The fully digital front end for active speakers

[JonFo]

This thread is about options and implementations of pre-pro solutions that allow for advanced feature sets like Room Correction (DRC) and digital crossovers (DXO) in a single unit.

Several of us have been chasing the goal of having the highest performance, greatest flexibility pre-pro for fully active speaker systems.

The goal is to keep the signal in the digital domain for as long as possible, applying all the signal manipulation within a single box or a linked (clock synchronized) set of boxes to minimize any jitter-induced anomalies.

For 2 channel use, this goal is relatively achievable these days with multiple options, but for multichannel, there are fewer solutions on the market.

The signal chain we will be discussing is generally something along these lines:

• A digital source – Universal Disc Player, Network streamer, PC
• Input selection – HDMI switching, SPDIF Optical/coax, AES-EBU
• Signal Manipulation
  • Room Correction (Dirac, Audyssey, Trinnov, Convolution, etc.)
  • Gain, Delay and EQ
  • Digital crossovers
• Digital outputs
• Volume management signal
• Analog outputs (volume managed)
• Amps - digital direct drive (PowerDACs like NAD M2, TacT) or Analog
• Components of active speakers

The ideal outcome is to have as powerful and flexible a setup that’s combined with reasonable ease of use. These are often diametrically opposed to each other, as anyone who has used bleeding-edge tech knows

Some members here have made significant progress on this path and others like myself have achieved parts of this (but with too much complexity ). So I’d like to invite Ken (RUR) to share his config and experiences on this thread, and anyone else who has achieved the ‘one box’ goal.

Note: a key point of this discussion is integrating active-crossover speakers, as that’s where a lot of the complexity (as well as benefits) creeps in.

 

[RUR]

....with reasonable ease of use.

Aye, there's the rub

But, let's not get ahead of ourselves.......

Long ago and far away in 2008, I had an all-analog system. I thought it was absolutely terrific, but bass from the ML woofers was limited and, since the speakers were placed for best imaging, bad modal behavior was the rule of the day. "No worries!" I thought - "I'll just pick up a sub and place it so as to eliminate those pesky peaks and valleys". A fine plan, or so I thought, but what to do about a crossover? Well, Marchand made/makes an excellent analog device, so I bought one. Sure, it was an improvement, but I was still dissatisfied - and a single sub, no matter how well-placed, can only get you so far in mitigating modes.

OK, so now I needed another subwoofer, but I also wanted the ability to control delays so that the soundwaves @ crossover from the disparate transducers all arrived at the same time. And, I wanted the ability to use PEQ or other filters to finesse the results into the most linear behavior possible. miniDSP didn't exist, yet. Velodyne and a couple of other companies made simple PEQ boxes for subs, but their delays were fixed by design. What to do?

I'd done a ton of reading and, for me, it all pointed to TacT Audio. Pioneers in SoTA room correction, they had made terrific products since ca. 2000. I bought a gently used 2.2XP via A-gon, set it up, created a target curve, ran a calibration and......experienced an epiphany. Not only was the bass much better controlled, but the entire soundstage was instantly and quite considerably better defined. Drums were exactly there. Bass guitar was exactly over there, etc. Thanks to the now linear FR, masked frequencies became unmasked, and I was hearing the recording more clearly than ever before. SoTA frequency correction became something I would not do without.

Fast forward three years to March 2011. I'd learned the ins and outs of the (poorly documented) TacT, and I'd wrung every dB of improvement I could from its considerable capabilities. Meanwhile, the always-tenuous TacT company was on the wane and a french company, Trinnov, had introduced next-gen products with a number of improvements and a terrific reputation for SQ. So long TacT and hello Trinnov.

to be continued......

 

[Pneumonic]

Hi, Jonathan. Interesting thread. Timely too as I just bought a Sanders 10c speaker setup which is being run with DXO.

Setup chain is:

Lessloss DAC -> Sanders/Behringer DCX2496 -> Magtech (woofs) and Innersound (panels) amps -> Sanders 10c speakers.

Crossing over @ 172 Hz using 8th order, 48 dB/octave slopes.

Not a multichannel guy but, for 2ch, this setup not only affords me a great amount of flexibility (adjustable midrange/dynamic EQ along with time alignment of drivers, all in real time) but does so with incredible performance outcome. Once dialed-in, perhaps best I have ever experienced.

 

[Rich]

Timely too as I just bought a Sanders 10c speaker setup which is being run with DXO.

Congrats on purchasing the Sanders speakers, Pneumonic! Not to derail Jon's thread, but I would love it if you would post a review (in the off-topic A/V forum) of the speakers and any comparisons you can make to ML's electrostats. Sanders makes great products and I would love to read a review of his speakers from a former ML owner. When it comes time to replace my Summits, I may very well go with Roger Sanders' speakers.

 

[Gordon Gray]

Rich, as you know, I have great respect for Roger's work and for your opinions.

Regarding the speaker "shoot off", I suspect Roger's stat could very well perform better than ML within the very narrow (3" - 6") sweet spot.

Outside of that range, everything starts deteriorating rapidly. I know ML has issues with the sweet spot window but Sander's stat takes this limitation to the extreme.

If anyone who buys his product is aware of this issue, I think they've made a very wise decision.

Gordon

 

[Rich]

Honestly, Gordon, I think that Roger's speakers outside of the sweet spot don't sound much worse than ML speakers outside of the sweet spot. Inside the sweet spot, I think the Sanders speakers blow the ML's away by a long shot. I think the real difference is that there is a much bigger contrast between sweet spot vs non-sweet spot with the Sanders vs the MLs. Thus the difference becomes much more noticeable as you leave the sweet spot with the Sanders design. But I think this is because the sound in the sweet spot is so much better with the Sanders, rather than that the sound outside the sweet spot is any worse than the MLs. Unfortunately, I have never had a chance to compare them directly to each other to confirm this.

 

[Gordon Gray]

Valid perspective my friend. As with all things audio, YMMV.

Best,

Gordon

 

[JonFo]

to be continued......

Hi Ken, hopefully we can hear the rest.

I'm getting close to puling the trigger on a Dirac license to play around with. Just need to see a three consecutive weekend window open up, so far, no luck with that

Meantime, nothing notable at CES on this front. Hope we do not need to wait until CEDIA for the big boys to announce something. I know Harman had some very ambitious plans and then shelved the project. Maybe they will license the multi-sub optimization algorithms known as SoundField Mapping (SFM) for use in their BSS and/or DBX speaker processors. I'd hop on that in a nanosecond for my DriveRack 4800.

 

[JonFo]

Hi, Jonathan. Interesting thread. Timely too as I just bought a Sanders 10c speaker setup which is being run with DXO.

Setup chain is:

Lessloss DAC -> Sanders/Behringer DCX2496 -> Magtech (woofs) and Innersound (panels) amps -> Sanders 10c speakers.

Hi Pneumonic, Nice rig, that's the beauty of 2ch, it's much simpler to set up and configure. Still presents its challenges, but one can get to great results without too much effort. Glad you are enjoying it.
I'm always impressed when I get to hear a similar setup that Roger uses at trade shows. Generally one of the best sounding rooms at the events I've attended.

 

[RUR]

Hi Ken, hopefully we can hear the rest.

I'm getting close to puling the trigger on a Dirac license to play around with. Just need to see a three consecutive weekend window open up, so far, no luck with that

Meantime, nothing notable at CES on this front. Hope we do not need to wait until CEDIA for the big boys to announce something. I know Harman had some very ambitious plans and then shelved the project. Maybe they will license the multi-sub optimization algorithms known as SoundField Mapping (SFM) for use in their BSS and/or DBX speaker processors. I'd hop on that in a nanosecond for my DriveRack 4800.

Hey Jonathan,

Cool beans on the Dirac purchase! As a matter of fact, Dirac did provide some news at CES, where they demo'd Dirac Unity. Sorta like Harman's SFM, it uses all speakers to send compensating signals, on top of the primaries, to smooth response. A friend, who attended the show, was told that it operates up to 500Hz, but could go higher with enough DSP horsepower. He also said it sounded - and measured - terrific. Discussion and a link to a podcast with Matthias here, and more at Computer Audio, where username FLAC, a Dirac employee, discusses this and other things Dirac.

I'll try to get to the rest of my RC + active story later today. Much time-consuming chaos and destruction lately as we have remodeling done.

 

[Pneumonic]

Congrats on purchasing the Sanders speakers, Pneumonic! Not to derail Jon's thread, but I would love it if you would post a review (in the off-topic A/V forum) of the speakers and any comparisons you can make to ML's electrostats. Sanders makes great products and I would love to read a review of his speakers from a former ML owner. When it comes time to replace my Summits, I may very well go with Roger Sanders' speakers.

I will PM you a mini review, Rich.

Edit ..... sent.