opj_t1_getnmsedec_sig() method is passed (bitpos + T1_NMSEDEC_FRACBITS)
and then uses (bitpos - T1_NMSEDEC_FRACBITS) in its computations.
No need to add, and then subtract, T1_NMSEDEC_FRACBITS.
Likewise for opj_t1_getnmsedec_ref() method.
Original issue reported on code.google.com by boxe...@gmail.com on 11 Mar 2015 at 12:38
Original issue reported on code.google.com by
boxe...@gmail.com
on 11 Mar 2015 at 12:38