Does DSP belong in State of the Art Systems?

[the sound of Tao]

Maybe the pudding is like a trifle ...with a lot of stale leftovers sweetened with custard
I think it's just a leviathan that will take some turning .. but there is no doubting the sq that can be achieved

I just guess in purely cost no object systems active can establish itself more as an option… anecdotally that seems to be in some fairly specific cases to be a possibility but multiples of mundane amps in a high resolution system likely just become even more exposed for their constraints… and then at what other cost?

Distortion may be lowered as a potential but does it create a greater challenge in terms of bringing things together coherently at a summation level. Not saying that it can’t be achieved in extraordinary systems with phenomenal expenditure but in less optimal implementations it may prove to be a constraint of fully active as an approach.

 

[Young Skywalker]

So there are (possibly literally) a handful of fully actives (often with passive as default or optional) among all the crazy uber expensive speakers out there and that now means passive crossovers are all dead… sounds just like how digital was released and vinyl is now dead.

You have misinterpreted my argument. I did not say that passive crossovers were dead. I said that the path to maximum fidelity is via optimised active crossovers as opposed to optimised passive crossovers. That does not mean that passive crossovers have no place in the market. There are many well engineered products at the entry and mid-levels (by high end standards). There are also many excellent loudspeakers that don't try to be all things to all people which minimise the negative aspects of lossy passive crossovers, e.g., using shallower filter slopes where appropriate (some horn designs) or no filter at all in the case of wideband drivers.

I do believe that full range, so-called state of the art products costing as much as a home should be implementing active crossovers, be they purely analog or using DSP for systems with exclusively digital sources.

 

[Young Skywalker]

I just guess in purely cost no object systems active can establish itself more as an option… anecdotally that seems to be in some fairly specific cases to be a possibility but multiples of mundane amps in a high resolution system likely just become even more exposed for their constraints… and then at what other cost?

Distortion may be lowered as a potential but does it create a greater challenge in terms of bringing things together coherently at a summation level. Not saying that it can’t be achieved in extraordinary systems with phenomenal expenditure but in less optimal implementations it may prove to be a constraint of fully active as an approach.

The amplifiers that come after an active crossover have a very easy job, being bandwidth limited and seeing just the corresponding driver's voice coil, compared to an amplifier asked to navigate the impedance curve of a reactive passive crossover network limited by the least sensitive driver.

This argument that an active system mandates 3 times the outlay (for a three-way speaker) on very expensive stereo or mono block amplifiers is simply not true. I do hope that Roy Gregory makes good on his follow-up Clarisys review by trying some "affordable" Simaudio amplification on active Auditoriums (or one of the less expensive models).

 

[the sound of Tao]

You have misinterpreted my argument. I did not say that passive crossovers were dead. I said that the path to maximum fidelity is via optimised active crossovers as opposed to optimised passive crossovers. That does not mean that passive crossovers have no place in the market. There are many well engineered products at the entry and mid-levels (by high end standards). There are also many excellent loudspeakers that don't try to be all things to all people which minimise the negative aspects of lossy passive crossovers, e.g., using shallower filter slopes where appropriate (some horn designs) or no filter at all in the case of wideband drivers.

I do believe that full range, so-called state of the art products costing as much as a home should be implementing active crossovers, be they purely analog or using DSP for systems with exclusively digital sources.

Like most things it just comes back to context and implementation… I’d still suggest that because some crazy expensive setups manage to make active work for some (and for the sonic parameters that they are specifically chasing) just doesn’t then establish it as best option for speaker design at all.

For what may be a majority of speaker designers and listeners active may not be the best choice and that characteristic sound may not be an ideal outcome in terms of experience in music engagement or in sounding the least artificial.

 

[DasguteOhr]

This is the full Monty .. they now have a fantastic server so I am not sure if they will move to convolution in the pc for xovers etc. View attachment 131960

A nice curved plexiglass baffle would not have deteriorated the design. Then there would be no need for a subwoofer anymore, there would be enough bass.

 

[Young Skywalker]

Like most things it just comes back to context and implementation… I’d still suggest that because some crazy expensive setups manage to make active work for some (and for the sonic parameters that they are specifically chasing) just doesn’t then establish it as best option for speaker design at all.

For what may be a majority of speaker designers and listeners active may not be the best choice and that characteristic sound may not be an ideal outcome in terms of experience in music engagement or in sounding the least artificial.

All good. In my experience, most properly engineered active solutions are higher in fidelity compared to most passive solutions, regardless of price. Furthermore, this performance advantage is more replicable in a wider range of environments. I would never go back to passive, even if I came into millions of dollars and could buy whatever I wanted without financial constraint.

My remaining interest in passive loudspeakers is now purely about the gear nerd side of the hobby, e.g., I enjoy some of Kedar's videos with high sensitivity loudspeakers with minimalist passive crossovers driven by purist amplification, particularly when my preferred diet of classical music is playing (Shostakovich, Mahler, Tchaikovsky, Bruckner, Beethoven, etc.). After all, I started my serious journey into high end audio with an SET amp using 45s.

 

[the sound of Tao]

All good. In my experience, most properly engineered active solutions are higher in fidelity compared to most passive solutions, regardless of price. Furthermore, this performance advantage is more replicable in a wider range of environments. I would never go back to passive, even if I came into millions of dollars and could buy whatever I wanted without financial constraint.

My remaining interest in passive loudspeakers is now purely about the gear nerd side of the hobby, e.g., I enjoy some of Kedar's videos with high sensitivity loudspeakers with minimalist passive crossovers driven by purist amplification, particularly when my preferred diet of classical music is playing (Shostakovich, Mahler, Tchaikovsky, Bruckner, Beethoven, etc.). After all, I started my serious journey into high end audio with an SET amp using 45s.

I’m with you on that and find so many of Ked’s videos to showcase approaches to gear to be really aspirational to me and align very much to it and with music I very much enjoy along with it.

I’m not as purist with my choices in gear because unfortunately I can’t afford that at this stage of my life and run with a more modest setup, but the music I can play with it and how that sounds makes me completely happy. I suppose we get to a point where we struggle to justify having it all… even if we have given that mission plenty of ourselves along the way. More power to those that can, I think that’s awesome.

 

[pjwd]

A nice curved plexiglass baffle would not have deteriorated the design. Then there would be no need for a subwoofer anymore, there would be enough bass.

That's what I think as well .. in fact I believe it's actually a quadrapole.. or something like that
There must be a clear design intent for them to have done that but it's beyond my understanding .. perhaps imaging

I can say that it sounds stunning.. there is no sense of speakers in the room...the music just floats

 

[hopkins]

Passive has nothing to do with high fidelity.

What is your definition of high fidelity?

 

[DasguteOhr]

That's what I think as well .. in fact I believe it's actually a quadrapole.. or something like that
There must be a clear design intent for them to have done that but it's beyond my understanding .. perhaps imaging

I can say that it sounds stunning.. there is no sense of speakers in the room...the music just floats

I don't doubt it, it will definitely sound good. But 6×12"AE td woofers are really expensive ~4k€. With a curved baffle 60cm wide you would have achieved a much lower cutoff frequency with 3×12"(-3db 30hz)97db /1watt. No baffle speaker looks cooler

 

[pjwd]

I don't doubt it, it will definitely sound good. But 6×12"AE td woofers are really expensive ~4k€. With a curved baffle 60cm wide you would have achieved a much lower cutoff frequency with 3×12"(-3db 30hz)97db /1watt. No baffle speaker looks cooler

I know they are in the same phase so perhaps the idea is to cancel the output through the basket of both drivers hence its resulting difraction... that would then make a virtual circular baffle from the distance between the drivers.. just a wild stab

Also the AE dipole drivers are high Q and have a built in shelving type boost ... the 15" fr fall apart above about 250 hz but rise at nearly 6db slope to 30hz or so

 

[Holmz]

I just guess in purely cost no object systems active can establish itself more as an option… anecdotally that seems to be in some fairly specific cases to be a possibility but multiples of mundane amps in a high resolution system likely just become even more exposed for their constraints… and then at what other cost?

It might be easier to make the argument for systems that are less costly?
One can certainly use feedforward methods to lower distortion, and also get better transient response with an active system than with a passive XO.
And the damping factor of an amp is not shielded behind a passive XO.
And it is easier to replicate that multiple times when it is s/w based.

Distortion may be lowered as a potential but does it create a greater challenge in terms of bringing things together coherently at a summation level. Not saying that it can’t be achieved in extraordinary systems with phenomenal expenditure but in less optimal implementations it may prove to be a constraint of fully active as an approach.

There are no active systems that I can recall that do not allow for the time domain to be fixed.
Many system do not allow for phase rotation, nor for group delay, but some do… and they all do the time domain summation.

 

[Young Skywalker]

What is your definition of high fidelity?

I knew it wouldn’t be long before someone rightly requested that the term “high fidelity” be defined. After all, we don’t really know what the master is supposed to sound like.

For me, high fidelity audio playback manifests as the potential for greater discrimination between any two recordings (micro and macro dynamics, sense of original recorded acoustic space, tonality of voice or acoustic instruments, etc.). For that to happen, the system must feature low levels of distortion/colouration at each stage of the playback chain. If one changes a variable in the playback system, the question should not be “can I hear a difference?” But rather, “can I hear deeper into each recording such that the uniqueness of each is enhanced?”

 

[hopkins]

I knew it wouldn’t be long before someone rightly requested that the term “high fidelity” be defined. After all, we don’t really know what the master is supposed to sound like.

For me, high fidelity audio playback manifests as the potential for greater discrimination between any two recordings (micro and macro dynamics, sense of original recorded acoustic space, tonality of voice or acoustic instruments, etc.). For that to happen, the system must feature low levels of distortion/colouration at each stage of the playback chain. If one changes a variable in the playback system, the question should not be “can I hear a difference?” But rather, “can I hear deeper into each recording such that the uniqueness of each is enhanced?”

Fair enough. I guess everyone's definition may be somewhat guided by their experience with a variety of systems. I don't know if their is a unique "theoretical" definition. Let's see Bjorn's point of view.

 

[schlager]

High fidelity = true to the source (music), it's really quite simple. So if you are listening to an uneven frequency response, bad time domain, lot's of distortion from front end gear (tube amps, LP, etc.), you may like the sound from a subjective point of view, but you are technically, not listening to high fidelity sound.

 

[Hear Here]

High fidelity = true to the source (music), it's really quite simple. So if you are listening to an uneven frequency response, bad time domain, lot's of distortion from front end gear (tube amps, LP, etc.), you may like the sound from a subjective point of view, but you are technically, not listening to high fidelity sound.

That sounds like a comment from an ASR forum member! Measurements are far less important than they often claim. How many of an orchestra's instruments are measured using a microphone and computer and tuned according to measurements?

It's what the music sounds like that really matters and a little distortion (the aurally acceptable / appealing type) often does no harm at all. After all, the measuring kit doesn't have ears or a brain.

 

[schlager]

Hear here, that you really don't seem to distinguish between real instruments and a reproduction/playback of a recording of real instruments just speaks volumes of your general audio knowledge or lack of. No wonder you can't make DSP/EQ work

 

[hopkins]

High fidelity = true to the source (music), it's really quite simple. So if you are listening to an uneven frequency response, bad time domain, lot's of distortion from front end gear (tube amps, LP, etc.), you may like the sound from a subjective point of view, but you are technically, not listening to high fidelity sound.

I expected this answer from you I respect your approach, knowledge, and "sharing" attitude!

It is indeed easy and convenient to define high fidelity by what it is not, based on criteria that you can measure.

I think you will agree that you never get a perfect score on all criteria?

Moreover, some criteria may be more difficult to evaluate than others (in the sense that the measurement may only approximate the behavior of the system in the context of a real performance, not just test tones).

It is difficult to negate listeners' preferences for certain types of sonic attributes just by saying that they "prefer distortion" just because the system they enjoy may in fact perform poorly on one criteria, without trying to understand what it is that they enjoy and which measurements can in fact explain that, if at all possible.

Why do some people love the sound of a 1920s Western Electric horn speaker? Simply due to distortion?

Anyway, I don't want to debate this ad nauseum. I just wanted to point out that the decision between active/passive or even "crossover-less" may not be that simple.

 

[sigbergaudio]

It feels like we're discussing various aspects of / applications of DSP at the same time, which probably confuses the matter quite a bit.


Digital crossovers vs analog/passive crossovers = one discussion

Can applying DSP to "correct for the room" be beneficial, and/or will it detoriate the sound = a different discussion

Will DSP (digital signal processing), independent of use case / application, inherently reduce sound quality = possibly a third discussion.

Which one are we having / should we split it?

 

[sigbergaudio]

i'm open to hearing a fully dsp active system with serious sources and a serious room that sounds better than what i hear in my passive all analog acoustically designed and tweaked room. am i skeptical? hell double yeah! but i am open to it.

Having an active system with digital crossovers doesn't mean you should not also have an acoustically treated and tweaked room.

The test should surely be a passive system vs an active system in the same room?