the expected size of the insert and the whole plasmid? What marker is that (fragment sizes)? Have you tried running uncut plasmid on the same gel with the digests? Running uncut plasmid is often overlooked, but can be an important control as it can detect "dimer" insert plasmids. And Michael J. Benedik is right that in a complete digest pattern, the brightness of ALL the bands would be proportional to their size. So faint bands near the top of the gel are partial digest products of some kind, either incomplete products of SacI and BamHI or possibly a "star" activity of one of these.