Is your feature request related to a problem? Please specify.

For some viral sequencing experiments, the number of reads can reach the millions mapped to a small region. Generating a consensus for such a bam file even with extra cpus can take a while. There is an option for --min-depth but none for maximum depth.

Describe the solution you would like.

Similar to samtools mpileup which has --max-depth, it would be super helpful for samtools consensus to have --max-depth option as well (short form -D because -d is taken for --min-depth) .

A temporary solution would be to use samtools view --subsample to reduce the number of reads but it's not ideal as it could subsample regions with low coverage.