Considering going active on my current speakers (apogee centaur minors). Based on some other active projects with their big brothers it seems that a lower x-over point with LR4 electrical slopes is a good way to go. What I cannot find is any discussion over whether you need to put in a 360degree allpass delay on the high frequencies with LR4. There seems to be a consensus that we cannot hear a single cycle of anything but no conclusion that we therefore cannot hear a 360degree delay.
Does anyone have any experience of this?
Does anyone have any experience of this?
A delay of a single cycle could be audible, it greatly depends what frequency, and the corresponding time delay. It is measurable, I'm sure someone has tried that, but audible??? Contentious territory.
In my active two way, using 4th order active, i cannot hear the delay. The crossover is at around 2.8k so that's 0.35 milliseconds. Some may say I'm deaf, but i think an all pass to remove the cycle / 360* delay, would be more trouble than its worth. (i looked at it, and i needed 6-8 orders to do it, and the phase issues that caused put me off it entirely).
In my active two way, using 4th order active, i cannot hear the delay. The crossover is at around 2.8k so that's 0.35 milliseconds. Some may say I'm deaf, but i think an all pass to remove the cycle / 360* delay, would be more trouble than its worth. (i looked at it, and i needed 6-8 orders to do it, and the phase issues that caused put me off it entirely).
Adding an all-pass filter will render a crossover that most likely destroys the manifold advantages imparted by the LR filters (and other crossovers of like derivation).
Whilst LR2 (and like) crossovers result in a more pleasing delay characteristic, their audible advantages in two- and even three-way speakers are usually limited by the need to curtail the more audibly detrimental artefacts caused by operating drivers at their frequency extremes.
Note, however, that this discussion is separate from that of using a first-order all-pass filter as Linkwitz also describes in his 1976 JAES paper to compensate for the "lobing errors" produced by non-coincident drivers.
Check out Lipshitz and Vanderkooy's series of AES papers for a thorough guide to crossovers and (in particular) the addition of delay elements.
Whilst LR2 (and like) crossovers result in a more pleasing delay characteristic, their audible advantages in two- and even three-way speakers are usually limited by the need to curtail the more audibly detrimental artefacts caused by operating drivers at their frequency extremes.
Note, however, that this discussion is separate from that of using a first-order all-pass filter as Linkwitz also describes in his 1976 JAES paper to compensate for the "lobing errors" produced by non-coincident drivers.
Check out Lipshitz and Vanderkooy's series of AES papers for a thorough guide to crossovers and (in particular) the addition of delay elements.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.