Denon AVR-X4700 AVR Review (Updated)

[peng]

First of all, you cannot judge the quality of the sound coming out (or the quality of correction) looking at how flat the corrected curve is (especially from a single measurement) so that is a non-starter. Small variations not visible in the smoothed corrected measurements may have audible differences. All of my experiments have been in the last two years.

Wow, what are you talking about? I never judged the quality of the sound looking at how flat the corrected curve is, and how do you know I made only a single measurement. So before your blurbs about ....non-starter..,.etc, you should read people post and try to understand what they are saying and if in doubt you can ask for clarification before offering some useful feedback or any comments. My comments about my preference of flat bass curve referred to someone who mentioned most people prefer a curve that is not flat to the low end but one that has more response, like a up slope towards the low end, that's all. Whether I prefer it flat or slope up towards 10, 15 Hz from 120, 150, 200 Hz or what, is just my personal preference, nothing to do with good or bad sound quality.

The smoothed curve is just an approximation nice to look at and for detecting gross anomalies.

Sure, but don't you think others know such basic concepts too? Whenever someone, including me, who mention how smooth their curve is, it is just being informative to mention the smoothing applied, that does not mean the smoothing applied means the response is good, bad, smooth or what not. It seems that you are over thinking what others are saying.

I am here to learn too, but may be you can save you lecturing style without being asked. So condescending, amazing..

But like I said, once you start tweaking the target curves you can make Audyssey and ARC sound equally good if you know what you are doing but beyond the capability of most people buying an AVR.

Sure, tell me about it..

In terms of a brain-dead auto eq approach, I would recommend ARC over DIrac over Audyssey.

Edit: I don't own Denon or Anthem systems at the moment to have partisan views. Just relating my experiences.

If anyone is even hinting partisan views, it is not me, obviously anyone can have their opinion and/or preference. Just like there are Denon fans, Marantz fans, Yamaha fans, and there are people who despite AVRs regardless of brands.

 

[peng]

No worries. Here you go: https://www.audiosciencereview.com/...vr-x4700-avr-review-updated.14493/post-446465

Such a unique measurement, its a pity its not included in the review on page 1.

 

[peng]

(1)Add at this spot or do you mean something elses? "Since the preamp section is directly connected to the input section of the internal power amplifier, and the input section likely presents a very nasty load when the amplifier clips."

Of course, but I think you might have missed quite a few posts on the "bug" topic, the one that Amir/Denon finally discovered, that caused the extra noise over HDMI input, thereby dropping SINAD significantly. That is a separate issue from the degradation at above 1.4 V pre out due to the power amp clipping

 

[amirm]

Such a unique measurement, its a pity its not included in the review on page 1.

Added to the review.

 

[Chromatischism]

I am not disagreeing with the above.

As I said, i was comparing just plain default route (which is what an average AVR buyer would take). I didn't turn or off anything when I first used it.

An AVR usage typically has subs.

What you are saying is it can be tweaked if you can figure out some of the issues around Audyssey like you pointed out. Those are the ones discussed in AVS forums. I don't disagree with it. As I said, you can also tweak the target curve to potentially make it as good as the default ARC results. But I don't expect an average person to go read AVS forums or understand what target curves mean.

So you are actually supporting my thesis that as a brain-dead auto eq process, ARC has an advantage over Audyssey for people who don't need to understand anything about the process just follow instructions.

ARC and Audyssey are in the same class of equalization, just a difference in how they have implemented the UI/UX for the large majority that just want to click through most of the set up and what defaults they have chosen. Dirac is in a separate class of technology whose results cannot be duplicated by ARC or Audyssey regardless of how much you tweak them. But I don't like the default path Dirac provides in their UI/UX either.

I still don't understand why people have claimed that Audyssey suffers because people pay more attention and are more careful with ARC and Dirac but not Audyssey. The default set up process is pretty much the same in all, place the mic in the position indicated and click. If you get into actual tweaking - Arc is easier to tweak being less complicated but not as flexible as Dirac or Audyssey with the app. But all of them require significant knowledge of what one is doing or hangs out in audio forums to do the tweaking right.

Kind of. I'm actually arguing that Audyssey does not need a "target curve" because that is exactly what Dynamic EQ does, but it works at all volumes rather than just one. So, that is the default result for Audyssey.

 

[Bear123]

Hmm, maybe I missed something. Let's check the 96 SINAD conditions again:

• amps disabled (for the tested channels)
• only preamp outputs tested
• only 2 out of 13 preamp outputs active
• DSP/Audissey disabled
• (presumably) Everything but those 2 channels disabled: no zones, network, wifi, bluetooth, airplay, etc.

How's that "used as designed"?
This is an AVR with 11 amps, room-correction as main feature and 123 other features. Does the new Denon manual say "please never start the amps"?! Does the Denon marketing brochure say "please disable DSP and all other features"?!

When used as sold, as designed, as marketed, you get "less than 75dB" SINAD. I.e. 75dB measured from DAC/pre, plus additional amp distortion, plus DSP distortion, plus distortion/interference from 5+ other active channels, plus whatever interference/distortion from network/wireless/zones/etc. Somewhat optimistic, but let's pretend that you still get 70dB SINAD.

And yes, it's tested at 2V. Other AVRs were also tested at 2V and "decapitated"/trashed at 2V. I fully agree with that!

P.S. edit:
The x4700 measures best at 1.1V. Denon assumed that everyone will give them a 50% discount on the standard 2V. Wrong. But fine with me. I just want my 50% on price!
P.S.edit2
@amirm Understandable that you are very happy with Denon's response and want to encourage that. Kudos to that and to both of you.
The question still remains though: how did the x4700 AVR get more than "50% recommended"?

I think you are completely misinterpreting the results. Do you listen way above reference level? Ever? NO. No one does. Most folks don't listen louder than -15 from reference. Maybe -10. SINAD is 96 all the way up to reference level, which is far louder than 99% of users listen. End of story. 2V from preouts is only possible way above reference level. Many speakers can't even handle the power of Denon's internal amps without crapping the bed with distortion.

So if you live in fairy tale land and are worried about SINAD at 6 or 8 or 10 dB ABOVE reference level, then you are clueless and not qualified to comment. This is why ASR should use reference level when indicating volume level that we are seeing these numbers. We are advertising SINAD at volumes that no one will ever use the product....ever. With real world use, i.e. reference level and under i.e. 99.99% of all users at all times, performance is very good. Even having said all that, if you think you can hear .01% distortion from a 2volts pre-out signal while listening 6 or 8 dB ABOVE reference and your speakers are crapping out double digit distortion........


The sky is falling mentality with completely 100% unrealistic conditions that no one will ever see, ever, is ludicrous. Funny thing is, I bet most people couldn't tell the difference between oh no its the end of the world 70 dB SINAD and 96 dB. Excluding test tones or pink noise in an anechoic chamber or whatever ridiculous conditions it is claimed we can detect such miniscule levels of distortion.

Wow, yes ranting but seeing these ridiculous claims over and over and over again is old. If we were talking about cars, we'd be saying the world is ending when you drive your Corvette over 186 mph because NVH increases by 0.5% or something silly.

 

[enormity32]

I just spent an hour catching up on this thread. It at times has gone way off topic. Seems their is more talk about the 3600 and 3700 than the 4700. I'm also surprised nobody has posted a single user review that I could find on the 4700 in the new thread since Denon's reply.

I was one of the posters in the old thread that had a 4700 and hadn't opened it yet. I decided to install it a few days before Denons response. I bought it with the intention to use it as a pre-amp only. It gave me everything I needed in an 7.2.4 pre-amp for less than half the price of units that measured nowhere near as well, and those units didn't have half the features of the Denon. I have it mated to a Sunfire 7 X 400 WPC amp as well as a Sunfire 7 X 200 WPC amp. I am running Canton Ventos towers for the mains, a Ventos center, and Ventos bookshelves for the sides and rears. I haven't made a decision on my Atmos channel in-ceiling speakers yet.

I am currently only running the system in 2.1. I am waiting for my electrician to come and fish wires for the other 8 channels plus an extra for voice of god. So far I am extremely impressed with the unit in pre-amp mode when listening to music through Roon.

There are some quirks that I have found:

1) Unfortunately Denon uses airplay to stream music to the 4700 via Roon. All my music is down-sampled to 44.1/16. This is all airplay can handle. The Heos app does the same. Does anyone know if I can use Ethernet with Roon to stream my music files at their true resolution without re-sampling?

2) Volume range seems perfectly fine when streaming music. I usually listen between 40-50 on the volume. When I watch TV I have to listen around 75 on the volume and it's not as loud as listening to music at 40-50. TV signal comes through the receiver on the newest HD Comcast box. I could find no settings for gain in the Comcast box. This is extremely annoying. Does anyone have any ideas?

3) There are some other issues with Denon's Roon integration that I'll post when I have time. Nothing that can't be fixed with firmware. I will post later. I don't want those issues to add confusion to this post.

Any insight would be greatly appreciated. Love this forum

 

[Baff]

If we were talking about cars, we'd be saying the world is ending when you drive your Corvette over 186 mph because NVH increases by 0.5% or something silly.

That reminds me of an old comic strip, could have been The Far Side.
A kid walks in and says "Hey dad, there is something wrong with your car. It starts to shimmy when it goes over 180."

 

[Baff]

2) Volume range seems perfectly fine when streaming music. I usually listen between 40-50 on the volume. When I watch TV I have to listen around 75 on the volume and it's not as loud as listening to music at 40-50. TV signal comes through the receiver on the newest HD Comcast box. I could find no settings for gain in the Comcast box. This is extremely annoying. Does anyone have any ideas?

To adjust the input source level, press the Setup button to enter the "Setup" menu. Select Inputs>Source level and adjust the volume level.

 

[waynel]

You are clueless. Do you listen way above reference level? Ever? NO. No one does. Most folks don't listen louder than -15 from reference. Maybe -10. SINAD is 96 all the way up to reference level, which is far louder than 99% of users listen. End of story. 2V from preouts is only possible way above reference level. Many speakers can't even handle the power of Denon's internal amps without crapping the bed with distortion.

So if you live in fairy tale land and are worried about SINAD at 6 or 8 or 10 dB ABOVE reference level, then you are clueless and not qualified to comment. This is why ASR should use reference level when indicating volume level that we are seeing these numbers. We are advertising SINAD at volumes that no one will ever use the product....ever. With real world use, i.e. reference level and under i.e. 99.99% of all users at all times, performance is very good. Even having said all that, if you think you can hear .01% distortion from a 2volts pre-out signal while listening 6 or 8 dB ABOVE reference and your speakers are crapping out double digit distortion........


The sky is falling mentality with completely 100% unrealistic conditions that no one will ever see, ever, is ludicrous. Funny thing is, I bet most people couldn't tell the difference between oh no its the end of the world 70 dB SINAD and 96 dB. Excluding test tones or pink noise in an anechoic chamber or whatever ridiculous conditions it is claimed we can detect such miniscule levels of distortion.

Wow, yes ranting but seeing these ridiculous claims over and over and over again is old. If we were talking about cars, we'd be saying the world is ending when you drive your Corvette over 186 mph because NVH increases by 0.5% or something silly.

what is someone has a low gain amp or low sensitivity speakers or both?

 

[amirm]

You are clueless.

Please be professional. We don't speak to each other this way in this forum.

 

[amirm]

1) Unfortunately Denon uses airplay to stream music to the 4700 via Roon. All my music is down-sampled to 44.1/16. This is all airplay can handle. The Heos app does the same. Does anyone know if I can use Ethernet with Roon to stream my music files at their true resolution without re-sampling?

I used Heos app and managed to get 24 bits into it. But no, I don't know of a way to get Roon to stream high res to it. Note however that the performance of this AVR is barely above 16 bits so I would not lose sleep over 16 bit conversion in this case.

 

[amirm]

2) Volume range seems perfectly fine when streaming music. I usually listen between 40-50 on the volume. When I watch TV I have to listen around 75 on the volume and it's not as loud as listening to music at 40-50. TV signal comes through the receiver on the newest HD Comcast box. I could find no settings for gain in the Comcast box. This is extremely annoying. Does anyone have any ideas?

I noticed this too but forgot to note it. Seems like the Heos/streaming subsystem has its own gain.

 

[Vasr]

.. So before your blurbs about ....non-starter..,.etc, you should read people post and try to understand what they are saying and if in doubt you can ask for clarification before offering some useful feedback or any comments.
...
It seems that you are over thinking what others are saying.

I am here to learn too, but may be you can save you lecturing style without being asked. So condescending, amazing..

Sure, tell me about it..

This is such a disappointing approach to a discussion, so this will be my only response to you.

To follow this sub-thread

1. Someone commented that people got better results with ARC or Dirac because they took more care to not make mistakes with those systems. I disagreed with this explaining that the measurements method were all pretty similar between the three to say one would necessarily make more mistakes with Audyssey. Based on my experience, I also mentioned that Audyssey measurements were much more sensitive to mic placement and so it was possible to mess up there but that wasn't user error. I mentioned discussions in AVSF where the common wisdom is to keep the mic about a foot and not much more from MLP to get better results. The better results here has nothing to do with how smooth the curves are but sound output they liked. The reason this became an issue in those forums is because some people didn't like what they heard with their Audyssey setup and they were asked to tighten up the mic positions around MLP. This is not controversial there. Similar suggestion to up the bass a bit after setting up again to do with the the listening experience, nothing to do with any curve characteristics.

2. You posted in response discounting the AVSF suggestions by posting that your were not able to test the validity of those recommendations because your measurements showed that there was not much difference with the smoothing you mentioned when measured at different mic positions. You also wondered that given the measured closeness to the target being close to the accuracy of the mic, how ARC could do any better.

3. I responded by explaining that measurement curves and audibility are not coincidental and that differences in audibility (the point that brought up AVSF in 1 above) cannot be inferred from smoothed/averaged measurements since you were conflating the two. I have no idea how much you know or don't know. I can only go by what you post and the above is the inference. Your use of measurements to discount what the AVSF forum members have found and have a consensus on was a non-starter (i.e., non-sequitur) . 1 referred to a thumb rule designed to make the corrected sound sound better. Your disagreement in 2 referred to inability to validate that with measurements. So, I explained the difference.

Your characterization in response is a form of ad hominem (in the logical sense not the colloquial) based on your take of my intent or form which may or may not be true but is not relevant for the purposes of the discussion hence its use in your post a fallacious form of argumentation.

I would rather not waste my time on such things.

 

[Gedeon]

I noticed this too but forgot to note it. Seems like the Heos/streaming subsystem has its own gain.

In my experience it depends on the content. We all know most modern recordings fall short in DR. If I remember well from HEOS/Tidal I can listen classical at around 65-68 while with other recordings I can't go beyond 62-63.

In any case, I almost never raise the level above 70. Most movies around 68-70. Maybe 73-74 in very, very few cases.

 

[SpoOokY]

Kind of. I'm actually arguing that Audyssey does not need a "target curve" because that is exactly what Dynamic EQ does, but it works at all volumes rather than just one. So, that is the default result for Audyssey.

Well, that is completely wrong. Audyssey Dynamic EQ not only applies a dynamic bass boost depending on volume, it also boosts hights and the volume of the surround speakers resulting in a completely messed up tonal balance. Dynamic EQ is the worst implementation of loudness systems I have ever whitnessed. And no, boosting of hights and surround loudspeaker volume cannot be turned off.

That is exactly why you need a custom house curve even though it is not dynamic. Fortunately the new 2020 Denon generation lets you save two different Audyssey settings with two individual curves - one for regular daytime listening and one for lower volumes for example.

 

[Vasr]

Kind of. I'm actually arguing that Audyssey does not need a "target curve" because that is exactly what Dynamic EQ does, but it works at all volumes rather than just one. So, that is the default result for Audyssey.

That is not correct as I understand it. Dynamic EQ and use of target curves are two orthogonal concepts with different goals but related in implementation. You can have one without the other, in principle. The motivation for Dynamic EQ (as used in D&M products) is to adjust the sound to maintain the "reference level" mix at lower volumes (to compensate for variances in human hearing across frequencies based on volume levels and mostly applying to lower frequencies). The motivation for the target curve is to improve the tonal balance (over and above the flat target that takes care of room modes) that people prefer in practice over a flat curve.

One is not a substitute for the other.

As Audyssey explains it

MultEQ XT– This technology automatically customizes your home theater system to overcome the acoustical problems specific to your room. An optimized audio system provides the most clear and balanced sound possible. It is the basis for the rest of Audyssey’s technologies.

Dynamic EQ – As the volume is turned up or down, Audyssey Dynamic EQ maintains consistent bass response, tonal balance, and surround impression.

Edit: Oops @SpoOokY and I posted the same thing at the same time. Spooky....

 

[Gedeon]

Well, that is completely wrong. Audyssey Dynamic EQ not only applies a dynamic bass boost depending on volume, it also boosts hights and the volume of the surround speakers resulting in a completely messed up tonal balance. Dynamic EQ is the worst implementation of loudness systems I have ever whitnessed. And no, boosting of hights and surround loudspeaker volume cannot be turned off.

That is exactly why you need a custom house curve even though it is not dynamic. Fortunately the new 2020 Denon generation lets you save two different Audyssey settings with two individual curves - one for regular daytime listening and one for lower volumes for example.

DynamicEQ can be customized to select the reference loudness level.

In my experience, by default (0), is too aggressive, since it applies loudness like if you wanted to listen at reference levels.
I personally have it fixed at 10 for movies and at 15 for music. So when listening movies, the closer the level is to -10 (or 70), the less loudness is applied. At -10… no loudness comes into play.

I like the results a lot.

 

[SpoOokY]

DynamicEQ can be customized to select the reference loudness level.

In my experience, by default (0), is too aggressive, since it applies loudness like if you wanted to listen at reference levels.
I personally have it fixed at 10 for movies and at 15 for music. So when listening movies, the closer the level is to -10 (or 70), the less loudness is applied. At -10… no loudness comes into play.

I like the results a lot.

I knew that option and also used it when I still had my old Denon 3300 with XT32. Nonetheless it did never sound good. Video games were a very good example. When playing at night at low volumes the tonal balance was messed up even with DEQ set to 15 dB. Imagine a situation in which you stand in front of a sound source and you turn around so that the source is in your back. Surrounds were always too loud. Only with DEQ set to off the surrounds were as loud as the fronts. If DEQ would only increase bass based on the volume it would be great. But also turning up hights and surround volumes is a pretty bad decision.

 

[Gedeon]

I knew that option and also used it when I still had my old Denon 3300 with XT32. Nonetheless it did never sound good. Video games were a very good example. When playing at night at low volumes the tonal balance was messed up even with DEQ set to 15 dB. Imagine a situation in which you stand in front of a sound source and you turn around so that the source is in your back. Surrounds were always too loud. Only with DEQ set to off the surrounds were as loud as the fronts. If DEQ would only increase bass based on the volume it would be great. But also turning up hights and surround volumes is a pretty bad decision.

Just an example of different experiences. Nothing bad at all. I still do like a lot how it sounds.