DSP Active Crossover

I'm considering trying a DSP-based active crossover in my system. I did a search to see how much this has been discussed, and most of the posts are pretty old or about active speakers. DSP technology has changed a fair amount in the last 15-20 years.

My system is digital only, and my speakers are 3-way, so it's not particularly complicated. I've been looking at the Danville Signal dspNexux 2/8 which has two channel in (with digital inputs) and eight balanced analog outputs. This appears to be available with AKM AK4499 DACs which are fairly well regarded sigma-delta DACs (although I don't know how good their implementation is).

This product has a fairly rich DSP software environment for programming filters, time delays, etc., so it should be fairly straightforward to set it up to replace my passive crossovers. 

My biggest reservations are 1) giving up my Denafrips Terminator+ DAC and nice-quality DIY preamp, and 2) using the DAC's digital volume adjustments. 

This unit is about $3K (maybe a bit more with the AK4499 DACs), so isn't terribly expensive. From the limited research I've done, this unit appears to be higher sound quality than the miniDSP or DEQX boxes, but I could be wrong. All my amps have balanced inputs, so I'd prefer to use a unit with balanced outputs. 

So, what I'm wondering is if the benefits of active crossovers and dsp equalization will outweigh the lesser DAC quality (assuming this is the case) and lack of analog volume control (currently using a relay switched attenuator). I'm also wondering if there are other dsp audio processors that I should consider (digital inputs, at least six channels out, ideally with balanced outputs).


@ricevs - I have been using the dspNexus as the DAC/preamp, so I'm feeding it USB audio directly (no conversion from analog). I don't think it makes sense to use my Terminator + and preamp ahead of the dspNexus. 

I agree that the dspNexus could be improved. I think the biggest weakness is the power supply - the simple SMPS that is currently used could be improved a lot with a good linear supply with shunt regulation. I think if I decide to go this route, I will probably implement my own power supply. 

The biggest advantage I see to use a DSP crossover is to be able to implement steep filters with no added noise. If I implement analog active crossovers, the steepest I'd probably be able to get away with is 24db/octave. This is still better than the 12db/octave passive crossovers I have now though. 

I agree there is a bit of room for improvement in my current crossovers, but this is clearly diminishing returns. Still, I may consider it at some point. I've just got too many projects going on in parallel right now, and I'm having trouble actually finishing any of them 😀.

As far as facing all the drivers in the woofer towers forward, I have not been able to discern a difference. I had them facing forward in my triple towers and I think the quad towers are just as tight and clean sounding - probably a bit more so - than the triples. There isn't an easy way to turn them around anyway. All the holes for the wires are only drilled on one side. 

With such a tall tower, I'd be a bit nervous about having all the drivers facing forward since the weight distribution would be heavily toward the rear. Even the triples were slightly tippy. 

Hi Everyone,

I would like to discuss the choice of the switching power supply in the dspNexus. Commercial switching power supplies solve a lot of practical problems when converting AC mains to low voltage supplies in a commercial product. They are not ideal.

I evaluated 4 different switchers during the development process of the dspNexus. Most suffer by having insufficient current capability with the negative supply. This was clearly audible with the headphone amplifier even with 100 watt units. The one I chose was good in this regard. It is the only board in the dspNexus that I did not design.

That said, a large linear mains supply is not a panacea. There are three linear supplies between the DACs, ADC and Clocks. These are not all LDOs since high frequency power supply rejection is also important.  The goal is to have a low noise, low impedance supply located near each target device.

Designing good audio products is a process of making many small decisions. Collectively, the choices you make determine the outcome. I started making high end consumer audio products 45 years ago as an analog engineer. I have used DSPs for 30 years, so I have a bit of experience.  I welcome discussing design philosophies and I hope many of you will consider the dspNexus as a good solution for your systems. I will be at CAF and I hope to meet many of you there.

Al Clark
Danville Signal


@dsp - Al, I didn't mean to imply that the dspNexus isn't a well engineered product.  I would expect that replacing the SMPS with a linear supply that was good enough to provide a noticeable improvement to the sound quality would more than double the cost and significantly increase the size of the enclosure (also increasing cost).

As is, the dspNexus provides a lot of value for it's price point and is very nicely constructed, allowing easy upgradeability to the DSP and DAC modules. 

I look forward to meeting you at CAF. 


I am hal2010, AKA Richard Hollis.

If you have any questions about DSP crossovers for the dspNexus 2x8 system, will be happy to answer them.

Best regards,


I've heard several systems with the mini DSP, and have been unimpressed. 

I've also heard several very high end DSP systems, including the Kyron. And that system kills! Even Mr. analog, Michael Fremer loves it.

But my overall evaluation is, that it takes a very special DSP speaker system to best the best passive system.