Comparing amplifier THD+N in active speaker

[JEntwistle]

How do you compare the performance of an amplifier section in an active speaker to a separate amplifier + a passive speaker?

For example in the Genelec M040 review, we have these charts:

I think these include distortion from both the built-in amplifier and the rest of the speaker. How would you know whether performance is being held back by the amplifier?

i.e.: how would you be able to compare this to a state-of-the-art (in terms of SINAD) amplifier paired with a passive speaker? In the latter case, you can see the above charts for the passive speaker and you can see the measured SINAD of the amplifier.

And then the bigger question: is the amplifier in an active speaker usually a compromise compared to the best measured amplifier plus passive speakers?
Let's say for example (just making this up), three very highly regarded products here: A Genelec active speaker vs. a Benchmark amplifier + ELAC passive speaker. Would the Benchmark + ELAC always (usually) outperform the Genelec? Is the active speaker necessarily a compromise?

 

[amirm]

You would need DSP to match the DSP capabilities of active speakers that have them. You also need to bi-amp them so that the bass distortion doesn't bleed into tweeter.

But yes, in the absolute, there are amplifiers that are more powerful and less noisy than what is in an active speaker. All but a few show these limitations.

 

[JEntwistle]

"But yes, in the absolute, there are amplifiers that are more powerful and less noisy than what is in an active speaker. All but a few show these limitations."
This part is what I would have guessed. Is there a way to see this in your measurements, or is this something you just know from your experience?

But it sounds like you might be saying that even if a passive setup can match the DSP capabilities of an active speaker, you would need to double your amplification cost to manage the bass distortion bleed as well.

So does that mean that there is a definite trend toward the better active speakers with competent DSP ultimately becoming a better solution than passive speakers? So save your money on amps and invest that into active speakers?

And/or how does one choose which path to take?

 

[amirm]

"But yes, in the absolute, there are amplifiers that are more powerful and less noisy than what is in an active speaker. All but a few show these limitations."
This part is what I would have guessed. Is there a way to see this in your measurements, or is this something you just know from your experience?

Both although mostly the listening side of the reviews. It is kind of hard to give a solid answer as you may not want to listen as loud as I do. On measurements, as you posted, I do indicate if there is limiting. But that is not always the issue. Distortion often kicks in.

 

[amirm]

So does that mean that there is a definite trend toward the better active speakers with competent DSP ultimately becoming a better solution than passive speakers? So save your money on amps and invest that into active speakers?

For my needs, only the top of the line speakers from Genelec and Neumann have limits beyond what I need. For example, I would be happy with Genelec 8361A. With anything lower, I would use passive with very powerful amplification so it doesn't clip (and hence doesn't need bi-amping). Passive speaker needs to be very accurate to start and not need much EQ.

 

[amirm]

And/or how does one choose which path to take?

I think if you listen at mild to medium levels, you get better performance out of a DSP active speaker than a passive system you can put together in the same price range. That is, if you like their looks. There, you have many choices of active. If you go beyond this playback level, then it gets hard as there are a lot of choices in passive and comparatively few in active.

Note that i am excluding subwoofers in this equation. I think they are hard to integrate for most people. So all of my comments deal with speakers alone.

 

[JEntwistle]

Both although mostly the listening side of the reviews. It is kind of hard to give a solid answer as you may not want to listen as loud as I do. On measurements, as you posted, I do indicate if there is limiting. But that is not always the issue. Distortion often kicks in.

For my needs, only the top of the line speakers from Genelec and Neumann have limits beyond what I need. For example, I would be happy with Genelec 8361A. With anything lower, I would use passive with very powerful amplification so it doesn't clip (and hence doesn't need bi-amping). Passive speaker needs to be very accurate to start and not need much EQ.

I think if you listen at mild to medium levels, you get better performance out of a DSP active speaker than a passive system you can put together in the same price range. That is, if you like their looks. There, you have many choices of active. If you go beyond this playback level, then it gets hard as there are a lot of choices in passive and comparatively few in active.

Note that i am excluding subwoofers in this equation. I think they are hard to integrate for most people. So all of my comments deal with speakers alone.

Thanks. This is really helpful.

Have any of the active speaker manufacturers ever thought about partnering with a premier amplifier technology provider? Maybe on the business side it doesn't make sense; but could they license something like the the Hypex technology?

It does sound like if the cost of the amplification comes down enough, like it appears to be (e.g.: Fosi V3), active speakers may start taking over the market.

 

[Sokel]

Other than guessing have a look at Genelec 8020B chip-amp specs used in it's internal amp for example:

https://www.ti.com/lit/ds/symlink/lm1876.pdf?ts=1712744355761&ref_url=https%253A%252F%252Fwww.google.com%252F

 

[JEntwistle]

Other than guessing have a look at Genelec 8020B chip-amp specs used in it's internal amp for example:

https://www.ti.com/lit/ds/symlink/lm1876.pdf?ts=1712744355761&ref_url=https%253A%252F%252Fwww.google.com%252F

View attachment 365517

Oh, I did not even think about this. It would be a good start in comparing.

 

[tmtomh]

And we also know that most or all of the current Genelec Ones (8331 to 8361) use Infineon IRS amp modules.

I agree with @amirm ’s preference, except in my room with my listening-volume preferences and tolerance, I find that the 8351b will go as loud as - and louder than - I want without going into compression , clipping, or limiting.

 

[amirm]

Have any of the active speaker manufacturers ever thought about partnering with a premier amplifier technology provider? Maybe on the business side it doesn't make sense; but could they license something like the the Hypex technology?

JBL has. For the M2 speakers you can use their pro crown amps or mark levinson consumer amps. They don't support other brands though.

 

[highender]

Thanks. This is really helpful.

Have any of the active speaker manufacturers ever thought about partnering with a premier amplifier technology provider? Maybe on the business side it doesn't make sense; but could they license something like the the Hypex technology?

It does sound like if the cost of the amplification comes down enough, like it appears to be (e.g.: Fosi V3), active speakers may start taking over the market.

Sigberg Audio uses Hypex, and Sun Audio offers an external DSP amp with Hypex. I think there are more vendors out there but not sure.

 

[Waxx]

JBL has. For the M2 speakers you can use their pro crown amps or mark levinson consumer amps. They don't support other brands though.

But you can if you tune the dsp yourself. I did install some M2's with Hypex Ncore amps and a minidsp flex. But that means you need to measure the speaker in room yourself to design the crossover and other filters (DIRAC and so), which takes some skills that most consumers don't have. I once also did it with Lab Gruppen amps and their Lake dsp (that is build in). It's certainly possible, but JBL (and the mothercompany) won't support you if you don't use their homebrands (like Crown or Mark Levinson).

 

[JEntwistle]

So is it fair to summarize this as:

- There is no direct way to measure the performance of the amplifier inside an active speaker that would be comparable to amplifier reviews on ASR
- But we could infer the amplifier performance through published specs or by knowing which amplifier the speaker uses, if disclosed (this would still depend on a "competent" implementation of the amplifier design)

- The benefits of active speakers compared to passive speakers would be:
- Often built-in DSP
- Optimization of tweeter and woofer energy (so the bass distortion does not bleed into the tweeter)
- Potential cost savings from not having to buy a separate amplifier and DSP hardware and software

- Limitations of active speakers could be:
- Potentially not state-of-the art amplification
-- This could be for a number of reasons, including the design chosen, the competence of the implementation, or packaging concerns (e.g.: fitting the amplifier into the speaker housing; heat management)
- Lower amplifier power for the money (so you have limited SPL)
- Overall cost
- This is really just that there are limited options, and the best versions are expensive (for whatever reason)

It sounds like if one really wanted a passive speaker, they could replicate the DSP and other benefits of a well-designed active speaker. But this takes some know-how and effort and cost.

So with the advent of much less expensive and much smaller amplifier designs with great measured performance, is the future more active speaker options at lower price points? And would that be the best choice, more often than not?

 

[tmtomh]

Great comment, @JEntwistle ! I agree with everything you say.

I think it's important to try to be as objective and critical as possible and also to advocate for top-level engineering. And of course manufacturers work to minimize costs wherever they can.

But that said, I have a hard time believing that the top active speaker makers, whether targeting the pro or home markets - so for example Genelec, Neumann, Dutch and Dutch, KEF - are using amp modules that they believe create audible flaws in the sound that comes out of their active speaker systems. And since these makers tend to be quite measurement-oriented, their belief is likely backed by measurements.

In other words, most of us seem convinced by the available evidence that a properly implemented Hypex NCore is not going to sound audibly different than a Purifi, a Hypex Nilai, NCx and so on, even though the latter measure better than the original NCore.

Similarly, I think there is compelling - albeit circumstantial and by no means conclusive - evidence that the Infineon, Pascal, and other amp modules used in the top-tier active speaker systems produce audibly indistinguishable results compared to the Hypex and Purifi modules, except perhaps when it comes to noise levels and/or to distortion levels at certain frequencies when the amps are pushed to the higher levels of their output capabilities. I would guess that both of these characteristics are masked 99% of the time by (a) the self-noise of the active speaker system created by the internal gain-staging and lack of passive crossover filter network; and (b) built-in compression/limiting.

Of course it would be really interesting if @amirm had the opportunity to measure, say, a bespoke Genelec 8351 or 61 with Hypex or Purifi modules inside it instead of the Infineon IRS modules (with the rest of the amp and DSP tuned accordingly for the same target response of course). I suspect the measurements would be identical, although it would be really interesting to see if there were any measurable improvement in self-noise/hiss or in distortion at certain frequencies and SPL levels.

Finally, for the future, it seems that TI-chip based "chip amps" have upped their performance of late (and not just with reduced load dependance, which is irrelevant for actives), and so I would expect the cost-performance tradeoff in active speaker amps to essentially disappear in the future, as chip-amp modules will likely make that final leap to cross the transparency threshold at dirt-cheap prices, if they haven't already.

 

[Chr1]

I think that the same goes for class A/B amps in active speakers. I bought Neumann KH310s knowing the amps were class A/B but more importantly, that Neumann will have selected them for not only sound quality but also reliability. I decided that since the KH310s had been around for 20 years, the design and build was proven and therefore a safe bet, given the hefty (for me) financial outlay involved. Also, I don't need inbuilt DSP.

 

[JEntwistle]

But that said, I have a hard time believing that the top active speaker makers, whether targeting the pro or home markets - so for example Genelec, Neumann, Dutch and Dutch, KEF - are using amp modules that they believe create audible flaws in the sound that comes out of their active speaker systems. And since these makers tend to be quite measurement-oriented, their belief is likely backed by measurements.

In other words, most of us seem convinced by the available evidence that a properly implemented Hypex NCore is not going to sound audibly different than a Purifi, a Hypex Nilai, NCx and so on, even though the latter measure better than the original NCore.

Similarly, I think there is compelling - albeit circumstantial and by no means conclusive - evidence that the Infineon, Pascal, and other amp modules used in the top-tier active speaker systems produce audibly indistinguishable results compared to the Hypex and Purifi modules, except perhaps when it comes to noise levels and/or to distortion levels at certain frequencies when the amps are pushed to the higher levels of their output capabilities. I would guess that both of these characteristics are masked 99% of the time by (a) the self-noise of the active speaker system created by the internal gain-staging and lack of passive crossover filter network; and (b) built-in compression/limiting.

I think that the same goes for class A/B amps in active speakers. I bought Neumann KH310s knowing the amps were class A/B but more importantly, that Neumann will have selected them for not only sound quality but reliability. I decided that since the KH310s had been around for 20 years, the design and build was proven and therefore a safe bet, given the hefty (for me) financial outlay involved. Also, I don't need inbuilt DSP.

I would hope this is the case, and certainly, if you read Amir's Genelec 8361A review, you will see that he believes the end result is the best sound he has heard.

Part of what I'll say is admittedly the mindset that we (I) sometimes get here, which is chasing the measurements, even if we know the differences are not audible. We have seen some implementations of chip sets that are very competent and some that are very "meh". If we were looking to buy an amplifier most of us here would probably at least select from the top grouping of Amir's measurements. But what if an active speaker is using a second-tier design that would be holding back even better performance? We would really not know.

Of course it would be really interesting if @amirm had the opportunity to measure, say, a bespoke Genelec 8351 or 61 with Hypex or Purifi modules inside it instead of the Infineon IRS modules (with the rest of the amp and DSP tuned accordingly for the same target response of course). I suspect the measurements would be identical, although it would be really interesting to see if there were any measurable improvement in self-noise/hiss or in distortion at certain frequencies and SPL levels.

This is really what I would be curious to see, and what prompted my initial questions. And there may be some good reasons that the Hypex or Purifi modules can't work within the current Genelec design, anyway.

Without knowing where the cost is and what the pricing strategies are, with the products like the recently reviewed Fosi V3 Mono, the cost and size of top-tier amplification is becoming orders of magnitude lower. You would hope that would unleash a new lower-priced version of something like the Genelec.

But ultimately, I think there is a big transition looming between two verticals: continue to "optimize" a passive speaker setup, where you upgrade the amplifiers, DSP, and room EQ, or jump over to the active speaker side. It's two different worlds, and we can't even directly compare them. But I think I am beginning to understand that active will likely be the way to go.

 

[Audio Science Pal]

But what if an active speaker is using a second-tier design that would be holding back even better performance?

I bought Neumann KH310s knowing the amps were class A/B but more importantly, that Neumann will have selected them for not only sound quality but reliability.

Reliability is a very desirable feature.