• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Denon AVR-X4700 AVR Review (Updated)

OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
But I think you had similar issue with the Arcam, though you did include more speakers/channels and the SINAD went up right?
Yes, that vaguely sounds familiar. :) It is for this reason that when I saw the multitone clipping, I thought I had seen that before.

I have a JBL SDP-55 in for review. Since that shares architecture with Arcam, I should be able to verify this there.
 

peng

Master Contributor
Forum Donor
Joined
May 12, 2019
Messages
5,614
Likes
5,167
I will take all the eggs you want to throw at me for not doing so the last go around. :)

-----------
As always, questions, comments, recommendations, etc. are welcome.

You don't deserve any egg to be thrown at, because eggs are reserved for those who covered up their errors and wouldn't give credits to those who helped found the errors. So, I, for one, do appreciate your honesty, integrity and transparency.

So I suppose it is reasonable to expect the X3700H will be similarly good/bad right, and the SR6014 too except the HDAM thing may cause a couple dB in SINAD but may do a little better at higher output and/or tougher load. Will you be testing the 6014 first?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
So I suppose it is reasonable to expect the X3700H will be similarly good/bad right, and the SR6014 too except the HDAM thing may cause a couple dB in SINAD but may do a little better at higher output and/or tougher load. Will you be testing the 6014 first?
Most likely yes as the 3700 is not here yet.
 

North_Sky

Major Contributor
Joined
Oct 31, 2019
Messages
2,741
Likes
1,553
Location
Kha Nada
Oh this is cool, and it also confirmed my excellent experience with Denon in the past.
That was Denon Canada.

Now the X6700H, bring it on.
 

peng

Master Contributor
Forum Donor
Joined
May 12, 2019
Messages
5,614
Likes
5,167
@amirm I sorta agree here... most consumer level amps will require 1.6V. If I recall the X3600 was solid up to 2V and even performed pretty well to 4V, is that correct?

Oh no, you misunderstood.. The X3600H performed just as bad/or good as Amir's comment was "At first I tested it in normal mode and it could only go up to 1.5 volt before the amp clipped and sunk the DAC performance with it. "

Between 1.4 V and 1.5 V I would say ignore the difference as it could just be normal variations between units. Remember, every part's specs have "tolerance". The important point, is if you need more than 1.5 V or so, disconnect the front channel amps, or engage preamp mode if you don't need to use the internal amps.
 

peng

Master Contributor
Forum Donor
Joined
May 12, 2019
Messages
5,614
Likes
5,167
Oh this is cool, and it also confirmed my excellent experience with Denon in the past.
That was Denon Canada.

Now the X6700H, bring it on.

It also continue to explain why I have been so happy with my X4400H that replaced my AV8801. No loss of perceived sound quality going from an ex flagship AVP to a mid range AVR!!
 

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
I don't think they corrected anything, the issue was on this side.
Correct, I suppose I meant the fact such a big time mfr. bothered to follow up.
 

tparm

Senior Member
Forum Donor
Joined
Nov 5, 2019
Messages
352
Likes
223
Oh no, you misunderstood.. The X3600H performed just as bad/or good as Amir's comment was "At first I tested it in normal mode and it could only go up to 1.5 volt before the amp clipped and sunk the DAC performance with it. "

Between 1.4 V and 1.5 V I would say ignore the difference as it could just be normal variations between units. Remember, every part's specs have "tolerance". The important point, is if you need more than 1.5 V or so, disconnect the front channel amps, or engage preamp mode if you don't need to use the internal amps.
X3600 going back this afternoon and I'll decide between the X3700 & X4700 tonight. Either will be run in full pre-amp mode will no use of internal amplification. But I thought that's how it was tested, from the pre-out with amplifiers disconnected and resulted in best performance at only 1.1V. I guess I did misread as that would be a problem.... Now if Monolith would get their 5CH 5X back in stock! Thanks @amirm and @peng as you've been gracious with your time.
 

miike888

Member
Joined
Dec 8, 2019
Messages
34
Likes
17
Yes, that vaguely sounds familiar. :) It is for this reason that when I saw the multitone clipping, I thought I had seen that before.

I have a JBL SDP-55 in for review. Since that shares architecture with Arcam, I should be able to verify this there.

But hey Amir slow down now!;)

we are waiting for the Marantz 6014 review!:)
 

DrewMcG

Member
Forum Donor
Joined
Jul 2, 2020
Messages
52
Likes
33
I believe you mentioned, Amir, that you tested using one of the 4k (HDMI 2.0b) inputs. Is it possible that performance would vary on the 8k HDMI 2.1 input?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
I believe you mentioned, Amir, that you tested using one of the 4k (HDMI 2.0b) inputs. Is it possible that performance would vary on the 8k HDMI 2.1 input?
I would normally say "no" but under the circumstances, I better test. :)
 

Vasr

Major Contributor
Joined
Jun 27, 2020
Messages
1,409
Likes
1,922
I am glad it worked out for a win-win for all. But, there is still some explaining to do to figure out the root cause of the problem and whether this Denon unit is in the clear in all aspects... I could have misunderstood the problem fixed but it is useful to be clear about this.

The below isn't exactly right or unambiguous.

The output path form my PC however was 8 channels (determined by Intel GPU HDMI implementation in my Intel CPU). I had turned off all channels beside left and right assuming that would simulate simple 2 channel playback. Well, turned out this was the problem!

First, the number of channels isn't determined by the GPU HDMI implementation which as far as I understand, just provides the transmitted data to the audio and video drivers above it. All HDMI ports can transmit 8 channels at the HDMI protocol level.

Windows driver for HDMI (generic or chip vendor supplied) look at the EDID information coming over HDMI data to determine the capability of the downstream unit. For example, if you connect a typical TV directly, most TVs will only send Stereo capabilities in the EDID and so the Audio Device in Windows corresponding to the HDMI port will only show up to Stereo not any more. It will look something like this in the device properties (even when the HDMI is capable of sending multiple channels to something else connected to it). You cannot output more than 2 channels to such a device. For example, it may look something like this.

TV_HDMIJacks.PNG


A typical AVR would be sending ts capabilities over EDID (overriding TV audio capabilities if signal passing through) which I believe is 8 channels max (uncompressed, unencoded PCM - more channels are accommodated via encoding/bitstreaming). The HDMI drivers then recognize it as a 8 channel device and make that capability available in the HDMI properties. Most AVRs will send the max channels it is capable of rather than number of speakers currently attached since you don't want the source to be reconfigured every time, you change a zone or upgrade speakers.

If you are using a ASIO or Wasapi driver to deal directly with the device, that is what the source will see as device capabilities. If you are going through Windows Audio engine, then there is additional configuration you can do in Windows Device Manager to select the number of speakers actually attached, whether they are large/small etc and can do its own up/down mixing which will affect what is going out in each channel. But this only comes into play when you are using the internal Windows mixer in shared mode.

Here is the unexplained part:

When you say "I had turned off all channels beside left and right", it is not clear what this is referring to. In the source on the PC for the sounds being generated or in the AVR for the number of speakers attached? The distinction is important.

From what I understand, there is a problem when the unit is receiving more channels than the AVR is configured for in number of speakers attached and this should not happen. This is not an erroneous input. AVRs can receive 7.1 input when configured for 2, 2.1 or 5.1 speakers.

There are two things AVRs can do and ideally based on a user configurable setting - Either just drop the other channels or down-mix if less speakers than channels received. What is the default configuration on this unit? Neither of these should generate noise/clipping/distortion if done correctly.

First of all, your PC source should be sending signals only in the channels you have selected for use in the source (not the AVR). It is not clear to me what your source is actually doing and whether it is sending the same signal via all channels when it looks at a 8 channel audio device. This is something that should be looked at and standardized across all tests over HDMI.

If source PC is sending nothing over the other 6 channels, down-mixing to stereo or dropping the channels in the AVR should have no effect on the measurement. (It would also be useful to ensure that there is no noise being sent over in those channels in this kind of configuration). So, if this was the case here, then there is a problem with the AVR.

If it is sending the same signal over all 8 channels, there should be no effect on the measurement IF down-mixing is disabled in the AVR (either explicitly or by default). If this was the case here, this is still a problem with the AVR.

If there is down-mixing in the AVR and it received signal in all 8 channels, then any down-mixing that creates clipping/distortion/noise is a bug in the down-mixing implementation. There is a lot of philosophical differences in how the down-mixing should happen (i.e., in what proportion those channels should me mixed) but in no situation should this result in clipping or significant distortion. This kind of situation can arise in many, many 5.1 installations where a 7.1 source content is sent over. Or when listening in stereo mode in another zone.

Perhaps Denon didn't expect the boundary condition of 8 channels of 0db signal sent over over all channels to be down-mixed into 2 channels but then that is an engineering failure to be corrected in firmware.

But to be really sure, I would ask Amir to verify and measure what comes out of the source PC in each channel and document and make it standard for all tests of any future HDMI based devices. This is just setting the test standard and being clear about what one is doing for the test.

Based on that, we can determine if the Denon passed or failed the boundary condition if not the typical case.

The above would apply to any AVR not to single out this one.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
[,,]Anyway, I am happy to now recommend the AVR-X4700H. I will take all the eggs you want to throw at me for not doing so the last go around. :)
No need to worry.:) As far as I understand this issue - and I'm not an expert in multichannel audio or HDMI, so take whatever I write now with a bucket of salt - your configuration (send 8 channels with 6 of them having zero signal, to an AVR in 2 channel speaker configuration) is perfectly valid. I use my AVP just like this, and customers use a BD player or a PC and not an AP to feed their AVR.

I would expect that an AVR or AVP mixes those 8 channels down to 2, but I would not expect it to do so with clipping the output channels. I would rather expect the output level being lowered by a factor of 4 (12 dB) max since 6 of 8 channels contain a zero signal (assuming a very simple averaging mixdown algorithm which is certainly wrong). So in my view the AVR-X4700H has a problem with this configuration and it should be fixed in a firmware upgrade (which I think should be possible).

It also shows that an AVR or AVP needs to be measured in many configurations if one wants to be sure to find all hidden flaws. Surely those tens of thousands of tests Denon says to do (which I do not dispute) did not cover this configuration - and I'm very sure that Denon will add this configuration to their future test suite (if not already done).

EDIT: @Vasr just beat me with a much better explanation.:cool:
 

Dj7675

Major Contributor
Forum Donor
Joined
Jan 12, 2019
Messages
2,116
Likes
2,781
Appreciate the hours digging into the initial poor results and creating a new review showing the very good results. I think that establishing inside contacts at Denon (and other brands) could be the silver lining in this. It sounds as if the are very, very good to work with so if any future issues are found, contacts are already established. In my opinion Denon is doing a tremendous job at balancing performance and value in the 3xxx series and and 4xxx series. They have an amazing list of features and standards to get in these AVRs and also to hit a price point and they have succeeded in my opinion. What will be interesting is to see if the 6xxx and 8500 will show better performance, or just more channels and features. Hopefully one day Denon can put out a no holds barred performance based Processor. ASR has shown that many do in fact care a great deal about performance. It would be very interesting to see what their engineering teams could do with a 15 channel predecessor with 4 independent sub outs with a SINAD of 110-115. I wonder if it is even possible in an AV Processsor.
 

SimpleTheater

Addicted to Fun and Learning
Forum Donor
Joined
Jun 6, 2019
Messages
927
Likes
1,789
Location
Woodstock, NY
From what I understand, there is a problem when the unit is receiving more channels than the AVR is configured for in number of speakers attached and this should not happen. This is not an erroneous input. AVRs can receive 7.1 input when configured for 2, 2.1 or 5.1 speakers.
Agreed. While it's nice the Denon 4700 has decent measurements when it is getting the same inputs it is expecting, it's still a flawed design.
 

Nick72

New Member
Forum Donor
Joined
Jun 21, 2020
Messages
3
Likes
7
If your source is 8 channel and you play 5.1, it will cost you about 2 dB. So not that much of a hit.

Assuming I understood correctly,
This seems to indicate that there is indeed an issue with this receiver when it need to combine a HDMI source containing a larger number of audio tracks to output them to a smaller number of speakers that are actually setup.( I.e.. 8 channel blue ray outputted to a 5.1 or lower speaker configuration.. The less number of actual speakers, the worse it might be..).

Assuming my understanding is correct, it sounds like Denon is not blameless here and hope that it tries to address this "bug".

Not sure why Amir needs to be egged for finding this issue.
If anything it will make the product better if it can be addressed.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
If source PC is sending nothing over the other 6 channels, down-mixing to stereo or dropping the channels in the AVR should have no effect on the measurement. (It would also be useful to ensure that there is no noise being sent over in those channels in this kind of configuration). So, if this was the case here, then there is a problem with the AVR.
This is correct. And why this took so long to troubleshoot. It was only when I tested this on another AVR without this issue that I realized there is a problem here.

As I noted in the review, Denon engineering is investigating why this is occuring:
AVRs have channel mapping to deal with configuration differences between input audio streams and playback capability. You may for example use the AVR with just two speakers for living room sound while playing 7.1 channel Blu-ray content. AVR will then use its internal mapping to mix down the high input channel count to stereo. For reasons that are unknown at this point but is being investigated by Denon engineering, when feeding only two channels to AVR but in 8 channel configuration, and the AVR is configured as just Left and Right speakers, the noise level goes up substantially and there is potential for clipping on maximum level signal.

A normal mixer would apply a gain to every channel, perform the sum and handle potential overflow. If higher than input resolution is used for intermediate sum (which one hopes is the case), then dither is involved as is gain management in the AVR. Both or each may be responsible for the clipping and increase in noise floor.

The reason I am not losing sleep over this is because almost everyone using these AVRs uses it with more than a stereo pair of speakers. But yes, I hope a firmware fix is provided once it is confirmed to be the down mixing algorithm.
 
Top Bottom