Before you dismiss Meridian, take a listen to the latest products with all-new electronics, drivers and operating system: DSP8000XE and DSP9. Technology marches on…
Does DSP belong in State of the Art Systems?
- Thread starter musicfirst1
- Start date
DSP for room correction is difficult to get right, and most automatic systems fail to do so well. The best approach is to apply it sparingly and mostly to the bass.
While I have exclusively active, digital crossover speakers, none of the systems have active room correction. Any applied DSP is manually applied below 100-150hz only. Start with good speakers and an acoustically decent room, and that is all you need.
I am going to try dsp amps on my bass towers at some point, mainly to free up more power to the main speakers from my good amps, i doubt i will be satisfied, but will try it. Mainly because it would be a cost efficient solution if it sounds good, buying a second set of MBL 9011 amps are $100K these days.Okay. Many interesting designs yet to try then!
You have not addressed the need for an AD converter for analog sources. I also question the fidelity of multiple low cost DACs in a digital active speaker compared to my MSB Reference DAC with Reference Digital Director.1. You will lose one of the main advantages of a digital crossover as a developer, the ability to quickly iterate and experiment. The development time and cost will skyrocket compared to a digital crossover. As a natural outcome of this, the end result will not be as good.
2. A competent digital crossover will be at least as transparent and more efficient than an analog crossover. It also have capabilities to do more and different things. As an example we use 1.order, phase linear crossover filters, but they are asymmetric. So further down, the slopes get steeper. Which means you get the benefit of the phase linearity and natural sound of a traditional 1.order filter, combined with the driver protection of a steeper filter.
3. Sound better than what? The SGR Audio speakers? It is not my place as a manufacturer to comment directly on competing products, and I am not familiar with their products. That being said, truly full range active speaker systems is actually our thing.
An extremely iterative approach to designing and building an active line level crossover is not necessary when you have a deep understanding of driver design, driver measurements (both free air and in cabinet) and electronics design (particularly ground plane optimisation).
Yes, designing a quiet sealed enclosure is expensive, as is designing and building your own drivers.
I am not suggesting that DSP and digital active crossovers have no place in high ene audio. Quite the contrary actually. But implementing these things at a level beyond DIY is neither inexpensive or within the technical abilities of just anyone.
1) Are there some links to show performance?
2) We/you have not even yet touched upon being able to correct motor non-linearity that leads to many distortions.
(This is independent of whether people like distortions or not… but being able to not have them, seems like an argue of higher fidelity.
The AD/DA in a competent digital speaker will be just as transparent as an analog crossover, and the sound signature of anything earlier in the chain will not be lost.
You only need one ADC/DAC per speaker, not multiple, and they don't necessarily have to be cheap.
The amps in our speakers use a high-end DAC from AKM (AKM AK4454 Velvet sound, some information here: https://audioxpress.com/news/Next-G...hannel-DAC-Series-from-AKM-Reaches-the-Market).Agreed, and I have never stated or suggested that just anyone should start to design and sell active speakers. I assume we agree that not just anyone can sit down and design and build high end passive speakers either.
But I agree that one of the reasons we don't see more high-end active speakers is because the traditional speaker companies currently building passive speakers may not have the competence needed to do so. They could of course aquire that knowledge if they wanted to, but the incentives are probably low as long as most of the current customer base are happy with passives.
Last edited:
To your first question:
I am not sure exactly what you are asking for, and I am also not completely sure what the forum rules are about sharing this about our own products, so please let me know if I am overstepping somehow.
With regards to subjective information, there is a review of our Manta speaker system in Norwegian Stereopluss available here. The review also discuss the fact that consumers may not be ready to embrace active systems (and why that might be a lost opportunity):
Note that the language may be a bit weird here and there as it is translated from Norwegian. For some context, this is a traditional Norwegian stereo magazine (so reviewing mostly analog gear) that has been around for a decade, and everyone working there has been connected to other publications a couple of decades beyond that.
This speaker system got a product of the year award, and was rated 10/10 for sound quality vs price for a system that cost ~25,000USD in Norway.
If you are looking for measurements, that is also available (shared below). Note that these show measurements of the main speakers only. This is a 4-way speaker system consisting of subwoofers, so the true response is 25-20,000hz +/-2.5dB anechoically, and extends below 20hz in rooom.
To your second question:
Are you thinking of active servo systems, or are you thinking of adjusting amplitude/frequency response irregularities more precisely than you can with a passive crossover?
Last edited:
I do not have “my own products:, just a background in science and some S/W and DSP skillz.
Do you have step function and impulse response?
That is an area where a digital XO can really do well, as you mentioned with… ‘steep filters, but not <jacking> up the phase and timing.’
Maybe more like feed forward.
If we know that the BL curve is not linear, then at some arbitrary position Z, we know in order to get the acceleration needed at that point, to be representative of what the signal requires, then we need to command the motor to move a bit more or less than it would be commanded to move without a BL adjustment. (Or any other motor non-linear behavior like suspension stiffness.)
Since one would need to do this on a motor by motor basis, it sort of makes sense to do it as part of a digital XO.
I actually got an Octo DAC-8 to do this, but have not even unwrapped it, and just picked up another passive-XO based speaker.
We don't take the "extreme" approach to digital crossovers. There are no FIR filters, and no high order slopes either. We use phase linear asymmetric 1.order filters (becomes steeper further down/up in frequency).
I would rather start with good, over engineered drivers than try to correct them like that. Keep it simple. Then we can stay well within the linear excursion of the drivers. The Manta system mentioned above is <0.5%THD at 96dB@1m.
Drivers has become better over the years as well, so something like subwoofers with active servo to correct for non-linear drivers also have less merit now (assuming good drivers are used of course).
It's more likely you don't like the active speakers you have heard.You keep on talking about active versus passive. I don't like the sound of active digital speakers, active speakers with analog crossovers are definitely contenders for great sound, the digital versions not so much !
As mentioned before, a correct comparison would be between the same speakers or where you start with the similar speaker design.
Not sure what you refer to by "digital". When using a DAC, a digital processing happens and is the same. Is using a DAC something detrimental to sound quality as well to you? I do assume you understand all music undergoes AD and DA conversion.
I am talking about the difference between active analog and active digital speakers, i have heard excellent analog active crossovers. Not all music undergoes AD and DA conversion. I try not to use a DAC if i can avoid it. When i play AAA analog recordings on my TT they have never been touched by the mark of Cain.
I totally respect if you cannot hear the difference, in my system i can. … When i play AAA analog recordings on my TT they have never been touched by the mark of Cain.
Should these should labeled as “Able recordings”?
Never to be played in the AM… only as the Eve approaches.
Hi, thank you!
They are anechoic.They are taken with the Klippel software in the anechoic chamber at Seas (I live around 20 minutes away from the Seas factory here in Norway).
Yes. When there's no crossover in the speaker. Like a large electrostatic speaker.
But it would still be greatly improved with a DSP with a more linear response and improved lows.
And biggest improvement would be to cross it actively to a separate bass/subwoofer solution with perfect time alignment, and avoiding the weakness of an electrostatic speaker - which is obviously the lowest frequencies.
Another poor strawman argument with no relevation to facts and to audible improvement.
Just showing the need to defend own choices, and that type of attitude it partially why high end audio has been declining for years. People defending crap and things they have no understanding of.
Thats fantasticHi, thank you!
They are taken with the Klippel software in the anechoic chamber at Seas (I live around 20 minutes away from the Seas factory here in Norway).
View attachment 132525
Pretty jealous
Graphs and measurements are certainly useful for "engineers", but as a consumer, I judge products based on listening.
Funny that you state that "high end audio has been declining for years", when in fact you could argue that it is precisely the push towards increasingly better measurements that have caused this decline.
No one in their right mind would argue that "all other things being equal", a flat and extended frequency response, is not preferable. The problem is, it always comes at a cost. If you are not "aware" of this, maybe it is because your experience of listening is limited, or that your obsession with measureable performance creates a bias.
Theoretical discussions are one thing. Listening is another. But since your products are "top secret", listening won't happen, and all we are left with is listening to your theories, which we have heard many times before.
Similar threads
