so i guess, you are trying to use it for comb-filtering effects? for future developments, i actually have it on the agenda. the main reason for leaving it out was efficiency - i currently round the delay-time to the nearest integer number of samples - for delay-times that are somewhat large (large enough to make for seperately perceived sonic events for each delayed copy), the delaytime error does not really matter because the relative delaytime error will still be small in this case. but when we go down to delaytimes that are really small (entering the realm of comb-filtering), i would have to go with sub-sample precise delaytimes, which in turn calls for interpolated delaylines, making the whole process much more computationally expensive.Reflected wrote:It will be useful to remove the 10ms (and 0.0100 beats) limit.
but at some stage, i probably want to include modulation capabilities anyway, so i'll have to implement interpolation anyway. i guess, i could make it optional per delayline, so we can still benefit from the efficient implementation when modulation (i.e. sub-sample precise delaytimes) is not really required. we'll see...
