Denon AVR-A1H High-end AVR Review

[Brambo67]

Beautiful piece of engineering resulting in fine performance for home theater use as well as music listening in almost every format available. Expensive too as well as big (huge) and heavy. For my use to expensive and big. Engineering these machines and bringing them to the market results in continuous (incremental) improvements in the lower tier products as well. If you look at DM’s product portfolio there’s an AVR/AVC for almost every requirement. Nice to see Denon is also raising the bar all the time.

 

[mtbdudex]

Who here has a A1H besides me?
I had the 8500H prior.

3 Denon remotes; 4308CI from 2008 I still use for multi room audio, 4500H family room 5.1.4, and A1H basement HT 9.3.6

The A1H remote definitely a premium feel, metal top with wrap.

Side there is a light button on the A1H

Like I posted earlier though, I use the iPhone Denon remote app for digging into the AVR menu items quickly. FWIW I use a Harmony Elite in my basement, the hand remote 90% of times but sometimes the iPhone app also.

 

[Atanasi]

Because I was running the tests from my laptop, I could not drive the unit with HDMI.

Is there a problem with connecting a dongle to output HDMI? So, a dongle is not compatible with the tests.

 

[peng]

(BTW, I used to own a Parasound power amp. I've owned lots of amps, and hoped and expected the Parasound would be great, but it sounded poor)

That doesn't look great, but here are some examples of what good looks like.

Denon A1H , JBL SDR-35 , Lyngdorf MXA-8400:

View attachment 427209View attachment 427210View attachment 427211
MXA-8400:

Thanks again, but a) the JBL opened up a gap at just slightly over 30 W and the gap reached about 5 dB at rated output, it also has poorer performance in absolute sense. Still it is class G so I would say it is fair to compare with the characteristics of class AB amps to a large extent.

b) The Lyngdorf is class D, so it is a different animal and the lesser frequency dependent should be expected, yet, it also has a gap of > 10 dB, better behaved than the class AB A1H but again that's expected as it is not class AB.

After researching this issue on ASR measured amps, I made the following observations:

1) It seems that if the measured amp has higher SINAD in the 1 kHz test, it had a greater chance of a larger gap between the SINAD at 5, 10 kHz and 1 kHz.
2) As above, the gap tended to be smaller if the amp's measured SINAD is relatively high in the 1 kHz test to begin with.
3) Class AB amps measured seemed to, as Amir alluded to, performed less good in the SINAD vs frequency measurements.

To quote Amir, who commented on that measurement of the A1H:

This is a typical class AB amplification showing rise in distortion proportional to frequency above 1 kH

So I would say this to any potential A1H owners, don't be alarmed by that one test, as it isn't a "failure", but rather, "typical". That is, Denon should, or could have done better, than achieving just "typical" results in that test. Then again, as cited many times before, at higher than 1 kHz tests, lots of the resulted harmonics that make up the THD are in the in audible band, and even the resulting IMD would not have audible effects due to the low magnitude and high frequencies, except of course to golden ears.

So did you manage to find class AB amps that Amir measured, that has significant better performance than the Denon, in terms of THD+N vs frequency and output, still love to see a couple?

For reference, let's compare the A1H with the SR8015, that most likely have very similar amp section, if not virtually identical:

The difference in the performance is very interesting, if you look carefully (that, the coloring scheme and the graphs scales/axis are different), the Marantz curve looks better/flatter, but in absolute values, you would have to pick the Denon because while it's SINAD was higher, if you factor in the different scale, aspect rations change between the two graphs, the Denon has higher SINAD all the way to 50m, and the one for the Marantz actually extends to 500m only so we have to just extrapolate the curves to 50m in order to compare.

Then if you compare both from 500m to 200 W, the Denon clearly better than the Marantz, by at least 5 dB between the frequency extremes!

Since their amp sections are so similar, such difference in frequency dependency may be resulted from the power supply and the shielding, so I suspect the Marantz performed better in terms of SINAD vs frequency at lower output level because it has the same or lower noise overall, whereas the Denon likely benefit from having a larger power supply, and the pre out signal is a 5-7 dB cleaner so it should, and does perform better in terms of SINAD vs output, and more than made up for the perhaps a little worse in SINAD vs frequency. If it was possible to compare the two based on THD, not SINAD, the curves would most likely have very similar shape, though the Denon would still have lower THD in terms of absolute values.

Considering Marantz does seem to pay more attention to shielding, and the use of toroidal transformer (not always better but in this case I would bet it is), it would make logical sense that the better noise suppression of the Marantz would explain the different shape. What's you take on this (EE to EE talk lol..)?

 

[celroid]

Just learning about AVRs. What's the intended audience at 15 channels + 4 subwoofer ones (19 channels total)? What would be the advantage of individual subwoofer LFE channels instead of daisy chaning subwoofers which have individual DSP control? Is it for 11.x.4 or 11.x.6 systems instead? I would assume there would have to be a lot of diminishing returns going from something like 9.x.2 or 7.x.4 to 11.x.x unless you have more than 1 row of seats, which already sounds like going into dedicated theatre rooms.

 

[GXAlan]

Ha! I got stuck at 42, but I was there nearly a decade before it became mainstream. I see no reason to go bigger in our room.

But when your TV eventually dies, you won’t be able to find a premium TV below

What would be the advantage of individual subwoofer LFE channels instead of daisy chaning subwoofers which have individual DSP control?

Software like Dirac Live Bass Control and Dirac ART (and Trinnov Waveforming) can independently adjust the DSPs simultaneously.

With individual DSP control, you can optimize each sub in isolation but it’s harder to do it in plural. Tools like MSO (multisub optimizer) are how you can achieve it with multiple subs, but it’s a lot harder.

You can also do directional bass.

I would assume there would have to be a lot of diminishing returns going from something like 9.x.2 or 7.x.4 to 11.x.x unless you have more than 1 row of seats, which already sounds like going into dedicated theatre rooms.

100%. I think beyond 5.1.2 you get diminishing returns, and definitely beyond 7.1.4 (since that’s how many Atmos tracks are hard encoded).

On the previous page I do mention the role of Zone 2/3 which is one application of these beefy AVRs. Older homes which may have whole home audio in the pre-WiFi era may already be wired to have everything to a closet. You can upgrade your HT at the same time you upgrade your accessory rooms, all while using the same software interface.

 

[rvsixer]

What would be the advantage of individual subwoofer LFE channels instead of daisy chaning subwoofers which have individual DSP control?

1) There is only one LFE channel (not multiple LFE channels). Not to be confused with bass managed channels.
2) Daisy chained subs, with individual DSP on each, are limited to sub tweaking only. You really want a device/method that will properly DSP multiple subs (such as AVR with independent sub outs and capable DSP, if not built into the AVR an external device to do same, etc.).

I would assume there would have to be a lot of diminishing returns going from something like 9.x.2 or 7.x.4 to 11.x.x unless you have more than 1 row of seats, which already sounds like going into dedicated theatre rooms.

Not sure if the subject is still subs, or has switched to Atmos top speaker configuration (standard Dolby nomenclature is x.y.x; where x=number of standard plane speakers, y=number of subwoofers, z=number of top speakers):
1) If subject is still subs; in general one (x.1.x) can be good for a single listening position, two (x.2.x) can be good for multiple seats, after four (x.4.x) there are diminishing returns
2) If subject is top speaker config; be aware x.x.2 is still channel-based, you won't get object-based Atmos until x.x.4 or above. Since the whole point of Atmos is placing objects in the room, I'd say x.x.2 is diminished (though I have not personally heard a x.x.2 system myself, all have been x.x.4 or higher).

 

[peng]

Thanks for measuring this beast!

A question for the more experienced engineer-types among us - this amplifier measures very well for an all-in-one AVR, but as a DAC/preamp is still well behind the best measuring Toppings et al.

Is this due to failures or oversights (or simply different priorities) of Denon's engineers, or is it just an unavoidable consequence of packing so many different electronics so close together in a small space like this? Are you just never going to get SINAD better than ~105 when you've got that massive power brick and all those video processing circuits and all the other fun stuff in a top-end AVR sitting next to the DAC chips?

Sorry about my long response.

I would reference/quote Dr. Rich (I have done so a few times on ASR in the past) who wrote a bunch of technical articles on AVRs/AVPs for hometheaterhifi.com.

AVR - Audio Video Receiver - Build Quality: Part I - HomeTheaterHifi.com

AVR - Audio Video Receiver - Build Quality: Part 1

hometheaterhifi.com

A key takeaway: circuit quality in the direct mode (stereo or 7.1) is almost always invariant to AVR prices in the range of $400 to $2,000. As examples, the $250 Yamaha RX-V367 and Marantz AV8801 ($3000) use the same Renesas LSI chip (R2A15220FP). With the LSI analog chip in these products, the sound of the direct mode is relatively constant, although a more robust power supplies, addition a quality output buffer and enhanced DC blocking capacitor quality can make small differences.

This is my favorite quote because he really nailed it in simple non technical term, its purely logic that the strongest point = the weakest link/point.

Now if we apply the logic to the Denon A1H, and the Marantz AV10, they both use the same DAC, and volume control IC, the opas are most likely the same as well, and the Anthem AVM90, is the only one in the <$8000 list price that use ESS's real flagship DAC IC, namely the ES9038 Pro, then why none of them managed to reach 110 dB SINAD, the ES9038Pro in the Anthem, and the ES9018K2M used in the Denon and Marantz unit all have 120 dB SINAD!?

The logical answer, Dr. Rich's point about one possible weak link causing it, that is the volume control IC.
Thanks to Masimo/D+M, they made the following public knowledge, I wish Anthem, NAD, Arcam would do the same!

https://www.nisshinbo-microdevices.co.jp/en/pdf/datasheet/NJU72343_E.pdf

You can clearly see that the volume IC limits the preout SINAD to about 0.0004 to 0.0005%, that is about 108 dB at the most.
We also know the opas used in those units are the

The next potential limit would have been set by the opa used, in this case, they are the NJM8080 and NJM8068:

https://www.nisshinbo-microdevices.co.jp/en/pdf/datasheet/NJM8080_E.pdf

You can see that the best it can do it also about 0.0005% or 106 dB SINAD.

Given such plain facts, I have to give D+M the credit, that their implementation in the pre out/dac circuitry as a whole is superb!!
They did it in the X3800H, C50, C70 too, managed to reach >90 dB SINAD in the pre out/DAC at the optimum voltage output, using the PCM5102A dac chip that tops out at 93 dB.

Lastly, based on the volume control IC, let alone the use of mediocre opas, there are not going to be a single multichannel AV preamp processor or AVR that could achieve the kind of 120 dB or even 115 dB SINAD that many desktop DAC could do. Those DACs typically use much better opas and no volume control chips to degrade the signal.

Inspired by Dr. Rich's articles, I have read a ton of service manuals and found that Arcam once used the CS3318 Cirrus Logic volume control IC, but while it has impressive specs, would still limit pre out SINAD to 112 dB maximum, DR, at 127 dB is excellent.

D+M has upgraded to the NJM72343 volume chip about 7-8 years ago after being called out by Dr. Rich, that allows them to achieve >100 dB SINAD in models that use the better DAC chips. If they want to break through 107 dB, they would have to switch to something as good as the Cirrus Logic CS3308/3318 and use OPAs better than the NJM8068 and 8080 they are currently using. For the likes of the A1H, C30, X6800H, AV10, there are numerous OPAs so the cost will add up quickly in they are to use those found in many desktop DACs in the 115-120 dB SINAD range.

May be the next gen Trinnov, Storm audio kind of AVPs will get us results comparable to desktop DACs. The engineering is there, so it boils down to money/cost.

I doubt such upgrades will happen anytime soon, because manufacturers know going from 105 dB SINAD to 125 dB SINAD makes little sense because most consumers who enjoy their movies/videos don't care enough about going from 100 to 115 dB SINAD, even crazy ones like me, would not want to pay $500 more for the better ICs, as 105 dB SINAD is transparent enough after allowing for the minimum to reach 80 dB at the frequency and output voltage extremes. The only chance for that to happen could be when China, or India finally can made the best chips at much lower price than Taiwan, US, and EU countries can lol..

Note: A quick check on Mouser, the OPA1612 used in the smsl su-9 costs about 10X that of the NJM8080g used in those AVPs/AVRs.

Also worth mentioning that when you do the bottleneck analysis based SINAD, DR, Slew rate specs on the presout/dac signal path, most of the claimed advantage of the HDAMs discrete OPA clearly made no sense because even if it has the specs of a straight wire because they are just one part in the path that has volume control, dac, opas that are all ICs anyway.

CS3308/18 | Cirrus Logic

 

[welwynnick]

Nice. Rich's articles were all very interesting. I did some similar bottleneck analysis here:

https://www.avforums.com/threads/av-processors-with-a-digital-output.2326976/page-21#post-32782316

As you say, volume control is critical and not cheap or simple to do right.

 

[celroid]

1) There is only one LFE channel (not multiple LFE channels). Not to be confused with bass managed channels.
2) Daisy chained subs, with individual DSP on each, are limited to sub tweaking only. You really want a device/method that will properly DSP multiple subs (such as AVR with independent sub outs and capable DSP, if not built into the AVR an external device to do same, etc.).

Does it do automatic DSP for multiple subs without requiring manual user measurements afterwards? Is it just a one time only process or is it dynamic, e.g changing depending on the content, presets...? I imagine such solutions like Dirac Live Bass Control and Dirac ART mentioned by GXAlan above requires paying a subscription. For a one time process I personally would not be interested.

 

[Gwreck]

I bought this unit direct from Denon last year for 30% off. For $4500 I could not turn this bargain down. I realize that it measures just “ok” but better than other avr’s and definitely works better than the Arcam avr-20 it replaced and the NAD t777 prior. I was able to simplify my setup and get rid of the extra amp’s I was using for my 9.3.4 setup. I am surprised about the amplifier headroom this thing has which is pretty impressive compared to cheaper avr’s. Anyway I would not have bought it at $6500 but at $4500 I’m very pleased.

 

[Kal Rubinson]

The other hurdle is that Dante has a fixed sample rate. I am not sure if Milan/AVB/AES67 is any different.

It doesn't have to be Dante (which is burdened by expensive licensing by Audinate), there are several options for an immersive system:

AES67/Ravenna is capable of rate-switching. I am currently running it in 8 channels with a Merging Hapi DAC at up to DSD/DXD & 384KHz.

 

[rvsixer]

Does it do automatic DSP for multiple subs without requiring manual user measurements afterwards? Is it just a one time only process or is it dynamic, e.g changing depending on the content, presets...? I imagine such solutions like Dirac Live Bass Control and Dirac ART mentioned by GXAlan above requires paying a subscription. For a one time process I personally would not be interested.

This is ASR, so of course you would want to check/measure the system afterwards to know if you have met the target response you were after . The Denon in this review can indeed do automatic DSP for multiple subs. Which auto-eq best suits you is up to you. I personally had no luck with Denon's built-in Audyssey (it has some very well known issues), Audyssey MultEQ-X was better, but I got excellent measured/subjective results with A1 EVO. YMMV or course. No experience with Dirac, sorry.

Probably best to get back to the review subject, though applicable its straying a bit OT.

 

[peng]

I bought this unit direct from Denon last year for 30% off. For $4500 I could not turn this bargain down. I realize that it measures just “ok” but better than other avr’s and definitely works better than the Arcam avr-20 it replaced and the NAD t777 prior. I was able to simplify my setup and get rid of the extra amp’s I was using for my 9.3.4 setup. I am surprised about the amplifier headroom this thing has which is pretty impressive compared to cheaper avr’s. Anyway I would not have bought it at $6500 but at $4500 I’m very pleased.

For 30% discount I would too if I am ready to go back to AVR, but also not at full price. Same with the AV10 or AVM90, 30% discount would make them attractive enough for me to trade in my AVM70.

 

[welwynnick]

AES67/Ravenna is capable of rate-switching. I am currently running it in 8 channels with a Merging Hapi DAC at up to DSD/DXD & 384KHz.

Do you use a preamp, Kal, or do you connect your Merging directly to your power amps?

I seem to recall that Merging announced Dante compatibility for the HAPI Mk III?

 

[Old_School_Brad]

I came across some pictures of a disassembled unit, and it looks pretty decent.

Spoiler: Picture 1

Spoiler: Picture 2

Spoiler: Picture 3

Spoiler: Picture 4

Spoiler: Source

Denon's New AVR-A1H Home Theater Receiver, Marantz AV10 and AMP10 Home Theater Separates - A Deep Dive - HomeTheaterHifi.com

At CEDIA 2022 Masimo Consumer Audio (formerly Sound United) took the wraps off of the new Denon AVR-A1H home theater receiver and the Marantz AV10 and AMP10 home theater separates. These new components both support up to 15.4 channels of operation and have been designed from the ground up to...

hometheaterhifi.com

 

[peng]

Nice. Rich's articles were all very interesting. I did some similar bottleneck analysis here:

https://www.avforums.com/threads/av-processors-with-a-digital-output.2326976/page-21#post-32782316

As you say, volume control is critical and not cheap or simple to do right.

That's a very nice chart you did, I downloaded it, thank you very much for that.

There is something that really puzzled me at the time when I saw the low SNR for the X4800H at rated output, but I forgot to alert Amir for perhaps some investigation on that particular measurement.

The obvious question is, how could the AVR-X4800H did so much worse than the X3800H and even the Cinema 70.

With the better layout, shielding, I can understand why the Cinema 40 would do slightly better in DR, but there is no way it could so much better, a full 12 dB better? There is no reason whatsoever for that. As a quick check, the X3800H had the following:

So even the X3800H did 9 dB better, no way!!

In fact, checking further back, the made in Japan AVR-X4800H has the worse/lowest SNR/DNR ever measured on the many recent Denon AVR measurements, including the relatively poorly measured X3500H at full power.

I wonder if @amirm might chime in and offer some possible explanation.

I don't/won't question Amir's measurements, but in this case, I think something might have gone wrong, seriously wrong, resulting in such an outlier results.

 

[Kal Rubinson]

Do you use a preamp, Kal, or do you connect your Merging directly to your power amps?

Yes.

I seem to recall that Merging announced Dante compatibility for the HAPI Mk III?

Yes. Will be interesting to see if they get Dante to respond to changing sample rates.

 

[EWL5]

That's a very nice chart you did, I downloaded it, thank you very much for that.

There is something that really puzzled me at the time when I saw the low SNR for the X4800H at rated output, but I forgot to alert Amir for perhaps some investigation on that particular measurement.

The obvious question is, how could the AVR-X4800H did so much worse than the X3800H and even the Cinema 70.

With the better layout, shielding, I can understand why the Cinema 40 would do slightly better in DR, but there is no way it could so much better, a full 12 dB better? There is no reason whatsoever for that. As a quick check, the X3800H had the following:

So even the X3800H did 9 dB better, no way!!

In fact, checking further back, the made in Japan AVR-X4800H has the worse/lowest SNR/DNR ever measured on the many recent Denon AVR measurements, including the relatively poorly measured X3500H at full power.

I wonder if @amirm might chime in and offer some possible explanation.

I don't/won't question Amir's measurements, but in this case, I think something might have gone wrong, seriously wrong, resulting in such an outlier results.

View attachment 427297



View attachment 427296View attachment 427296

Adds fuel to the conspiracy theorists of why the 4800 got a "recommend" when the 3800 didn't!

 

[beagleman]

I don’t think it can be had. But happy to wait for someone to show us…

post #68, I just said;
"I have said this many times.................

No one would actually PAY for SOTA performance from an AVR with this much stuff packed in it, but for sure, many will bitch about how it is not SOTA.

It is almost like there is some "Anti-AVR" sub-crowd that comes into every AVR thread to announce that its still not SOTA performance and Stereo separates can do better."

Post#70;

"For $6,500, I'd be getting separates."

Intentional thread crap or not? I will let your wisdom decide....