Denon AVR-X3500H AVR Review

[Haint]

Apologies if this this has already been addressed, but in layman's terms, are any of these measured issues actually audible in real world use? If so, in what kind of scenarios (e.g. max volume, 4ohm/inefficient speakers, etc...) and to what degree (significant, minor, moderate)? Do you expect these issues would be a sufficient tell to point out a difference in a double blind ABX?

 

[doodlebro]

Apologies if this this has already been addressed, but in layman's terms, are any of these measured issues actually audible in real world use? If so, in what kind of scenarios (e.g. max volume, 4ohm/inefficient speakers, etc...) and to what degree (significant, minor, moderate)? Do you expect these issues would be a sufficient tell to point out a difference in a double blind ABX?

No, not really. The dynamic range, or difference between the quietest sounds and the loudest sounds, of most music stays well below 20dB (with some exceptions in classical genres). Movies don't get much higher than that either. So, even the worst measuring DACs still have plenty of DR for any content you can throw at them. You would need the noise floor to be pretty high from a source in order to see negative impacts, especially when recent (last 15 years) audio equipment can easily hit >70dB of dynamic range. I also think many conflate HT and desktop/headphone gear which doesn't help: headphone environments are incredibly low noise, home theaters can be but I'd argue most of us are dealing with less than ideal background noise in our HT setups. Amir does a pretty good job of organizing things to keep 'em separated, but many members think that they are comparable.

Next time you turn your theater on, take note of the level of background noise and how much volume it takes to drown that out so that it isn't audible. It's not much.

I think the most practical thing you can take from these reviews is the peace of mind that you've optimized as much as you can for your system in your environment. These days most of us have a few sources connected directly to our receivers, which then connects to speakers (let's say you're using an external power amp because it helps the scenario by adding complexity). So, if one of those sources is a reel to reel which has a higher noise floor and thus, lower SINAD. Optimizing the signal path after the reel-to-reel is wise (it keeps the noise floor/distortion only as high as the reel-to-reel gives off). As long as every other component has a SINAD higher than the reel-to-reel, ill effects will not be noticed outside of the limitations of that reel-to-reel player. Ultimately it's about optimizing for the most signal and the least noise through the audio-chain but the bounds of reality have to still be considered.

It's worth noting that SINAD, bit-resolution, and noise floor matter much, much more on the recording side than they do on the playback side. I think many of us here fail to remember that when we get caught up in the numbers.

The mistake many people make with interpreting ASR is they see pretty, well-organized graphs and numbers along with Amir's commentary and suddenly they too agree that the engineering behind this product is poor and needs to be looked at. In reality things are much more complicated than this. We cannot possibly interject on how well something was implemented without first understanding the initial design principles and goals. I give Amir a pass to make these statements since he's embedded in the industry and has the chops, but the rest of us are just consumers.

We will do better by understanding where these issues are audible and how to bring manufacturers into the discussion. Imagine if a Denon engineer came into this thread to explain why the ak4458 was implemented the way it was and what their design assumptions were. We'd get a lot further than we have with our admittedly limited perspective on the matter.

 

[Haint]

No, not really. The dynamic range, or difference between the quietest sounds and the loudest sounds, of most music stays well below 20dB (with some exceptions in classical genres). Movies don't get much higher than that either. So, even the worst measuring DACs still have plenty of DR for any content you can throw at them. You would need the noise floor to be pretty high from a source in order to see negative impacts, especially when recent (last 15 years) audio equipment can easily hit >70dB of dynamic range. I also think many conflate HT and desktop/headphone gear which doesn't help: headphone environments are incredibly low noise, home theaters can be but I'd argue most of us are dealing with less than ideal background noise in our HT setups. Amir does a pretty good job of organizing things to keep 'em separated, but many members think that they are comparable.

Next time you turn your theater on, take note of the level of background noise and how much volume it takes to drown that out so that it isn't audible. It's not much.

I think the most practical thing you can take from these reviews is the peace of mind that you've optimized as much as you can for your system in your environment. These days most of us have a few sources connected directly to our receivers, which then connects to speakers (let's say you're using an external power amp because it helps the scenario by adding complexity). So, if one of those sources is a reel to reel which has a higher noise floor and thus, lower SINAD. Optimizing the signal path after the reel-to-reel is wise (it keeps the noise floor/distortion only as high as the reel-to-reel gives off). As long as every other component has a SINAD higher than the reel-to-reel, ill effects will not be noticed outside of the limitations of that reel-to-reel player. Ultimately it's about optimizing for the most signal and the least noise through the audio-chain but the bounds of reality have to still be considered.

It's worth noting that SINAD, bit-resolution, and noise floor matter much, much more on the recording side than they do on the playback side. I think many of us here fail to remember that when we get caught up in the numbers.

The mistake many people make with interpreting ASR is they see pretty, well-organized graphs and numbers along with Amir's commentary and suddenly they too agree that the engineering behind this product is poor and needs to be looked at. In reality things are much more complicated than this. We cannot possibly interject on how well something was implemented without first understanding the initial design principles and goals. I give Amir a pass to make these statements since he's embedded in the industry and has the chops, but the rest of us are just consumers.

We will do better by understanding where these issues are audible and how to bring manufacturers into the discussion. Imagine if a Denon engineer came into this thread to explain why the ak4458 was implemented the way it was and what their design assumptions were. We'd get a lot further than we have with our admittedly limited perspective on the matter.

Thanks for the detailed response. My first impression was to wholesale write Denon's off as trash based on my ignorant interpretation of some of the graphs and commentary. I understand XT32 + Audyssey app is a fairly capable system, so it would be a shame if anyone else as dumb as me decided to buy less capable separates/AVR's just cause they measured better.

 

[Tempest]

No, not really. The dynamic range, or difference between the quietest sounds and the loudest sounds, of most music stays well below 20dB (with some exceptions in classical genres). Movies don't get much higher than that either. So, even the worst measuring DACs still have plenty of DR for any content you can throw at them. You would need the noise floor to be pretty high from a source in order to see negative impacts, especially when recent (last 15 years) audio equipment can easily hit >70dB of dynamic range. I also think many conflate HT and desktop/headphone gear which doesn't help: headphone environments are incredibly low noise, home theaters can be but I'd argue most of us are dealing with less than ideal background noise in our HT setups. Amir does a pretty good job of organizing things to keep 'em separated, but many members think that they are comparable.

Next time you turn your theater on, take note of the level of background noise and how much volume it takes to drown that out so that it isn't audible. It's not much.

I think the most practical thing you can take from these reviews is the peace of mind that you've optimized as much as you can for your system in your environment. These days most of us have a few sources connected directly to our receivers, which then connects to speakers (let's say you're using an external power amp because it helps the scenario by adding complexity). So, if one of those sources is a reel to reel which has a higher noise floor and thus, lower SINAD. Optimizing the signal path after the reel-to-reel is wise (it keeps the noise floor/distortion only as high as the reel-to-reel gives off). As long as every other component has a SINAD higher than the reel-to-reel, ill effects will not be noticed outside of the limitations of that reel-to-reel player. Ultimately it's about optimizing for the most signal and the least noise through the audio-chain but the bounds of reality have to still be considered.

It's worth noting that SINAD, bit-resolution, and noise floor matter much, much more on the recording side than they do on the playback side. I think many of us here fail to remember that when we get caught up in the numbers.

The mistake many people make with interpreting ASR is they see pretty, well-organized graphs and numbers along with Amir's commentary and suddenly they too agree that the engineering behind this product is poor and needs to be looked at. In reality things are much more complicated than this. We cannot possibly interject on how well something was implemented without first understanding the initial design principles and goals. I give Amir a pass to make these statements since he's embedded in the industry and has the chops, but the rest of us are just consumers.

We will do better by understanding where these issues are audible and how to bring manufacturers into the discussion. Imagine if a Denon engineer came into this thread to explain why the ak4458 was implemented the way it was and what their design assumptions were. We'd get a lot further than we have with our admittedly limited perspective on the matter.

I have the 3500h connected to an Outlaw 755. It has a 1.43v impedance. The max volume I can listen to comfortably in my small room is -10. When I first read the measurements I was concerned. But after reading the feedback I am feeling somewhat better as I am not coming even close to pushing the Denon to the testing levels. It would be interesting to see tests at reasonable playback levels to see how these measure for the average listener. Similar to car reviews as I don’t drive 140 mph and rarely taken corners at 80 mph. That being said Amir’s Reviews do serve a purpose to show what these units should be doing when pushed to the limit. Just not sure how many people actually do that.

 

[Bear123]

Some folks read these reviews and think the units are garbage. I'd love to see better measured performance and hope these reviews lead to that(apparently they have for some products), but what I see here is distortion well below audible thresholds using both the internal amps and pre-outs(85 dB, 95 dB), especially if kept below 1.5 volts, with more power and better measured performance than plenty of much higher priced gear such as NAD.

Also get XT32 and subeqHT with which EQ can be limited to below Schroeder. All for under $600. Seems like a hell of a bargain honestly. I have the 3300 and taking a realistic look at the measurements, it seems like "upgrading" to a much higher priced AVR or pre-pro to get better measured performance will only yield better measurements, not audibly better sound. I'd still be interested to possibly "upgrade" to an external amp for LCR, but the real world benefits of this for my particular use are infinitesimally small since I don't listen to movies near reference level where multi channel power output of the AVR would become a concern.

Again I hope these detailed measurements do result in better engineering and measured performance, but I also think we are a long ways off from "the sky is falling", and I think the measurements can be really misleading to folks who think using an external DAC to get a cleaner signal would somehow improve sound quality. In reality, eq below Schroeder along with properly integrated dual subs will provide drastically better sound quality than using an external DAC in 2.0.

 

[jsy]

It's worth noting that SINAD, bit-resolution, and noise floor matter much, much more on the recording side than they do on the playback side. I think many of us here fail to remember that when we get caught up in the numbers.

Are there reasons for this besides that the recording is "upstream" of the playback? I get that 24 bits of playback can't improve an 8 bit recording - garbage in, garbage out. Any other factors?

 

[doodlebro]

Are there reasons for this besides that the recording is "upstream" of the playback? I get that 24 bits of playback can't improve an 8 bit recording - garbage in, garbage out. Any other factors?

Since much of audio engineering and sound creation is plugging things into other things to create a chain, squeezing as much headroom as possible at the start leaves more headroom as the song gets mixed and mastered. So it's still garbage in garbage out but the chain is what makes it so.

Certain effects like compressors do poorly with higher noise floors as well. A compressor is just a chain of effects at the end of the day.

 

[jsy]

Since much of audio engineering and sound creation is plugging things into other things to create a chain, squeezing as much headroom as possible at the start leaves more headroom as the song gets mixed and mastered. So it's still garbage in garbage out but the chain is what makes it so.

Certain effects like compressors do poorly with higher noise floors as well. A compressor is just a chain of effects at the end of the day.

Word, I was with you then. This makes sense to me.

As to your mythical AVR engineer facing a panel of angry, pitchfork-wielding forumites, in my head this unfortunate person just says "our thresholds were set where we thought our buyers wouldn't be able to tell the difference. Once we hit them, we spent all the rest of our time on integration/size/heat/manufacturability/shininess/whatever."

 

[mpbuggins]

Thank you for testing this. However my conjecture was that “pure direct” mode improves the amp’s, not the DAC’s performance on this machine (much better sound in pure direct through the outboard ChordQutest compared to the onboard DAC. No difference heard in “auto” mode). If my experience is tracking the objective reality, the only way it can be happening is if amplifier’s 84 SINAD improves substantially in the pure direct mode.

So if I was in pure direct mode on my “cheesy” AVR-X3500H and using my pre-amp outs to a Emotiva BasX A300 to power my front channel Klipsch RP280s I would be ok?

 

[AnderS52]

Its a mystery why not at least one of these companies, on the bottom of the charts, study the measurements and see if there is anything that can be done to get better results. They could use this in marketing, i mean if they say:
"we listened to our customers and decided it was time to improve blah blah blah" - they would get my money.
I am thinking about the surround receivers, the NAD t758, Pioneer and such.
Because they can all do it, they have models that both measure good, and sound good.
I am aware it is a question of price, but i believe you can do very much with good design choices. If one of these DAC chips needs clean power to perform good, then maybe a small revision on the board layout would be welcome. Why not call it " t758 Mark II" ?
How hard can it be ? If Behringer or Crown can stick some transistors, some caps and a transformer together for a very low price,
and it measures nice, why cant Yamaha ?
They dont have to re-invent the wheel, just use common sense.

The second mystery is: Why does t758v3 get such good reviews, if it sounds like shit ? It gets "recommended" from SoundandVision, HometheaterHifi, AvNirvana (reviewer heard distortion, but said it was in the recording !), HomeCinemaMagazine, and EISA.
Are all these reviews wrong ? Common sense would have me conclude: that they are not.
In fact, the amp could be regarded as one of the better options, except for an endless number of problems with firmware,
ps4 HDMI handshake and Dirac troubles in the owner's thread here:
https://www.avsforum.com/forum/90-receivers-amps-processors/2895049-nad-758-v3.html

But still, the effect of having someone post horrific charts can not be underestimated,
as normally the first thing a person does when buying anything, is looking it up on the net. Or am i the only one ?

The 3rd mystery is this: I see a lot of talks about ECo mode and standby current draw.
Please: Turn your power off when not using the equipment. You can get these remote operated power outlets, you can even connect them to Smart Houses integration, and turn them off on a schedule, or together with TV or light switch.
As soon as there is power, the clock is ticking, and every component has a life expectancy. And there is risk of fire.
The heat inside for instance a Subwoofer, (yes Klipsch sw115r-i am looking at you), can be enough to kill it in short time.

 

[markus]

The second mystery is: Why does t758v3 get such good reviews, if it sounds like shit ?

Who said "it sounds like shit"??? Is that what people take away from reading post #1? Looks like @amirm needs to adjust his commentary.

Regarding "good reviews": ever seen a device NOT getting a good "review" when the manufacturer sends out a press release at product launch? That's exactly what those "reviews" are worth. Parroting by the press followed by even more parroting from "expert" reviewers. Welcome to the wonderful world of marketing.

The 3rd mystery is this: I see a lot of talks about ECo mode and standby current draw.
Please: Turn your power off when not using the equipment.

People want instant-on. The 758 draws only 0.5W if "CEC/LAN/BluOS IN STANDBY" is disabled.

 

[617]

I have an x1500h and it does a good job of making my TV loud. I honestly don't hear much in video sound channels which really demands truly hifi reproduction, just ambiance, dialog and ****** music. It would be nice if my hifi and tv could live together but it seems like that's not happening for less than $10K.

 

[North_Sky]

I have an x1500h and it does a good job of making my TV loud. I honestly don't hear much in video sound channels which really demands truly hifi reproduction, just ambiance, dialog and ****** music. It would be nice if my hifi and tv could live together but it seems like that's not happening for less than $10K.

In real life the sound channels we don't hear much either.
But with fake movies...Avengers, Batman, Superman, Star Wars, ... the surrounds are actively fake and loud.

You can watch some films (Roma on Netflix as an example) with one of the best and most realistic sound mix done and you'll get the full impact of it with an inexpensive multichannel setup.

That's my opinion for a realistic and tasteful and intelligent world of entertainment today in this world of Parasite and 1917 we live in.

 

[AnderS52]

i generally conclude that if something measures poorly, it must perform poor as well. Thats just me.
What else is the point ? And these are not Tube Amps

For the standby mode, i have no comment. It is in general, just a bad idea.
Plus: you can still get instant on, if you use remote operated wall sockets.
At the very least, it will reduce the number of boxes connected to mains power.

https://www.fireprotectiononline.co.uk/info/why-shouldn-t-leave-electrical-plugged-in/

 

[markus]

i generally conclude that if something measures poorly, it must perform poor as well. Thats just me.
What else is the point ? And these are not Tube Amps

If a piece of wood is 0.1 inches off, it is not a POS per se. It's simply 0.1 inches off. Depends on the context if that's good or bad "performance".

For the standby mode, i have no comment. It is in general, just a bad idea.
Plus: you can still get instant on, if you use remote operated wall sockets.
At the very least, it will reduce the number of boxes connected to mains power.

https://www.fireprotectiononline.co.uk/info/why-shouldn-t-leave-electrical-plugged-in/

Yeah, life is dangerous. By the way, plugging devices in and out is what kills them (besides continuous heat, corrosion and mechanical stress).

 

[North_Sky]

It's the same with people; plugging and unplugging them stresses them ...

 

[AnderS52]

So: Regarding poor measurements NOT affecting sound quality nor decision to recommend or not recommend any product:
1: You never had this discussion before.
2: You have had it before, but have not reached any conclusion.
3: You have discussed it, and concluded that it can in general: be disregarded.
4: You prefer to put your trust in God and Snake oil
5: Anyone who raises the subject, is expelled.
6: The purpose of measurements is not to conclude on quality of product.

Regarding the unplugging/plugging subject, by all means, keep yours on 24/7. Happy life

 

[North_Sky]

Keep it pluggin' ... good audio philosophy.

 

[doodlebro]

i generally conclude that if something measures poorly, it must perform poor as well.

I hope you realize how limited that perspective is. Not only does it leave no room for subjectivity (which is still important) it doesn't really make sense of where these measurements cross the audible range.

 

[capt.s]

This is a review and detailed measurements of the Denon AVR-X3500H Home Theater Audio Video Receiver (AVR). It is kindly loaned to me by a member. The AVR-X3500H is a 2018 model but it is still available from Amazon for US $599 despite having a retail price of US $999. If it performs, this is a very attractive price.

The front panel of the unit is the same as any other AVR:

View attachment 39924​

The volume control though has a much nicer feel than most other AVRs.

If there is a setup button on the front, I must have missed it as I could not find a way to program the unit without the remote control.

The back panel is as expected:

View attachment 39925​

The speaker connectors were beefy enough for my heavy wires I use which I appreciated.

For testing, I reset the unit which put me in a wizard which was useful but then got stuck without a cancel button on room EQ. Somehow I got out of it but was not easy.

The on-screen graphs were not that attractive but did the job.

The amplifier modules use the same flimsy heatsinks many other AVRs use. Here, they made a mistake of running two wide ribbon cables over the left and right channels which likely block convection cooling fair bit. In use, during my normal testing those two challens felt pretty warm, eventually leading to amp limiting power output to some 35 watts or so. I then ran my regulated power test which caused the unit to shut down and go internal protection. I put a fan on it and it would again shut down. Clearly these AVRs are not designed for continuous duty operation.

I had a heck of a time getting my PC's Nvidia HDMI to route through the AVR-X3500H. It would only provide 1080p resolution at 4:2:0 causing color bleed due to interpolation there. And a fuzzy screen on my 4K display. Messing with menus and settings in Windows did not help. Worse yet, ASIO4ALL interface would not see the Nvidia HDMI audio channels at all. If you are planning to use a PC to drive this AVR, better be sure you know what you are getting yourself into.

DAC Audio Measurements
For this portion of the testing, I did not connect any speakers to the output terminals as to avoid stressing them. I then ran my HDMI tests by using Roon using WASAPI interface to play a couple of test files starting with our 1 kHz tone:
View attachment 39927

The default output at full volume was quite a bit higher and clearly clipping. So I lowered that to 2 volt but unfortunately performance was still lousy as you see. SINAD is dominated by harmonic distortion and not noise. Clearly the output is beyond its comfort factor (shown later). For now, the ranking is poor:

View attachment 39928

And among AVRs:
View attachment 39929

Here is the jitter test showing copious amount of jitter components:
View attachment 39930

As noted, I needed to use another interface in order to be able to run the rest of my tests so I opted to use Coax input for the following. First, let's compare the dashboard to see how it performs relative to HDMI:

View attachment 39933

Overall score remains the same but disappointing to see additional sidebands around our main tone of 1 kHz. Jitter test shows that in detail:

View attachment 39934

The sidebands beyond that are also elevated.

Intermodulation distortion relative to level shows that this DAC is designed for low level of output:
View attachment 39935

Even we back off from clipping though, performance still falls short.

Fortunately noise rating is good:

View attachment 39936

Filter response is typical:
View attachment 39937

Linearity was surprisingly good:
View attachment 39938

I could not run multitone test because it requires 192 kHz support and the coax input doesn't go that high.

AVR Amplifier Audio Measurements
I tested the amp using both digital (coax) and analog input. Let's start with analog:

View attachment 39939

Nice to see above average performance here relative to other AVRs:
View attachment 39940

It is also above average for amplifiers in general but of course well shy of state of the art:
View attachment 39942

Frequency response extended quiteh high indicating it is not digitized:
View attachment 39943

This is without running Room EQ.

Signal to noise ratio was not as good as the DAC section:

View attachment 39944

Crosstalk was especially poor:
View attachment 39945

This is what I see in $30 amplifiers. The concern here is not that you don't have enough channel separation but that what bleeds into the other channel many be worse than its own distortions.

Most important is power versus noise and distortion. Let's start there with 8 ohm and analog input:

View attachment 39946

The AR-X3500H massively outperforms the two NAD products in noise and distortion while matching the power rating. And on that front, the amp is exceeding its rated power.

For variety and to better simulate use case, I tested the Coax input to drive the amp. Here is the dashboard again:
View attachment 39948

Sadly noise floor has gone up causing SINAD to drop a couple of dBs. One things digital input would be preferred to an analog input from noise front but here, we take a step back. There is a reduction of spurious tones though higher up in frequencies so it does some good there.

Anyway, running with that to measure 4 ohm performance gets us this:
View attachment 39949

Again, the NAD T777 V3 is left in the dust with respect to noise and distortion.

Note that the amplifier is not rated for 4 ohm but works anyway.

Here is the regulated output test that caused the amplifier to shut down (sweep is from 20 khz down to 20 Hz):
View attachment 39954

I later ran the test from 20 to 20 kHz and it produced almost the same power starting at 20 Hz. But oddly again it shut down at the same 60 Hz point. The mains noise may be confusing the analyzer causing it to stay there longer leading to shut down. Not wanting the amp to be damaged, I did not try to run the test any further.

Finally, here is the FFT spectrum of a 1 kHz tone:

View attachment 39951

Lack of design hygiene shows spurious tones but they are very low in level and at any rate, much less than class D implementations.

Conclusions
Let's get the worst out of the way first: the Denon AVR-X3500H DAC implementation is quite poor. Clearly no care was included to make it perform well. It is much worse than the amplifier at the limit which is embarrassing. I was about to give it the worst score due to that (headless panther) but the amplifier pulled through. I appreciated the analog inputs not going through mysterious transformations. And low noise and distortion which bettered most AVRs I have tested.

Thermal performance has some issues but not as bad as some other budget AVRs such as Pioneer.

Overall, you have a lousy DAC but a good amplifier with Room EQ at attractive price in Denon AVR-X3500H. As such, it makes a decent option if you have such needs. If it goes on sale, it may be even a better bet.

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

I need to hire someone to write better jokes for me. If you have been suffering through my bad jokes then I hope I can count on generous donations from you using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Amir - You seem to be complementary of the analog input. I'm not sure if you measured this without the amplifiers in line but what is the performance of the analog input straight through to the pre-outs in pure direct mode? i.e. Does it make a reasonable and convenient volume control for a desktop DAC such as The Topping E30 or the SMSL Sanskrit MkII you just reviewed without being too destructive to the SINAD?