If you only need to quantify expression at the transcript/gene level, use Salmon or Kallisto. If you need to align reads to the transcriptome or genome go with STAR.

Not sure about the best aligner, but there are some aligners that widely used, such as STAR or HISAT2. I saw some article comparing aligners using simulating data. If I remember it correctly, GSNAP and STAR did pretty good jobs.