Good loci lost during phasing

I noticed that some of my samples that sequenced at highest coverage have several loci missing after phasing. After some testing with samtools, I think the problem is that by default max coverage is set to 256 (flag -D). It seems SECAPR code does not change this flag, so maybe setting it to something higher will eliminate the problem.