Will require reading a chuck of data to have enough different scores to determine encoding - if possible. Have a read of https://en.wikipedia.org/wiki/FASTQ_format