Review and Measurements of Denon AVR-4306

[maty]

Class D amps have a problem with great phase shift > 1 kHz.

Almost nobody comments that weakness but there it is. A recording with acoustic instruments can be problematic. Modern music, with electronics, low DR and... no problema.

That is why I gave up months ago buying an amplifier with the Icepower 1200AS2 module.

I have a thread:

Phase shift in class D amplifiers. How it affects the sound?

https://www.diyaudio.com/forums/class-d/321578-phase-shift-class-amplifiers-affects-sound.html

 

[graz_lag]

Given that these have built in amplifiers and a DAC ...

To me that's the way to go, definitively for a secondary system. (I say secondary considering the speaker cabinet dimensions and typical listening distance.)
Better integration and tuning of the critical components, as a whole engineered and integrated electronic.
Less cables that mess around.
Strangely, Dynaudio of Denmark has discontinued what was probably their best collection of active speakers with integrated D/A conversion: the AIR.
The specs as well as the listening experience was impressive with those AIR's.
The AIR's where designed for the studios, so they did not get intercepted by the radars among the community that proudly call itself "audiophile".
I was hoping other manufacturers would have followed the path that Dynaudio opened with the AIR.
But as a matter of fact, no.
As far as the tower cabinets : I was recently impressed by the Triton series from GoldenEar, which features built-in powered subwoofers.
For these, the value for money is terrific, you get a lot of high-quality loudspeakers for your money.

 

[JJB70]

The idea of active digital speakers seems to be viewed with some suspicion in the hi-fi segment, it's the polar opposite of the more general sound segment where active digital speakers now dominate to the point that many non-specialist hi-fi shops don't sell anything else than Bluetooth speakers. I do think we will see a steady increase in the adoption of active digital speakers but I also don't see passive speakers and the associated amplifiers etc going away, or not for a long time at least.

 

[Maki]

Thanks a lot for doing these speaker amplifier measurements. It's unfortunate most of these products don't measure up to even budget headphone gear, but I suppose that just comes with the much higher power requirements of speakers. If it weren't my only power amplifier, I'd certainly send my Emotiva BasX A-100 in for measurement.

 

[andreasmaaan]

Class D amps have a problem with great phase shift > 1 kHz.

Almost nobody comments that weakness but there it is. A recording with acoustic instruments can be problematic. Modern music, with electronics, low DR and... no problema.

That is why I gave up months ago buying an amplifier with the Icepower 1200AS2 module.

I have a thread:

Phase shift in class D amplifiers. How it affects the sound?

https://www.diyaudio.com/forums/class-d/321578-phase-shift-class-amplifiers-affects-sound.html

There’s no doubt such phase shift occurs, but there’s also no reason to believe it should be audible.

Research has never established that phase shifts of anywhere near that low a magnitude are audible in that frequency range, to my knowledge.

Moreover, the type of phase shift is minimum phase and is exactly what occurs in any tweeter as a result of its natural supersonic roll-off, not to mention most DA converters owing to their anti-imaging filters, and indeed all recordings owing to the high-frequency roll-off of recording microphones, analogue tape, and/or anti-aliasing digital filters in the ADC.

Compared to these phenomena, the extent of phase shift produced by a relatively gentle low pass filter in a class D amp is just not significant.

However, I didn’t get through your whole diy audio thread - perhaps I missed some evidence there regarding audibility?

 

[gorg]

A link to archive of tests done by stereo.de crew (pdf files): https://www.stereo.de/verweise/messtechnik/

 

[maty]

The question is:

Why class A/AB amplifiers try to limit the phase shift to +/- 6.5º?

Why not the class D amps?

The speakers limit the phase shift to +/- 45º. If the transition is smooth then the ear will hardly appreciate it.

IcePower 1200AS2 has a phase shift of 50º at 1 kHz, you know, the human voice. And 70º at 10 kHz.

I can not insert the images.

http://maty.galeon.com/WP-imagenes/hard/IcePower-1200AS2-output-impedance-phase.png

My KEF Q100 speakers.

Shift phase and Impedance graphs:

http://www.salonav.com/arch/2012/09/sp_graf/kef_imp.jpg

With this module:

1 kHz -> +40º + 50º = +90º >> +45º

10 kHz -> -32º + 70º = +38º

Will the voices sound the same?

In my main system (kidnapped by TV and family), the music sounds much better with Direct than with room equalization (Yamaha). With very good records (maybe 80% of all my music) the difference is overwhelming.

 

[amirm]

Will the voices sound the same?

Most likely so.

The enemy of phase detection by the ear/brain is the room. Sound reflects from all the surfaces and arrives with different time delays at the ear. In other words, in any room you have massive bombardment of phase distortion. As I explain in my article on perceptual effects of room reflections, a simple frequency sweep that looks like this:

Becomes like this due to all of those reflections adding up:

Same thing happens in our everyday life with other voices interacting heavily with the environment. If we could hear it, then people's voices would constantly change if they just moved an inch. But that is not what we hear. Constant distortions like this are perceptually ignored.

Studies show that with headphones where room reflections are not an issue, we could hear phase delays with special signals. But with music, this again becomes impossibly hard. See this post: https://www.audiosciencereview.com/...surements-of-jds-labs-el-dac.4850/post-108285

 

[andreasmaaan]

That degree of phase shift (<360 deg @ 1KHz) has never been shown to be audible under controlled experimental conditions. The experimentally established audibility thresholds, even if slightly wrong, are just nowhere near the 40-50 deg we’re talking about here...

Also, those graphs you posted are for electrical impedance, which is a completely separate question.

The KEF Q100 uses an LR4 xover IIRC, so the phase shift at 1KHz is likely to be about 360 deg from the crossover alone...

 

[RayDunzl]

Probably why I like my electrostats (blue) more than cones and domes (red).

Reflections (relatively untreated but lived-in room)

It probably contributes to the way the stats throw a measurably flat phase to the listener at the sweet spot. Woofer/panel crossover is 180Hz.

DRC in use on both, though the panels throw a phase-flat measure without DRC. Blue includes the Cheezewoofers, Red does not.

I find they sound "cleaner", though that may be due to less harmonic distortion. I definitely prefer the imaging, much more focused. I suppose deadening the walls/ceiling would help, but I won't.

 

[dwkdnvr]

Also, those graphs you posted are for electrical impedance, which is a completely separate question.

IMHO this is the key point to be made - the phase of the amp output impedance is a pretty irrelevant measurement. Assuming we're not talking about a tube amp with a very high output impedance, the amplifier output impedance will be swamped by the speaker impedance, and since they're in series the phase of the amp impedance becomes negligible.

If you're concerned about 'phase shift', it has to be the acoustic phase shift of the amplifier+speaker combination. I strongly suspect that the amp would have to be rather defective to greatly influence the aggregate result.

The KEF Q100 uses an LR4 xover IIRC, so the phase shift at 1KHz is likely to be about 360 deg from the crossover alone...

Actually, I think the Q100 uses a simple cap on the tweeter and coil on the woofer. 1st order electrical, although I'm not sure what the resultant acoustic behavior is.

 

[maty]

The speaker response is influenced by the amplifier. In class A/AB amplifiers the output impedance is resistive -> the speaker response does not vary. Only change the magnitude.

But in class D the output impedance is not only resistive -> the system response is modified by the amp. How much? Human ear can detect it? I do not know.

 

[maty]

By the way, KEF Q100 filters are only first order!

http://maty.galeon.com/WP-imagenes/kef-q100/KEF-Q100-crossover.jpg

My first Q100 cheap tweak.

http://maty.galeon.com/WP-imagenes/kef-q100/KEF-Q100-crossover-styroflex-bypass.jpg

 

[RayDunzl]

Constant distortions like this are perceptually ignored.

Not to me, when critically listening, imaging is affected, using these speakers, my room, YMMV and all that.

When not sitting in the sweet spot, and just going about my business (like right now) no problem.

I think I hear fake width and indistinct location vs an initially smaller but ultimately preferable presentation..

 

[andreasmaaan]

By the way, KEF Q100 filters are only first order!

http://maty.galeon.com/WP-imagenes/kef-q100/KEF-Q100-crossover.jpg

My first Q100 cheap tweak.

http://maty.galeon.com/WP-imagenes/kef-q100/KEF-Q100-crossover-styroflex-bypass.jpg

Fair enough, I was mistaken about the filter order. However, I suspect the q100 are electrical first order, acoustical second order. See the step response for the q900. That would imply 180 deg phase shift at 1KHz as a result of the crossover (and inverted polarity of the midwoofer).

And I agree that class D amps will generally interact more with the load impedance of the speaker in terms of frequency response. I doubt this will be of much significance in most cases, but yes it means the frequency response is more dependent on the impedance of the speaker being driven.

 

[FrantzM]

Not much Love for Class D in this forum ...

 

[bravomail]

what about class T? Does it have Phase shift?

Not much Love for Class D in this forum ...

 

[amirm]

Not much Love for Class D in this forum ...

There is from me....

 

[Blumlein 88]

Not much Love for Class D in this forum ...

I have some. I use Wyred4Sound amps. I've had some large expensive high performance amps before, and these match up or exceed all them. The only one that maybe sounded better is a Spectral. And I never had those side by side to see.

 

[March Audio]

There is from me....

Me too