Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Matt]

I've heard from several people that are doing speaker cross-overs with MC.  It's possible to do this well in MC today using Parametric Equalizer.

However, I'm wondering if a new filter would make it easier for people.  I'm imagining picking an input channel, whether you're bi-amping or tri-amping, then picking the output channels, frequencies, and slopes for each range.

It seems like one filter could take the place of several, making it easier for people to setup a bi-amp / tri-amp configuration.  The change is probably not too hard for JRiver since it's just a new user interface on the existing algorithms.

Would this be helpful to anyone?  If so, any advice?

Thanks.

[mwillems]

I've heard from several people that are doing speaker cross-overs with MC.  It's possible to do this well in MC today using Parametric Equalizer.

However, I'm wondering if a new filter would make it easier for people.  I'm imagining picking an input channel, whether you're bi-amping or tri-amping, then picking the output channels, frequencies, and slopes for each range.

It seems like one filter could take the place of several, making it easier for people to setup a bi-amp / tri-amp configuration.  The change is probably not too hard for JRiver since it's just a new user interface on the existing algorithms.

Would this be helpful to anyone?  If so, any advice?

Thanks.

I currently use JRiver for bi-amping, and I love it.  I think an integrated filter sounds great, and would be helpful for people getting set up for the first time. I think there are one or two things that might be important to consider as part of the implementation:

1) An all-in-one filter should probably have a Linkwitz-Riley filter option.  The current crossover filters in PEQ are butterworth filters, but symmetrical butterworth filters do not sum flat at the crossover point.  There's a 3dB bump at the crossover point if you use a butterworth low pass and a butterworth high pass with their filter "knees" set to the same place (and potentially some undesirable phase consequences depending on the order of the filter).  Some people account for this by using asymmetrical filters (butterworths of different orders and/or different set points) with subtractive delay.  But that's really fiddly, and is why most speaker systems don't use Butterworth crossovers.  Flat frequency response and phase coherence across the crossover region are very important, which is why the far and away most popular way to do a symmetrical crossover is an even order Linkwitz Riley filter (4th order being the most common).  Symmetrical LR filters sum flat at the crossover point and (with a 4th order filter) are phase coherent.  In the current PEQ, this isn't hard to do because you can create a 4th order LR by stacking two identical 2nd order Butterworths, but if you create your proposed integrated all-in-one filter stacking might or might not be possible.  If you implement a "bi-amp" or "tri-amp" filter, I think it's important to provide a choice between butterworth and LR topographies along with the choice of slope.  That choice wouldn't require any new programming as LRs are just stacked butterworths.  I, personally, would like to see some additional, not currently achievable in JRiver, crossover filter options (3rd order butterworths, Bessel filters, etc.), but I think an all-in-one filter should support, at a minimum, LR filters because they are (probably) the most popular crossover, and the 4th order LR is the crossover most likely to work correctly "out of the box" without additional filtering (i.e. should probably be the default option for a bi-amp filter).  Obviously I'm glossing over some things and simplifying (it's more complicated than that with a real speaker), but you get the idea.

2) I think it's important to retain the ability to set the crossover frequencies, slopes, etc. separately for each filter portion (high pass, low pass, etc.).  It sounded from your description that you had planned to implement it that way, but I wanted to register a vote for that approach.  It's useful to be able to create asymmetrical filters, provided you can get the delay sorted out, which brings us to:

3) Almost all bi-amped or tri-amped systems need some delay.  Many crossover filters need some delay to work correctly (4th order LR being the main exception), and delay is needed to account for different driver spacing. So it might make sense to have an integrated "delay" setting as one of the configurable options for the bi-amp/tri-amp filter.  Obviously this is easily done in PEQ, but if the goal is to create "one stop shopping" for bi and tri-amping, delay is probably worth including.  

Just a few thoughts off the top of my head.  

[hulkss]

I agree, that is the way to do it.

[InflatableMouse]

While you're at it, would it be possible to show the filter in the analyzer (graph line) with a tickbox to enable/disable please?

[Wungun]

It you can find LSFilter, it's the best crossover VST I've tried...4-way, a slew of slope/dB settings, all the popular alogorithms....