Hi,
I think it's the first one. If concern was the time / phase alignment then one would be better of with DSP which can make it what ever you want, for some particular listening axis. If concern is frequency response anomaly due to the crossover phase mismatch, there is lot more effect due to edge diffraction (no roundovers) so to me stepped baffle like this seems unnecessary tinkering and marketing trick.
I think it's also important to try and listen how would small phase mismatch, or frequency response issues, be audible. In my own room with my own speakers with my current listening skill, if I listen too far the tweeter can be delayed multiple milliseconds with hardly any audible effect. Get listen closer up where room early reflections reduce from perception and now it's more audible. Considering a lot of images of setups seen on forums, also setups I saw in a hifi show, most people have speakers setup so that there is so much room sound I think the time alignment really isn't that audible. Perhaps some listen close enough, or perhaps power response gets bit better so sound on far also improves, which then affects perceived sound on these room sound setups. But to get better power response even better with those same drivers you could use some other tricks, like c-c distance and facets I think.
As disclaimer, I have no such speaker and the above is based on logic only.
For crossovers, one cannot decide crossover before hand as it must be tailored with the physical construct and measurements of it. A crossover can be optimized for any set of measurements, so the speaker quality depends on how it measures (before crossover in place) and after that what crossover you implement. It really is trivial to come up with any crossover with VituixCAD, takes some time and effort to get good measurements, but eventually the sound quality is not from crossover. However, it is of course fine goal try and make crossover such that it makes the system sound as good as possible, which phase matching of acoustic LR4 filter provides, but it doesn't mean something else couldn't be even more suitable in your application.
While crossover could certainly ruin sound it cannot change the physical structure of the speaker better, which is the real limiting factor. The physical structure is something you must figure out before considering a crossover at all. You must make speaker that measures good, and then crossover falls in place and is no kind of an issue. And, before you can make speaker that measures good you must know what measurement sounds good to you in your practical situation, your room, how your auditory system affects things and so on. Crossover is about the last step building speakers and no problem at all, it's much more harder to come up with proper design that measures well which eventually sounds good to you in your room with everything in place.
In case you already have some drivers, you might try and figure out what kind of a system they could suit and whether that system suits you. If not, sell them and start from scratch. If you must use them, you could build something, measure, make crossover and if it sounds fine it's fine. If your ambition is higher than this, consider building prototypes until measurements are up to bar and the system sounds like you wish.
Have fun 🙂
Last edited:
I have never had any need to invert the polarity in a typical 2 way design, except with Tannoy's coaxials.
The right polarity is the one that gives you the fullest sounding experience. I can hear it by listening to
an MLS signal.
The right polarity is the one that gives you the fullest sounding experience. I can hear it by listening to
an MLS signal.
DSP can be used, but I wasn't thinking of that in this particular case, but it's also not a perfect solution as off axis any digital delay in the tweeter will lead to errors which wouldn't happen if the (as an example) driver magnets were actually aligned.
@Lojzek No, neither have I but I've had to pick the filter order based on the phase match, rather than picking a filter order and then adjusting for phase, but an inverted tweeter in a 2-way is common.
@Lojzek No, neither have I but I've had to pick the filter order based on the phase match, rather than picking a filter order and then adjusting for phase, but an inverted tweeter in a 2-way is common.