Review and Measurements of Yamaha RX-A1080 AVR

[RichB]

Of course, but a) the harmonic distortions spectrum of class AB amps, are not drastically different, and b) we are talking about <0.05 % THD+N 20-20,000 Hz, so no, I wouldn't worry about even, odd, low, high order.

I worry about distortion of any order. IMO, there is no good distortion there is only worse distortion.

- Rich

 

[GrimSurfer]

You are not an AVR user as far as I know. If you are, you would find that most Onkyo, D&M and Yamaha's are within the ball park to produce 85 dB average at volume 0 if you sit 10-13 ft from the speakers, without attenuation (may be a few dB at the most).

You don't have to be an AVR user to know how they (or sound) work. Indeed, there are many, many owners who don't have a clue how their AVRs work because they haven't bothered to read their owner's manuals.

The issue in my mind isn't whether any manufacturer's device can operate in the "ball park". The issue is whether an audio device can live up to its claims when operated within its stated and implied parameters. We can discuss what that might mean in the HT context... but I warn you that manufacturers themselves are long on claims and short on specs.

I think, however, that using terms like "ball park" or saying things like "without attenuation (may be a few dB at the most)" betrays a lack of confidence that is informed by the underlying knowledge that AVRs aren't the high fidelity products they're cracked up to be. @restorer-john may have more to offer from a design perspective...

 

[peng]

I worry about distortion of any order. IMO, there is no good distortion there is only worse distortion.

- Rich

Fully agreed, apparently studies shows not everyone prefer even (e.g. 2nd) harmonics to odd (iirc, even Nelson Pass has said something like that but..), some do prefer 3rd to 2nd, again just examples. I am just saying people can't tell the difference between 0.05% and 0.03%, if measured properly, full bandwidth and are "total"+N, or not. You may be an exception, I have the impression from your past posts that you are more sensitive to minute differences than a lot of ppl.

I bought my Bryston and Halo amp because of their low distortion specs, same for my preamp, external DACs. But can I hear difference? I don't know, but I doubt I can past any blind test, even single blind..

 

[scooter]

Bi-amping in AVRs may not provide any additional headroom. Since frequency bands are not split within the AVR all amplifiers still see the full signal voltage. There is essentially no increase in voltage headroom, and current headroom depends upon the load.

Frequency bands are not split but now 4 "amplifiers" are used instead of 2, and each of them drives lesser number of drivers, more impedance and less stress on each amplifier.
I know that total power comes from a single transformer, but now it is more evenly shared across 4 amps (one cannot overload and distort as quickly as before, when it was driving woofers and tweeters together). So why there is not any benefit if it has individual amplifiers?

 

[peng]

You don't have to be an AVR user to know how they (or sound) work. Indeed, there are many, many owners who don't have a clue how their AVRs work because they haven't bothered to read their owner's manuals.

The issue in my mind isn't whether any manufacturer's device can operate in the "ball park". The issue is whether an audio device can live up to its claims when operated within its stated and implied parameters. We can discuss what that might mean in the HT context... but I warn you that manufacturers themselves are long on claims and short on specs.

I think, however, that using terms like "ball park" or saying things like "without attenuation (may be a few dB at the most)" betrays a certain line of thinking. That thinking is informed by the underlying knowledge that AVRs have become, as @restorer-john has said, a bit of a joke whether you wish to admit this or not.

He's talking about "most" room, and requires over 20 dB of attenuation, that's why I said what I said and in response to his post only, so please do not take things out of context. Quote: "0 as "reference" in my room (and I expect most rooms) this requires over 20 dB of attenuation."

May be I misunderstood his post, or I took him out of context myself, but no offense, I would rather hear from him directly.

 

[Blumlein 88]

Here is what THX says about reference level:

https://www.thx.com/blog/questions/what-is-the-reference-level/



I've always found this dubious.
First, 85 dB at home is VERY loud and I suspect uncommon for folks using AVRs.
Second, 0 as "reference" in my room (and I expect most rooms) this requires over 20 dB of attenuation.
What for, to get a particular number. I am not a numberist.

There is another method, set the volume to your preference

- Rich

This explains it better than THX did.
http://www.acousticfrontiers.com/2013314thx-reference-level/

I think you are somewhat confused on the 0 db reference level. It doesn't mean you'll automatically use maximum or 0 db on your volume control of your gear.

What you do is play a pink noise signal which is -20 db from maximum digital signal level. While playing this you measure SPL at the listening position and adjust your volume so you get 85 db SPL. Your volume control may read -26 db or any other number. If it were -26 db for your gear, this is your nominal THX 0 db reference level. It allows for 20 db of headroom over normal average levels (or if your reference level were -26 db you'll have 26 db headroom and 6 of it will never be needed). Your speakers would then play max level in room at 105 db SPL (assuming they are up to it). If you listen at this level you are hearing exactly the sound levels the people who made a movie sound track intended for you to hear. Just keep your volume control set to - 26 db if that is what gave you reference on the pink noise signal.

Now if you found your volume was at some point higher than -20 db to meet reference SPL in your room, your gear might not be capable of full sound level to meet THX spec. If things were just right you'll get lucky and find -20 db on your volume control gives you the in room pink noise at 85 db SPL.

I do adjust my video rig this way. I find it works pretty well. It may be a touch loud on some movies, but I rarely have to lower it more than - 6 db and usually not more than about 3 db. Now for plain TV for other programming than movies or other purposes I may lower volume some more.

 

[RichB]

Fully agreed, apparently studies shows not everyone prefer even (e.g. 2nd) harmonics to odd (iirc, even Nelson Pass has said something like that but..), some do prefer 3rd to 2nd, again just examples. I am just saying people can't tell the difference between 0.05% and 0.03%, if measured properly, full bandwidth and are "total"+N, or not. You may be an exception, I have the impression from your past posts that you are more sensitive to minute differences than a lot of ppl.

I bought my Bryston and Halo amp because of their low distortion specs, same for my preamp, external DACs. But can I hear difference? I don't know, but I doubt I can past any blind test, even single blind..

Come to MA., bring your Bryston, and we'll do a blind against the AHB2. I have no idea how this will come out. Too bad you can't find a local AHB2 to try out. I funded a good portion of my upgrades by parting with the gear I was no longer using.

Class A/B with similar protection, power supply design, and construction are going to sound about the same. I did a level matched comparison of the ATI AT4000 and AT522NC and these amps were distinguishable.

- Rich

 

[GrimSurfer]

Finding audible differences between a Bryston and a Benchmark would likely be difficult because I don't think we'd be looking at several orders of magnitude differences at normal levels.

But compare a Bryston and a 1080 in 2ch mode, I'd be astonished if a reasonably trained ear couldn't distinguish between two amp stages that ARE magnitudes of difference apart.

And that's what this is all about: audio fidelity. It's not about making excuses for a product because the manufacturer decided to cram every square inch with components AND kick it out the door at $1500.

 

[DonH56]

Frequency bands are not split but now 4 "amplifiers" are used instead of 2, and each of them drives lesser number of drivers, more impedance and less stress on each amplifier.
I know that total power comes from a single transformer, but now it is more evenly shared across 4 amps (one cannot overload and distort as quickly as before, when it was driving woofers and tweeters together). So why there is not any benefit if it has individual amplifiers?

No, "lesser number of drivers" does not mean "more impedance" and "less stress". You'd have to look at the impedance curve of your speakers to estimate the impedance -- it changes over frequency. Further more, if your speaker was an ideal 8-ohm load at all frequencies, then when you split the bass and treble then each amp still sees 8 ohms. The impedance does not automatically double when you bi-amp.

The voltage output of each amplifier is not changed in that scheme so I would not expect much if any gain. Spreading the load among four amplifiers means the overall power draw from the transformer increases because the amplifiers are not 100% efficient -- there is a little "standing" power in each amp and since they all swing the full voltage there is more power used by four amplifiers than by two even if the same power is delivered to the speaker.

 

[RichB]

This explains it better than THX did.
http://www.acousticfrontiers.com/2013314thx-reference-level/

I think you are somewhat confused on the 0 db reference level. It doesn't mean you'll automatically use maximum or 0 db on your volume control of your gear.

What you do is play a pink noise signal which is -20 db from maximum digital signal level. While playing this you measure SPL at the listening position and adjust your volume so you get 85 db SPL. Your volume control may read -26 db or any other number. If it were -26 db for your gear, this is your nominal THX 0 db reference level. It allows for 20 db of headroom over normal average levels (or if your reference level were -26 db you'll have 26 db headroom and 6 of it will never be needed). Your speakers would then play max level in room at 105 db SPL (assuming they are up to it). If you listen at this level you are hearing exactly the sound levels the people who made a movie sound track intended for you to hear. Just keep your volume control set to - 26 db if that is what gave you reference on the pink noise signal.

Now if you found your volume was at some point higher than -20 db to meet reference SPL in your room, your gear might not be capable of full sound level to meet THX spec. If things were just right you'll get lucky and find -20 db on your volume control gives you the in room pink noise at 85 db SPL.

I do adjust my video rig this way. I find it works pretty well. It may be a touch loud on some movies, but I rarely have to lower it more than - 6 db and usually not more than about 3 db. Now for plain TV for other programming than movies or other purposes I may lower volume some more.

Actually, this is not what THX says on their site:

On every THX Certified AV Receiver the “0” on the volume dial represents THX Reference Level, the exact volume level used by the moviemakers and sound artists in the studio.

You have a preference, I have a preference and neither of us would attenuate to achieve the published THX specification for an AVR or processor.

- Rich

 

[RichB]

Finding audible differences between a Bryston and a Benchmark would likely be difficult because I don't think we'd be looking at several orders of magnitude differences at normal levels.

But compare a Bryston and a 1080 in 2ch mode, I'd be astonished if a reasonably trained ear couldn't distinguish between two amp stages that ARE magnitudes of difference apart.

And that's what this is all about: audio fidelity. It's not about making excuses for a product because the manufacturer decided to cram every square inch with components AND kick it out the door at $1500.

The definition of fidelity is to have an audio amplifier that when playing music into load has inaudible distortion, linear and non-linear.
The argument could be made that the current measurements could miss a particular amps disfunction and that new and different technology may required additional measurements to identify new misbehavior. I believe this is possible.

However, amps and other components with obvious flaws are unacceptable. I am immune to protestations of toe-tapping rhythm.

- Rich

 

[amirm]

I wonder if measurements at "0db volume" is the most appropriate approach to these AVRs, since the manufacturer might use the volume level display as an arbitrary setting for marketing (or whatever) purposes.

I don't use 0 dB to mean anything. I go by what output voltage is produced. Stand-alone DACs produce 2 volts for RCA and 4 volts for balanced. I adjust the volume to achieve these and I report what I had to do with the volume control to get there. I often report results at more than one level.

Looking at various specifications in AVRs printed manuals, it looks like for this family of devices the industry standard for preouts is 1V.

That's because their internal amplifiers don't have a lot of power so they need a lot of voltage on the input to get there. A more powerful amplifier used externally will need more input voltage for the same voltage gain.

The AVRs are capable of producing the nominal voltages. The problem is that they don't shut down the power to their power amps causing them to severely clip. That in turn screws up the power for the DAC, etc. causing their performance to degrade.

 

[blueone]

No, "lesser number of drivers" does not mean "more impedance" and "less stress". You'd have to look at the impedance curve of your speakers to estimate the impedance -- it changes over frequency. Further more, if your speaker was an ideal 8-ohm load at all frequencies, then when you split the bass and treble then each amp still sees 8 ohms. The impedance does not automatically double when you bi-amp.

Perhaps I'm not understanding what you're saying. I'm reading this as you are contending that not only is the voltage the same on both amplifier outputs in a passive bi-amp pair, but the current is the same too. So, to extend the thought, a 40Hz signal will be delivered to the midrange-tweeter section of the crossover at the same current level as it is to the woofer section of the crossover. Does that mean you think the 40Hz current delivered to the midrange-tweeter section is dissipated as heat in the crossover components?

The voltage output of each amplifier is not changed in that scheme so I would not expect much if any gain. Spreading the load among four amplifiers means the overall power draw from the transformer increases because the amplifiers are not 100% efficient -- there is a little "standing" power in each amp and since they all swing the full voltage there is more power used by four amplifiers than by two even if the same power is delivered to the speaker.

It is correct that Class A & AB amplifier output stages have a standing bias (and it can be considerable in some designs), so powering four Class AB amplifiers in place of two will consume more power just sitting there idling. With the Class D amplifiers in many receivers I'm not sure how significant this factor is. (I suspect it is insignificant, but I don't know for sure.)

 

[HammerSandwich]

Q1: Not really. Think about how you measure spl. You (or the amp/AVR) do a frequency sweep. The only thing that changes is the frequency. So there is no peak to measure.

For the 85dB calibration, okay. But is the digital signal steady-state at 0dBFS, when you get 105dB in your chair? (I'll grant the DR0 Iggy Pop album, but I hope no one plays that at reference level in the house. If they do, either hardware or ears will fail soon.)

So if it can sustain 85 dB spl at the seating position, hitting 15-20 dB during a very short duration peak is likely. There are many reasons for this, of the least of which include the capacity of a good amp's power caps and its excess cooling capacity.

Too many variables, IMO. Or tautology from "good amp."

 

[GrimSurfer]

Too many variables, IMO. Or tautology from "good amp."

No tautology required.

The variable is whether you have an amp and speakers, sufficiently sized with the space, that can sustain 85 dB at the listening position all day long and have the headroom (in the case of the amp) to peak when needed at 100-105 dB without clipping or tripping a thermal protection circuit.

THAT is a good amp.

OTOH, one could always set things up to THX standards and play the 1997 pressing of "Raw Power" knowing that the peaks will stay within ~ 3dB of reference levels!

http://dr.loudness-war.info/album/view/162547

 

[Blumlein 88]

Actually, this is not what THX says on their site:



You have a preference, I have a preference and neither of us would attenuate to achieve the published THX specification for an AVR or processor.

- Rich

I think you still aren't thinking about this correctly. Yes every THX certified device will have 0 db Reference at 0 db on their volume. If however you don't have such a device it will be somewhere else. It also would mean their 0 level would give max signal output at 105 db SPL in room.

Now I did have a mistake in my last post. If you get 0 db for the 85 db SPL signal, you still might have room for the full scale 105 db SPL signal, but it will depend upon the particulars of your system. So my example of that wasn't quite right.

And yes I do actually set my video rig this way. I send it the right signal and set it for 85 db SPL using the two front channels. Which would be equivalent to 82 db from one channel. I do attenuate for that because my front speakers are relatively efficient and my power amps are overkill.

 

[GrimSurfer]

Yup. Without knowing the distance to the listener and speaker sensitivity, and in the absence of in room calibration, that 0 dB read-out on the display is utterly meaningless.

 

[DonH56]

Perhaps I'm not understanding what you're saying. I'm reading this as you are contending that not only is the voltage the same on both amplifier outputs in a passive bi-amp pair, but the current is the same too. So, to extend the thought, a 40Hz signal will be delivered to the midrange-tweeter section of the crossover at the same current level as it is to the woofer section of the crossover. Does that mean you think the 40Hz current delivered to the midrange-tweeter section is dissipated as heat in the crossover components?

Nope, the opposite, never said current is the same, unless I made a typo somewhere. It depends upon the frequency distribution and impedance over frequency. I used an example of both treble and bass being 8 ohms as an illustration of how load impedance might not change since the OP insists that it is "more" when you bi-amp. The speaker's impedance does not change when you passively bi-amp though is split into two parts.

Couple of examples follow (I did not look up their crossover frequencies).

B&W 702 has low impedance spots at high and low frequencies:

ML Montis has its lowest impedance in the HF region:

Power distribution is frequency dependent so power, and current, will change even if voltage does not. Most audio amplifiers are designed as voltage sources and so if the voltage clips in one amp it will clip in all in the AVR passive bi-amping scenario. In fact there will be a bit of headroom improvement if power is significantly different due to less drop across the output transistors, but that is typically very small (<1 dB) IME. And that is only at or near clipping. But as was noted above, Amir has found some amps that clip "poorly", but I suspect those are outliers. In any event, if you are clipping, chances are you need more like 3~6 dB of additional headroom and thus 2 to 4 times the power.

Two things are getting conflated that I was trying to explain (obviously poorly): impedance is not necessarily "more" when you bi-amp, and power delivery depends upon impedance and the frequency content in the source. The usual argument about frequencies is that treble takes much less power than bass. That is not always true, either -- depends upon where the crossover is for the speakers, and perhaps even less so for systems that have subwoofers and thus offload the deepest bass.

It is correct that Class A & AB amplifier output stages have a standing bias (and it can be considerable in some designs), so powering four Class AB amplifiers in place of two will consume more power just sitting there idling. With the Class D amplifiers in many receivers I'm not sure how significant this factor is. (I suspect it is insignificant, but I don't know for sure.)

Yes, I agree, I do have some experience with various amplifier topologies though limited in audio. Class D's efficiency is also worst at low output. The quiescent current is usually lower, sometimes much lower, than for class A or AB amplifiers but is not zero and thus using four amps, even class D amps, will use more power than two amps for the same total output power. I believe that is all I said (made no attempt to quantify it; I have seen numbers from maybe 1 to 10 W but not something I track). For class D in AVRs implementing "passive" bi-amping it is worse because the amplifiers are are swinging (and slewing) full voltage even if there is less power delivered to the load. Of course, they are more efficient, but the net power still goes up. A truly bi-amplified system, using active or passive line-level crossovers before the power amps, is more likely to provide additional headroom (voltage and current), but even that is not guaranteed. Again depends on the relative frequency distribution and impedance of the speaker over frequency.

IME/IMO - Don

 

[restorer-john]

The problem is that they don't shut down the power to their power amps causing them to severely clip. That in turn screws up the power for the DAC, etc. causing their performance to degrade.

Even if they are using completely separate t/xf windings for the various stages and fully regulated supplies for the digital circuitry, the current draw on the main amplifier tappings is enough to adversely affect those low level (digital supply) rails.

It's just poor design, yet again. The big money is in the transformer/s and PSUs, and it's clearly not up to the task.

Back in the day, multiple transformers, multiple tappings and regulated PSUs were used if the product was a single box. I've got preamplifiers with three transformers- one for each channel and one for the control circuitry. I've got a TOTL D/A converter with 3 toroidal power transformers and goodness knows how many individual regulated rails.

Everything starts and stops with the power supply. Just like getting fuel and air into your car engine.

 

[blueone]

Nope, the opposite, never said current is the same, unless I made a typo somewhere. It depends upon the frequency distribution and impedance over frequency. I used an example of both treble and bass being 8 ohms as an illustration of how load impedance might not change since the OP insists that it is "more" when you bi-amp. The speaker's impedance does not change when you passively bi-amp though is split into two parts.

Power distribution is frequency dependent so power, and current, will change even if voltage does not. Most audio amplifiers are designed as voltage sources and so if the voltage clips in one amp it will clip in all in the AVR passive bi-amping scenario. In fact there will be a bit of headroom improvement if power is significantly different due to less drop across the output transistors, but that is typically very small (<1 dB) IME. And that is only at or near clipping. But as was noted above, Amir has found some amps that clip "poorly", but I suspect those are outliers. In any event, if you are clipping, chances are you need more like 3~6 dB of additional headroom and thus 2 to 4 times the power.

Two things are getting conflating that I was trying to explain (obviously poorly): impedance is not necessarily "more" when you bi-amp, and power delivery depends upon impedance and the frequency content in the source. The usual argument about frequencies is that treble takes much less power than bass. That is not always true, either -- depends upon where the crossover is for the speakers, and perhaps even less so for systems that have subwoofers and thus offload the deepest bass.

Excellent. I suspected that I misunderstood you. The current flowing on the two cables for one speaker will be different at different frequencies.

I agree that making gross generalizations about the power requirements of woofers versus mid-high sections may be incorrect. For example, in RichB's prized Salon2s, the woofer section is more like a subwoofer section, as the crossover is at 150Hz. The mid-tweeter section may on quite a bit of material need more power delivered from the amplifier than the woofer section does.