Graeme Gill wrote: > Nikolay Pokhilchenko wrote: > > > colprof -v -al -qu -iD50 -o1931_2 (-f) -bn -nc preliminary > > > > So there isn't a bug, because I didn't include *.ti3 data. > > OK - is there any reason you're using -nc + applycal rather than simply > letting the calibration flow though ? Yes, the reason is the volume of *.ti3 data vith high resolution spectrum. In some cases with more than thousand patches the *.ti3 data are 1.5 or even more megabytes by size. Since I can do a couple of profiles while experimenting, I'd like to not include ti3 data into the profile. > I don't think -bn is doing anything though, as it only applies > to input profiles. -bl is better if you want to reduce B2A table > generation (or the undocumented -B). May be there was a bug with 1.1.0_RC1, because with -qu -bl options the colprof seems to do -bu-quality B2A table. And it was impossible to wait till colprof -qu -bl completes the profile. I didn't check it the RC2, but with v1.0.4 -bl have work as it should.