Review and Measurements of Benchmark AHB2 Amp

[Vasr]

What these sorts of tests show of course is that all the 'audiofool' reviewer nonsense you read is just that. .

The flip-side of this is that chasing SINAD is also an "audiofool" nonsense. If you cannot reliably hear differences between amps with a wide variance of measurements then the measurements become an end in itself not means to an end (other than detecting broken things).

But it is a hobby as much as buying a well-engineered wrist watch that can work 600' under water or keep time within 0.00000000001 sec. So, no harm in either case (other than to a wallet).

Audio is what you believe you hear.

 

[RichB]

Rich,

Once all the amplifiers were level matched we used the Main Listening Volume adjustment which in parallel raises or lowers all channel pairs and therefore XLR connected amplifiers. The auditioned got to choose any volume they wanted with the Marantz remote in hand (Though we agreed not to go above -10dB on the MLV as a precaution against source clipping since the individual channels were boosted into the +6dB +7dB range on a couple of the amps. (each amp had positive channel trims because of the low sensitivity ML Renaissance speakers)). The auditioner then could then switch amplifiers from the main listening position at their own whim as fast as they could throw light switches.

For the lesser portion where someone was controlling the switchboard out of sight for the auditioner (to test if any possible perceived preference from their time commanding the switching device would remain consistent when they couldn't actually see which amp switches were flipped) the tested auditioner was always in the MLP as well.

We used a voltmeter and sine wave sweeps in the original 2013 testing as part of our setup process, and it took a much greater amount of time to set up, but those doing the set up agreed after going through that process that it didn't matter in this specific type of audition/testing. Why? It's just one pair of speakers, with four amps connected to it and the auditioners goal is trying to detect a difference - if they can't, then the sine wave doesn't matter (also under the expectation the amps should be producing an absolutely identical frequency response sweep unless an amp is broken) IF OF COURSE - it was determined the auditioners could detect and reliably identify a difference consistently - - at that point we'd want to ferret out the difference. So that is simply a time constraint. We had agreed this would be a half day thing going in. Even still, it took us about 3 hours to carefully set up in a double blind manner - and then we auditioned thereafter for 2-3 hours.

Here's a video from our 2013 test. We didn't capture video this time, but this will kind of show you the idea.

I applaud your efforts.

I find it far easier to set the volume, play a 1kHz sine wave and measure the voltage.
In the past, I find the volume settings to produces 2.83 volts (or as close as possible) which represents the 1 watt into an 8 Ohm speaker.
There is no issue with sound interference in the measurement. Once your amps setup in parallel, it should have taken under 10 minutes to obtain the volume output adjustment required for most amplifiers.

The AV8805 channels may be using different DACs (may not matter but should be noted).
All Channels Stereo is a bit of a black box. It may well uses DSPs to copy all channels but it may not. Marantz, after all, picked a slow roll off filter that attenuates below 20 kHz because they feel it sounds "better". Maybe they have done something in All Channel Stereo to make it sound better.

I prefer splitting the XLR and comparing two amps at a time to reduce fatigue and perform tests over a few days. I know this is not always possible but it help improve my confidence levels.

I have SBT compared the ATI AT4002 (class A/B) and ATI AT522NC (NCore) and found these amplifiers distinguishable. I preferred the AT4002 which sounded a bit more natural driving Revel M20s. However, I really liked the dynamics of the AT522NC so I ended up buying the 5 channel version for the family vacation home that also uses these speakers. For this test, two volume levels were used, one at 2.83 volts and one at 10 dB higher, voltage matched with the Oppo UPD-205 split to both amps. The source was via the USB DAC input, that has no possible processing, driven by Roon connected RPI.

In the end, these tests are these people, that room, those speakers which (IMO) is suitable for personal buying decisions.

- Rich

 

[Coach_Kaarlo]

OK, I will bite back.

Just because something might be electrically infinitesimally different as measured by laboratory test equipment, is not proof that the human ear - an organ which massively less capable than such devices - has any chance of detecting these changes.

It's like the businesses trying to convince people that 24/192 sounds better than 24/48 or 24/96, for example, when the human ear has no chance of hearing anything above 20 KHz, and for the old fogeys like me who follow this nonsense - probably not much over 15 KHz. Or the silly "stepped sine wave" trying to show how hires audio is better, when no steps exist in lower res audio anyway. It's quack science.

The entire industry is jam packed full of such guff, seeking to explain audible differences which do not exist. And I can see why, to admit that they do not would be the demise of a multi-million (billion?) pound business scam.

I am quite sure some people are convinced different speaker cables sound different. That of course is their prerogative. As it is mine to be 100% certain in my mind that they are fooling themselves and that they are wrong.

(I am not talking about edge cases with extremely high capacitance cables or 1000 ft cables or what have you. I am merely suggesting that a bog standard pair of Canare 4S11 will sound INDISTIGUISHABLE to some esoteric nonsense costing 10x, 50x, 100x as much. It is snake oil, pure and simple.)

So we agree. Great.

 

[misterdog]

where someone was controlling the switchboard

So the Marantz pre/pro with SINAD of 91.5dB was not considered a limiting factor ?
https://www.audiosciencereview.com/...urements-of-marantz-av8805-av-processor.6926/

Considering that I kept the output level at 4 volts, i.e. well below max, the rising distortion is worrisome.

-Amir

You used of a Panasonic disc player as a source with SINAD of 92dB.

Given that the AHB2 Amp has SINAD of 112.5

And yet you set the input for the AHB2 at the lowest input of 2V .

the Benchmark amp's input sensitivity position switch was in the default 2vrms setting.

even though connected via XLR.

we had Ryan's wife make the XLR connections to the Marantz prepro

So you used a mediocre source connected to a mediocre A/V preamp connected to four amplifiers with dozens of metres of cable, and a 'switchbox'.
All fed into -
Loudspeakers perched on the top of beer crates ?

And concluded that all amplifiers sound the same

Sorry but I'm out.

 

[RichB]

So the Marantz pre/pro with SINAD of 91.5dB was not considered a limiting factor ?
https://www.audiosciencereview.com/...urements-of-marantz-av8805-av-processor.6926/
-Amir

Your used of a Panasonic disc player as a source with SINAD of 92dB.

Given that the AHB2 Amp has SINAD of 112.5

And yet you set the input for the AHB2 at the lowest 2V .



even though connected via XLR.



So you used a mediocre source connected to a mediocre A/V preamp connected to four amplifiers with dozens of metres of cable, and a 'switchbox'.
All fed into -
Loudspeakers perched on the top of beer crates ?

And concluded that all amplifiers sound the same

Sorry but I'm out.

The listening tests used the ceiling and surround channels and a DSP mode called "All Channel Stereo" which could also be named "Extra Special All Channel Stereo". The only thing we know for sure, is that is absolutely not Stereo

I am not saying that the test is invalid but the SINAD of the AV8805 measured here are for the FR/FL channels in direct mode.
There is no measured SINAD for All Channel Stereo mode on those channels. SINAD may be the same but probably less.

- Rich

 

[Pdxwayne]

So the Marantz pre/pro with SINAD of 91.5dB was not considered a limiting factor ?
https://www.audiosciencereview.com/...urements-of-marantz-av8805-av-processor.6926/
-Amir

You used of a Panasonic disc player as a source with SINAD of 92dB.

Given that the AHB2 Amp has SINAD of 112.5

And yet you set the input for the AHB2 at the lowest input of 2V .



even though connected via XLR.



So you used a mediocre source connected to a mediocre A/V preamp connected to four amplifiers with dozens of metres of cable, and a 'switchbox'.
All fed into -
Loudspeakers perched on the top of beer crates ?

And concluded that all amplifiers sound the same

Sorry but I'm out.

Not to mention the sitting position in the room may have added additional adverse effect on SNR....

 

[misterdog]

Though I often sit at 90 degrees, to one side of my system, to hear the benefit of one channel, in mono, through one ear first .

Never tried it with a Candy Land pizza though.

 

[JimB]

I give this effort significant credit for what it was attempting - find a clear preference under reasonable home conditions. This is not proof that no one could ever hear any difference under any conditions with any load. But, I am way more suspect of the significance of uncontrolled claimed distinctions, than honest inability to find clear, consistent differences.

I too wondered about the use of CANDY LAND 'supports' - decidedly not audiophile grade.

 

[anmpr1]

The flip-side of this is that chasing SINAD is also an "audiofool" nonsense.

Not necessarily. In certain applications super low noise might be necessary. Also, one may want to own engineering excellence for its own sake, and then SINAD (along with other factors) would be important criteria. There is nothing foolish about excellence in engineering.

What would be foolish would be to pay the same amount (or more) for a lot worse. For example, one company out there sells a 2.3 watt (that's two point three) per channel SET for the same price as an AHB-2. S/N is claimed to be almost 80dB down.

 

[Pdxwayne]

I give this effort significant credit for what it was attempting - find a clear preference under reasonable home conditions. This is not proof that no one could ever hear any difference under any conditions with any load. But, I am way more suspect of the significance of uncontrolled claimed distinctions, than honest inability to find clear, consistent differences.

I too wondered about the use of CANDY LAND 'supports' - decidedly not audiophile grade.

It proves to us that in a mediocre room, in mediocre seating position, with mediocre source and mediocre preamp, all one needs is mediocre amp. ; )

 

[Archaea]

So the Marantz pre/pro with SINAD of 91.5dB was not considered a limiting factor ?
https://www.audiosciencereview.com/...urements-of-marantz-av8805-av-processor.6926/
-Amir

You used of a Panasonic disc player as a source with SINAD of 92dB.

Given that the AHB2 Amp has SINAD of 112.5

And yet you set the input for the AHB2 at the lowest input of 2V .



even though connected via XLR.



So you used a mediocre source connected to a mediocre A/V preamp connected to four amplifiers with dozens of metres of cable, and a 'switchbox'.
All fed into -
Loudspeakers perched on the top of beer crates ?

And concluded that all amplifiers sound the same

Sorry but I'm out.

Draw your own conclusions, or better yet test your own conclusions in some semblance of an objective blind test.

In our recent test we used the very best current gen flagship Pre-Pro from Marantz.
We tested the very best amplifier (tested here so far) against amplifiers with a SINAD score of 77 and 76.
So even if the Marantz pre-pro limited the SINAD to 92dB, the Behringer A800 and Crown XLS 1500 SINAD is much lower yet - a 15dB drop from the Marantz 8805's measurement. We left the AHB2 at default 2volt input to better match the other amps input voltage tested. The Marantz pre-pro doesn't have configurable pre-out voltage, so that particular complaint is really immaterial in this use case scenario anyway.

The testing we participated in simply echoed a real world use case for one of our local members with his speakers, and his prep-pro, and his room. He wanted to buy a stack of amps, and after our demo he did - armed with his own conclusions from our testing session.
I'm not telling anyone what to buy, simply sharing our testing scenario. Our group's conclusion is more appropriately summarized as - the four individual audio enthusiasts who participated in this test (aged ~35 - ~65), found our human ears could not reliably detect the objectively measurable differences between these amplifiers in blind audition, in a real world audio audition.
----
The video I linked was from our first test in 2013, and the speakers on the deck on "beer crates" as you put it was the informal break listening area where multiple sets of speakers were brought to just listen to and hang out and talk. The blind testing in that 2013 meet we did downstairs at that meet in the HT room with one set of speakers, the JTR 212HT. The following test we did in my HT room with Klipsch Reference Towers. That was the same year, 2013. Totally different amplifiers used in those first two tests, than what was used in this test. No overlap at all.

This recent testing was done in third different room with entirely different gear. The only thing the same was the DIY switching device (and the results that we participants were unable to distinguish anything except the lowest quality amplifier in the mix. The $20 Lepai T-Amp this time and the cheapest Onkyo HTIB money can buy receiver the first time).

----
As to not the right speakers, not the right song, weren't listening for the single specific audible difference on some esoteric music track with our head cocked just right?

Meh.

Run your own test, and pick your own material. Share the results! That's the best path.

Just do note that sighted listening is incredibly biased and can be wholly discounted if you want to be completely objective.

In our testing even when people were operating the switchboard and started thinking they had a preference, and stating such - they couldn't replicate or identify that preference when they didn't have control or eyes on the switchboard. It's really an interesting study in the way the brain works at some level. 3 of the 4 of us picked one amp that sounded subtly different when we controlled the switchboard. Funny because we didn't even pick the same one. It could have been that one of us said one amp sounds different to me, and then we all decided yes one must sound different. (Maybe some of us thought the AHB2 should sound better, Maybe some of us thought the 1980's refurbished amp would sound worse (as I did)).
This thought of being able to identify one amp of the four was proved false when the auditioner didn't control the switchboard and was clearly proven to have no idea when tested by someone else. Even identifying the same amp as the best and the worst moment apart as I called out in my main writeup.
We talked about this a little bit as a group after our test. How does this happen?
What we guessed was that perhaps it was as simple as some refrain, chorus, or particular note that we originally heard playing on one of the amps - we liked or disliked in that sample, and we picked that particular switch as being the one we liked or didn't, and that belief took increasing hold through the rest of our audition to think switch x sounds the best began to get stronger with some manner of unreliable confirmation bias. So point being - even in a blind test - where you don't know what switch is attached to what amp -- you can build a improper bias, and start to develop an affinity for a switch that isn't reliable or repeatable.

I encourage every enthusiast participate in something like this. It's fun.

 

[misterdog]

Just do note that sighted listening is incredibly biased and can be wholly discounted if you want to be completely objective.

Though if you use the lowest distortion source, into the lowest distortion amplifier, into the lowest distortion loudspeakers,you are most likely to hear.

A true representation of the recorded source.

No switch boxes required.

No blindfolds required.

Reproducing an earthquake 'effect' in a movie soundtrack is not the same as reproducing the subtleties of a Cello played by a maestro.
Most systems can perform, the former (even mine) though not all can perform the latter, and not all amplifiers 'sound' the same.
The quality of sound is dictated by the weakest link(s) in the chain.

It's really an interesting study in the way the brain works at some level.

 

[Vasr]

Also, one may want to own engineering excellence for its own sake, and then SINAD (along with other factors) would be important criteria. There is nothing foolish about excellence in engineering.

That is exactly what I said in comparison to buying a watch that is well-engineered. The "audiofoolery" is when one tries to relate SINAD to audibility beyond a certain point. That is no different from the "audiophoolery" of golden ears. Two sides of the same coin.

What would be foolish would be to pay the same amount (or more) for a lot worse. For example, one company out there sells a 2.3 watt (that's two point three) per channel SET for the same price as an AHB-2. S/N is claimed to be almost 80dB down.

The flip-side of this, of course, is that it would be foolish to pay X times as much for something that measures far better if it makes no difference in audibility (as in "audiofoolery"). As for paying for a work of art/engineering, sure. That is a subjective preference.

 

[Pdxwayne]

Draw your own conclusions, or better yet test your own conclusions in some semblance of an objective blind test.

In our recent test we used the very best current gen flagship Pre-Pro from Marantz.
We tested the very best amplifier (tested here so far) against amplifiers with a SINAD score of 77 and 76.
So even if the Marantz pre-pro limited the SINAD to 92dB, the Behringer A800 and Crown XLS 1500 SINAD is much lower yet - a 15dB drop from the Marantz 8805's measurement. We left the AHB2 at default 2volt input to better match the other amps input voltage tested. The Marantz pre-pro doesn't have configurable pre-out voltage, so that particular complaint is really immaterial in this use case scenario anyway.

The testing we participated in simply echoed a real world use case for one of our local members with his speakers, and his prep-pro, and his room. He wanted to buy a stack of amps, and after our demo he did - armed with his own conclusions from our testing session.
I'm not telling anyone what to buy, simply sharing our testing scenario. Our group's conclusion is more appropriately summarized as - the four individual audio enthusiasts who participated in this test (aged ~35 - ~65), found our human ears could not reliably detect the objectively measurable differences between these amplifiers in blind audition, in a real world audio audition.
----
The video I linked was from our first test in 2013, and the speakers on the deck on "beer crates" as you put it was the informal break listening area where multiple sets of speakers were brought to just listen to and hang out and talk. The blind testing in that 2013 meet we did downstairs at that meet in the HT room with one set of speakers, the JTR 212HT. The following test we did in my HT room with Klipsch Reference Towers. That was the same year, 2013. Totally different amplifiers used in those first two tests, than what was used in this test. No overlap at all.

This recent testing was done in third different room with entirely different gear. The only thing the same was the DIY switching device (and the results that we participants were unable to distinguish anything except the lowest quality amplifier in the mix. The $20 Lepai T-Amp this time and the cheapest Onkyo HTIB money can buy receiver the first time).

----
As to not the right speakers, not the right song, weren't listening for the single specific audible difference on some esoteric music track with our head cocked just right?

Meh.

Run your own test, and pick your own material. Share the results! That's the best path.

Just do note that sighted listening is incredibly biased and can be wholly discounted if you want to be completely objective.

In our testing even when people were operating the switchboard and started thinking they had a preference, and stating such - they couldn't replicate or identify that preference when they didn't have control or eyes on the switchboard. It's really an interesting study in the way the brain works at some level. 3 of the 4 of us picked one amp that sounded subtly different when we controlled the switchboard. Funny because we didn't even pick the same one. It could have been that one of us said one amp sounds different to me, and then we all decided yes one must sound different. (Maybe some of us thought the AHB2 should sound better, Maybe some of us though the 1980's refurbished amp would sound worse (as I did)).
This thought of being able to identify one amp of the four was proved false when the auditioner didn't control the switchboard and clearly had no idea when tested by someone else. Even identifying the same amp as the best and the worst as I called out in my main writeup.
We talked about this a little bit as a group after our test. How does this happen?
What we guessed was that perhaps it was as simple as some refrain or chorus or particular note that we originally heard playing on one of the amps - we liked or disliked that small short sample, and we picked that particular switch as being the one we liked or didn't, and that took hold through the rest of our audition that says switch x sounds the best began to get stronger with some manner of unreliable confirmation bias. So a difference that wasn't reliably there in reality with subsequent testing , became there, and the bias we were building was revealed beyond a doubt (to us anyway) as false when we were no longer controlling the switchboard and didn't know which switch was engaged during the audition. So point being - even in a blind test - where you don't know what switch is attached to what amp -- you can build a improper bias, and start to develop an affinity for a switch that isn't reliable or repeatable.

I encourage every enthusiast participate in something like this. It's fun.

How about you do a simple test of "multi CH stereo" vs "pure direct" sound quality comparison of AV8805?

I have AV8801. "Pure direct" is cleaner than "multi ch stereo" using main speakers in large setting and only main speakers amp on. More 3D with "Pure Direct".

Using multi ch stereo gives you worse SNR than what is measured by Amir.

 

[Vasr]

Though if you use the lowest distortion source, into the lowest distortion amplifier, into the lowest distortion loudspeakers,you are most likely to hear.

A true representation of the recorded source.

The point of the test is that if you heard a "true representation" of the recorded source in one setup and you heard a slightly different version of the recorded source in another set up and you could not make out any difference between the two, what benefit has been realized from listening to the true representation of the recorded source?

Is the true representation as performed by the artist? As recorded by the microphones? As mixed by the engineer? As digitized/processed into the distribution media? Because all of them are different. Why is one of any particular benefit?

You can have water from a natural spring transported in a special temperature controlled equipment that maintains the temperature and purity of the water as if you drank it directly from the source or you can have it collected by a plant that mass-distributes it and may be adds some additional inert elements when you consume it buying off the grocery shelf. If you cannot distinguish between the two in taste or suffer any side-effect from the latter, what exactly have you achieved? Other than as a conversation piece at a party for bragging rights?

 

[Archaea]

How about you do a simple test of "multi CH stereo" vs "pure direct" sound quality comparison of AV8805?

I have AV8801. "Pure direct" is cleaner than "multi ch stereo" using main speakers in large setting and only main speakers amp on. More 3D with "Pure Direct".

Using multi ch stereo gives you worse SNR than what is measured by Amir.

I think that’s a test better performed by Amir.
I only have a calibrated Omnimic which can do frequency sweeps and harmonic distortion, impulse response, decay waterfalls etc. It’s not a scope.
I would not expect there to be any difference in stereo vs pure direct if steps are taken to make stereo the same settings as pure direct. To my understanding the pure direct mode just disables all processing on the signal, which as I noted in my testing setup article explanation we absolutely did manually in the Marantz settings menus before our testing. If differences remain in SINAD from there, I don’t have the type of equipment that could capture them.

To make it clear what I can measure with my Omnimic - it’s more along the lines of something like this.
https://www.avsforum.com/threads/wh...g-what-is-going-on-behind-the-scenes.3169754/

 

[Pdxwayne]

I think that’s a test better performed by Amir.
I only have a calibrated Omnimic which can do frequency sweeps and harmonic distortion. Neither of which would I expect to be any difference in stereo vs pure direct if identical setup is used and steps are taken to make stereo the same settings as pure direct. To my understanding the pure direct mode just disables all processing on the signal, which as I noted in my testing setup article explanation we absolutely did manually in the Marantz settings menus. If differences remain in SINAD from there, I don’t have the type of equipment that could capture them.

To make it clear what I can measure with my Omnimic - it’s more along the lines of something like this.
https://www.avsforum.com/threads/wh...g-what-is-going-on-behind-the-scenes.3169754/

Do you have revelling speakers? Simply set your speakers to large. Then switch between stereo and pure direct. Change nothing else. This is what I am doing now. Stereo have muddier mid. Pure direct is cleaner.

My speakers is Paradigm Persona B. Very revealing.

 

[misterdog]

The point of the test is that if you heard a "true representation" of the recorded source in one setup and you heard a slightly different version of the recorded source in another set up and you could not make out any difference between the two, what benefit has been realized from listening to the true representation of the recorded source?

If you start with a 'distorted' source you will never hear anything other.

The greatest levels of distortion are from loudspeakers, then amplifiers then DAC's.

 

[Vasr]

If you start with a 'distorted' source you will never hear anything other.

The greatest levels of distortion are from loudspeakers, then amplifiers then DAC's.

 

[Archaea]

Do you have revelling speakers? Simply set your speakers to large. Then switch between stereo and pure direct. Change nothing else. This is what I am doing now. Stereo have muddier mid. Pure direct is cleaner.

My speakers is Paradigm Persona B. Very revealing.

Did you do these things first?

We disengaged Audyssey, all tone trims, restorer, and any form of EQ. We set all speakers to large to remove any crossover.

I don't have the Marantz 8805 I have the Denon X6700H, and my speakers are not full range speakers, so I need subwoofers.

But as to how revealing my own speakers are - they will be tested very shortly -- matter of fact. We shall both see.


https://www.audiosciencereview.com/...on-help-for-jbl-cbt-70j-1-measurements.17601/

I should probably make our blind test into its own post at AVS, so these types of things can be discussed without mucking up a owners/review thread with posts that some may consider off topic...