Are Active sytems making themselves irrelevant?

jfritzen posted:

That's approximately what Linn Exakt does. They have their own version of SPDIF however, it's the Exakt protocol and its transported over Ethernet cables.

I'm well aware of that.

Also, having tried various methods of applying mathematical DSP filters to the data in files (and comparing them to the unedited versions). I'm fully aware of the audio degradation caused by the application of mathematical processing to the data - just because it's a DSP based system doesn't mean it sounds better than analogue filters - nothing comes for free!

[@mention:1566878604016895] took the words out of my mouth. Indeed, Linn and Devialet keep the signal digital until the crossover (provided your source is also digital).

The analogue solutions from Bryston and Rane would be more suited for use with Naim since they work purely in the analogue realm. There shouldn't be any problem with their usage since they would, in theory, operate on a signal already optimised for a Naim power amp by a Naim preamp, and perform better than the passive crossovers in whatever speakers regardless of the fact they use lower end PSUs inside the units.

However, looking into it in more detail, the Bryston solution would only be suitable for speakers that had perfectly aligned mechanical stop points for all drivers. And that is rarely the case. The PMC Twenty/5 ranges acheive this by using a sloped fascia which puts all drive units on the same approximate virtical plane. But for anything else, you need, not only high and low pass filters, but a delay control also and information from the manufacturer about what the delay for each drive unit should be. 

It's all a bit of a minefield and one I think Naim could simplify by rethinking SNAXO from scratch as such a flexible product coupled with Naim's prowess in PSU isolation. I'd love to see a speaker agnostic SNAXO (full width 3-way Supercapable) and SNAXO-lite (shoebox, 2-way, HiCapable). The market for such a product would not be huge (the response to this thread tells us that - it is mainly me talking to myself). But being fully user configurable would lower the service and R&D cost to Naim also since they would not need to configure units at the factory or work on speaker specific implementations.

Wishful thinking.

Adding a high resolution digital delay line via DSP is trivial.

Huge posted:

jfritzen posted:

That's approximately what Linn Exakt does. They have their own version of SPDIF however, it's the Exakt protocol and its transported over Ethernet cables.

I'm well aware of that.

Also, having tried various methods of applying mathematical DSP filters to the data in files (and comparing them to the unedited versions). I'm fully aware of the audio degradation caused by the application of mathematical processing to the data - just because it's a DSP based system doesn't mean it sounds better than analogue filters - nothing comes for free!

I've heard some Linn Exakt systems and I think that Linn do know their digital stuff very well. Plus there is Space Optimisation too. In fact, regarding the digital domain, I don't think there are many companies that can compete with them. It's their amps and speakers I'm not a big fan of.

It doesn't matter how well they "know their digital stuff" the limits are in the maths that underlies the physics involved - you just can't escape the negative consequence, and it doesn't matter whether you use analogue or digital processing; the laws of physics still apply.

Perhaps one should tell Linn that they are on the wrong path ...

Not at all what I said - they just don't have a 'Magik' solution - it's constrained by physics exaktly like every other solution!

I don't think anyone is disagreeing with you Huge...

feeling_zen posted:

.

However, looking into it in more detail, the Bryston solution would only be suitable for speakers that had perfectly aligned mechanical stop points for all drivers. And that is rarely the case. The PMC Twenty/5 ranges acheive this by using a sloped fascia which puts all drive units on the same approximate virtical plane. But for anything else, you need, not only high and low pass filters, but a delay control also and information from the manufacturer about what the delay for each drive unit should be. 

It's all a bit of a minefield and one I think Naim could simplify by rethinking SNAXO from scratch as such a flexible product coupled with Naim's prowess in PSU isolation. I'd love to see a speaker agnostic SNAXO (full width 3-way Supercapable) and SNAXO-lite (shoebox, 2-way, HiCapable). .

 

Essentially making the Naim equivalent of the old ATC EC23, which sadly seems not to be in production anymore, or at least not available to the domestic market.

But you don’t need information from the speaker manufacturer re delays, as it is very simple to determine with sound monitoring software (e.g. REW) and a measuring microphone.

Huge posted:

The current situation for a lot of active systems with digital crossovers is a dog's dinner:

Digital source -> DAC -> preamp -> ADC -> DSP Crossover -> multiple DACs -> multiple poweramps -> drive units
just doesn't make any sense.

If you want to do digital crossovers, the only sensible solution is:
Digital source (S/Pdif) -> DSP signal control / Crossover  -> multiple DACs -> multiple poweramps -> drive units

Re final point, ideally, yes - and for me that would effectively mean 3 Daves (ouch!). However, with a stand-alone digital crossover the ADC and DAC filters can be precisely matched, which I gather maximises the fidelity of the processes. Whilst not ideal, I disagree with your suggestion that it doesn’t make any sense: In any given system, if speakers driven  actiively using a digital crossover sounds better than driven passively it makes sense. SimilarLy if a digital crossover sounds better than an analog one at whatever is the affordable price point (and sounds better than passive), it makes sense.

Active systems with amps and DSP in the speakers (Dutch & Dutch and Kii) are the way forward, providing time alignment and cartoid dispersion. The old naim approach looks to be less relevant going forward unless you have a lot of naim amps and money. My next move will probably be to active speakers. But then I don't have a large investment in naim amps, only the source.

Huge posted:

Also, having tried various methods of applying mathematical DSP filters to the data in files (and comparing them to the unedited versions). I'm fully aware of the audio degradation caused by the application of mathematical processing to the data - just because it's a DSP based system doesn't mean it sounds better than analogue filters - nothing comes for free!

With regard to artefacts or other degradation arising from the DSP process, what matters is whether they are audible, and if so whether they are greater or less than any audible degradation of the sound by an analog active XO or a passive XO.

(reposted with typos corrected)

feeling_zen posted:

[@mention:1566878604016895] took the words out of my mouth. Indeed, Linn and Devialet keep the signal digital until the crossover (provided your source is also digital).

The analogue solutions from Bryston and Rane would be more suited for use with Naim since they work purely in the analogue realm. There shouldn't be any problem with their usage since they would, in theory, operate on a signal already optimised for a Naim power amp by a Naim preamp, and perform better than the passive crossovers in whatever speakers regardless of the fact they use lower end PSUs inside the units.

However, looking into it in more detail, the Bryston solution would only be suitable for speakers that had perfectly aligned mechanical stop points for all drivers. And that is rarely the case. The PMC Twenty/5 ranges acheive this by using a sloped fascia which puts all drive units on the same approximate virtical plane. But for anything else, you need, not only high and low pass filters, but a delay control also and information from the manufacturer about what the delay for each drive unit should be. 

It's all a bit of a minefield and one I think Naim could simplify by rethinking SNAXO from scratch as such a flexible product coupled with Naim's prowess in PSU isolation. I'd love to see a speaker agnostic SNAXO (full width 3-way Supercapable) and SNAXO-lite (shoebox, 2-way, HiCapable). The market for such a product would not be huge (the response to this thread tells us that - it is mainly me talking to myself). But being fully user configurable would lower the service and R&D cost to Naim also since they would not need to configure units at the factory or work on speaker specific implementations.

Wishful thinking.

 

 

The new Bryston BAX-1 digital crossover can be used between Naim preamplifiers and Naim amplifiers but it is programmed at the factory for  specific Bryston active loudspeakers. However, in the UK Bryston is distributed by PMC so the BAX-1 could probably be programmed for PMC speakers if asked.

I just looked the Bryston unit up, it says it takes the analogue signal converts to digital, does the DSP corrections for the crossover and then back to digital. I wonder if it sounds any good, anybody heard it?

Dozey posted:

Active systems with amps and DSP in the speakers (Dutch & Dutch and Kii) are the way forward, providing time alignment and cartoid dispersion. 

Don’t forget the company that has been doing that in the domestic environment since 1989... Meridian.

Gazza posted:

I just looked the Bryston unit up, it says it takes the analogue signal converts to digital, does the DSP corrections for the crossover and then back to digital. I wonder if it sounds any good, anybody heard it?

No idea how the Bryston sounds... but there’s quite a lot of options for doing that mostly in the “pro audio” environment.

Gazza posted:

I just looked the Bryston unit up, it says it takes the analogue signal converts to digital, does the DSP corrections for the crossover and then back to digital. I wonder if it sounds any good, anybody heard it?

Not heard it, but their amps are excellent, and Bryston have a very good reputation for quality.

Notwithstanding Huge’s criticism of digital crossovers, they are mainstream these days, and widely used in the professional world - though that of course doesn’t mean that they would necessarily meet the demands of audiophiles. And regarding converting to digital and back, bear in mind that, quite apart from what Linn does, Naim now also do that in some products, though not (yet, AFAIK) applying DSP outside that of their streamer/DAC products.

As I mentioned, I have a Behringer digital XO and plan to compare with the ATC analog one some time - but whilst that comparison will require a lot of time, I can say that active with the Behringer sounds better than passive (PMC EB1i speakers).

The reason I question, is it tried a Naim Core/Chord Qutest into the Nova, seemed to suck some of the dynamics from the music. Then moved on to Core/Chord Dave into Nova.....sounded pretty much like a Nova, but £8500 dearer. Put it down to the ADC then DAC of the Nova.......the Chord lives to fight another demo, maybe, some day.

Gazza posted:

The reason I question, is it tried a Naim Core/Chord Qutest into the Nova, seemed to suck some of the dynamics from the music. Then moved on to Core/Chord Dave into Nova.....sounded pretty much like a Nova, but £8500 dearer. Put it down to the ADC then DAC of the Nova.......the Chord lives to fight another demo, maybe, some day.

Hmmm interesting. Have you tried Dave direct into your power amp?

cat345 posted:

feeling_zen posted:

[@mention:1566878604016895] took the words out of my mouth. Indeed, Linn and Devialet keep the signal digital until the crossover (provided your source is also digital).

The analogue solutions from Bryston and Rane would be more suited for use with Naim since they work purely in the analogue realm. There shouldn't be any problem with their usage since they would, in theory, operate on a signal already optimised for a Naim power amp by a Naim preamp, and perform better than the passive crossovers in whatever speakers regardless of the fact they use lower end PSUs inside the units.

However, looking into it in more detail, the Bryston solution would only be suitable for speakers that had perfectly aligned mechanical stop points for all drivers. And that is rarely the case. The PMC Twenty/5 ranges acheive this by using a sloped fascia which puts all drive units on the same approximate virtical plane. But for anything else, you need, not only high and low pass filters, but a delay control also and information from the manufacturer about what the delay for each drive unit should be. 

It's all a bit of a minefield and one I think Naim could simplify by rethinking SNAXO from scratch as such a flexible product coupled with Naim's prowess in PSU isolation. I'd love to see a speaker agnostic SNAXO (full width 3-way Supercapable) and SNAXO-lite (shoebox, 2-way, HiCapable). The market for such a product would not be huge (the response to this thread tells us that - it is mainly me talking to myself). But being fully user configurable would lower the service and R&D cost to Naim also since they would not need to configure units at the factory or work on speaker specific implementations.

Wishful thinking.

 

 

The new Bryston BAX-1 digital crossover can be used between Naim preamplifiers and Naim amplifiers but it is programmed at the factory for  specific Bryston active loudspeakers. However, in the UK Bryston is distributed by PMC so the BAX-1 could probably be programmed for PMC speakers if asked.

I didn't mention that model on purpose. I'd not bother with another A2D -> D2A conversion in the chain. 

I can't help feeling the analogue crossovers like the 10B-STD would be better. Of course I haven't compared. I've only ever heard Linn Aktiv and Naim SNAXO back in the 90s.

Innocent Bystander posted:

Gazza posted:

The reason I question, is it tried a Naim Core/Chord Qutest into the Nova, seemed to suck some of the dynamics from the music. Then moved on to Core/Chord Dave into Nova.....sounded pretty much like a Nova, but £8500 dearer. Put it down to the ADC then DAC of the Nova.......the Chord lives to fight another demo, maybe, some day.

Hmmm interesting. Have you tried Dave direct into your power amp?

Mentioned in another thread I will do at some point, but will now wait for the new streamers and then evaluate...one for another thread

Huge posted:

Not at all what I said - they just don't have a 'Magik' solution - it's constrained by physics exaktly like every other solution!

Linn claim to achieve zero phase distortion with the Exakt system. As opposed to analog crossovers and digital crossovers using IIR, according to Linn. If this is true, it's probably the best solution one can have.

I don't know if/how Linn achieve this in reality (if I would, I'd probably be working there), but in the digital domain you can do of course anything with a signal, you are not limited by math or physics. I can replace a mouse's squeak with a lion's roar if I like. Just perhaps not with digital filters which are limited in their features. Another challenge is probably to do this real time.

If you'd like to learn more about Exakt, perhaps google "Linn", "Exakt", "zero phase" and "Philbo".

jfritzen posted:

Huge posted:

Not at all what I said - they just don't have a 'Magik' solution - it's constrained by physics exaktly like every other solution!

. I can replace a mouse's squeak with a lion's roar if I like..

 

Novel idea - the acrptive XO Jazzing up Carnival of the Animals...

jfritzen posted:

Huge posted:

Not at all what I said - they just don't have a 'Magik' solution - it's constrained by physics exaktly like every other solution!

<snip>

I don't know if/how Linn achieve this in reality (if I would, I'd probably be working there), but in the digital domain you can do of course anything with a signal, you are not limited by math or physics. I can replace a mouse's squeak with a lion's roar if I like. Just perhaps not with digital filters which are limited in their features. Another challenge is probably to do this real time.

<snip>

Well digital crossovers, operate using mathematical principles and are limited by the laws of mathematics - so you can't get around those limitations.

Sound is a physical phenomenon and is limited by the laws of physics - so you can't get around those limitations either.

Finally real time crossovers are limited by the second law of thermodynamics, so cannot get around causality (but non-real time crossovers can bend the law to some degree).

So you add some latency, which does not hurt with unidirectional audio, in exchange for zero phase distortion through driver and crossover? If so this sounds like a good deal to me and I still maintain that Linn know their digital stuff.