This is a review and detailed measurements of the Anthem MRX 520 Home Theater Audio/Video Receiver (AVR). The MRX 520 costs US $1,399 so it is a premium product. The sample I have was kindly sent to me by a member.
The Anthem MRX 520 leaves a decent impression as far as look and feel:
The back panel is the usual assortment, sans analog multichannel input that is rather common in higher end AVRs:
The speaker terminals are a step above the other AVRs I have tested in this price category which I appreciated given my larger banana jacks i use.
You may notice that the cover for one of the Toslink inputs is stuck open. I tried to insert my Toslink into it and it would not seat. So I moved over to the next one and it fit fine. I have to look to see what is going on with its hinged lid.
The Anthem AVRs come with a very fancy room equalization kit which consists of proper measurement mic, stand, etc. This is in sharp contrast to the $2 "puck" style microphones that comes with other AVRs. While I did not test the Anthem Room Correction (ARC) in this unit, another product from the same family performed very well. ARC and Dirac in my opinion are a clear step above homebrew products from AVR manufactures and Audyssey although I have not tested the latest versions of that.
My testing goes to the core of the unit as far as performance. As such, I try to disable all that I can as far as processing. I set the unit to 2 channel, no sub, no surround, etc. On MRX 520, I failed in getting pass-through performance with frequency response still deviating substantially from flat. I assumed that was due to previous configuration using ARC. Unless my eyesight is failing me, I could not find a way to defeat that. So I performed a factory reset and that cleared things up. My apology to its owner which has to now re-run the calibration.
The MRX 520 runs fairly warm in use but not concerningly so. The heatsink are the older, more solid and beefy ones. But as is typical, pretty small for a unit with so many channels.
Recent updates to my PC seem to have remedied the problem I had with using HDMI from my PC to drive AVRs. So I performed my testing using that. That makes the results much more representative of how people use these units.
I was pleased to see Anthem as the only AVR maker that I have tested which has some spec other than vague amplifier ratings:
View attachment 33192
Kudos to them to publish this. Without it, how can you ever prove to a manufacturer your unit is not up to spec? Or whether they have any manufacturing or design target?
HDMI DAC Performance
As I have been doing with AVR tests, I like to first see how the front-end of the unit performs as far as input to analog output using pre-outs. Here, I was pleasantly surprised to see the Anthem output some 3.4 volts from its RCA terminals. That is quite good seeing how we like to see a nominal output of 2 volts. So I dialed down the volume control until it read 2 volts and here are the results:
View attachment 33189
Channel 2 performance is good but channel one was dancing around both in FFT and in SINAD (signal over noise and distortion). You can see a snapshot of it in red in FFT. That noise floor would climb up and then jump back down and repeat. This reminded me of a first generation Anthem AVR I tested where every second or two it would do the same thing. Clearly internal system activity is bleeding into one channel here.
Highest distortion products are at -105 dB. Since SINAD is sum of those plus noise and it is arriving at 95 dB or so, it tells us that SINAD is noise limited, not distortion.
The SINAD score as is, lands in the third tier of all DACs tested. In the narrower context of all AVRs I have tested, it is in the upper range of the two buckets:
View attachment 33190
If you go back to the FFT, you see a defect in its output: there is a spike between our main tone at 1 kHz and second harmonic at 2 kHz. I later measured it at 1.6 kHz. Not good since our hearing gets much more sensitive starting around 2 to 3 kHz.
From here on, unless I note it, I am testing at -5 dB since that gives us 2 volt nominal output. And that is how Anthem rates the unit.
Here is how dynamic range shakes out for example at that output level:
View attachment 33191
Channel 1 as noted was dancing back and forth.
My testing is without any kind of weighting so our shortfall relative to -110 dB can partially be explained. The rest may be test protocol differences.
Overall, we are clearing about 17 bits of dynamic range. So our 16 bit content is good, but not our 20 bit cinema sound.
I went to run frequency response test by setting the sample rate to 192 kHz and it was a head scratcher at first:
View attachment 33193
Instead of 96 kHz, we don't seem to have much above 22 kHz or so! Where did the rest of our bandwidth go? Well, we can answer that by stepping from 44.1 kHz sampling all the way up to 192 kHz with white noise (i.e. very wide spectrum) and see what the unit provides:
View attachment 33194
Yup. All sample rates above 44.1 and 48 kHz produce the same bandwidth as these two values! All sample rates above 48 kHz are divided by 2 or 4 to arrive at either 44.1 or 48 kHz. If this were a DAC, we would jump out the window in protest. But seems to be standard affair to secretly resample inputs in these AVRs.
Note also that there is a level error in resampler. The higher it samples down, the lower the level gets! This means you lose your level calibration as I doubt every much that the Room correction system is aware of this level shift. You get the highest level reduction at 176 and 192 kHz.
Best to put aside any aspirations you have for high resolution audio. For movies or music.
Intermodulation distortion versus level presented a puzzle in the larger picture:
View attachment 33195
Just about every AVR I have tested seems to be producing similar levels! Way higher noise than good desktop DACs like the Topping DX3 Pro in dashed line. But every similar to each other. It is as if they all got together in a secret room and decide what was good enough for all of us. "Let's give them 15 dB higher noise -- they can't hear it anyway. You say nothing, and neither will I."
Rant aside, I ran the test at -5 and 0 dB volume control and there is some difference as you would expect with the higher level getting a bit more distorted.
Linearity picture is ugly compared to what we are used to even in budget (read: $99) USB DACs:
View attachment 33196
The one channel really goes crazy as levels go down, producing nothing but random noise. The other one is more well behaved but you don't listen to one channel alone, do you? Here, you are good to 95 dB or 16 bits or so before the fan belt breaks on your engine.
Here is the jitter and noise performance:
View attachment 33197
I also show the Toslink for reference. Compared to that, the HDMI (in red) has wider "skirt" indicating random but low frequency jitter.
Remember I mentioned that unwanted spike below 2 kHz? It is here again. Something is whaling inside at that frequency and bleeding into sensitive analog/DAC output.
We can zoom in on that and see it more clearly:
View attachment 33198
As a way of reference, this is what a state-of-the-art headphone amplifier does:
No correlated noise. Flat noise floor. You get the picture.
Feeding 32 tones to the MRX 520 presents a pretty ugly picture as far as intermodulation distortion:
View attachment 33199
Now our distortion products leave us about 13.5 bits of dynamic range. With our best DACs, we get 20 bits.
AVR Power Amplifier Measurements
Given the fact that digital inputs provide such narrow bandwidth, I decided to test the power amp with analog input. Anthem nicely has an option to digitize the analog inputs or not. I selected not. That gave me a rather wide bandwidth:
View attachment 33200
Wider than 48 kHz sampling over HDMI anyway.
Here is our dashboard at 5 watt output into 4 ohm:
View attachment 33202
At full volume level of 0dB, we see a pretty high gain of 32 dB. I adjusted the input level accordingly to get our 5 watt output.
We see that noise spike again at 1.6 kHz indicating that it is bleeding into somewhere in the analog chain and is unrelated DAC portion itself. It is fully amplified now and actually exceeds the distortion products themselves! On top of that, our crazy channel is still acting as a variable noise generator. No wonder then that the overall SINAD is not very good. Not good at all:
View attachment 33203
I usually don't worry about crosstalk measurements as they are usually more than good enough. But here, seeing all of this interference, I thought I run it:
View attachment 33204
Gosh, this is pretty poor for an amplifier. Audibly it is probably still OK but you have more bleeding from one channel into another than the distortion it may produce!
Let's see what our power versus distortion at 4 ohm looks like:
View attachment 33205
Once again the "cartel" strikes again with strikingly similar curve to other AVRs tested. Here, even the power level is almost the same as the
NAD T758 I recently reviewed.
The very high noise levels relative to our reference Benchmark AHB2 shows that there is little "high-end" here as far as performance. Similar story exists at 8 ohm:
View attachment 33206
Notice the one channel wiggling while the other one is solid. We can see that if we just power on the unit and have it produce 50 watts:
View attachment 33207
It is like there is some boot up activity on its networking subsystem that settles down partially after a while. Good news is that the MRX 520 did not complain and kept going without getting too hot. This is far better than the
Pioneer AVRs which pull back their power level at some 35 seconds.
Peak power measurements were puzzling too, showing very little improvement relative to steady state that is used above at 4 ohm:
View attachment 33208
View attachment 33209
Seems like there is some current limiting stopping the 4 ohm configuration to produce more power. Shame as this is a more common speaker impedance than 8 ohm.
Was never a fan of these peak power measurements but seems like they are producing some useful insight here.
I better stop here as I am not sure how much grumpy reporting of measurements you all can take....
Conclusions
The Anthem MRX 520 started fine out of the gate with its DAC SINAD measurements. But seem to show a number of weaknesses after that. FIrst up, what is going on with these secret resamplings? In this day and age we can't have high sample rate playback? And no one wants to tell us that in the manual? Charge me another $100 for the DSP subsystem and do the correction at the native rate for heaven's sake. Or else, stop advertising these products as "high-end."
The noise in one channel of this AVr reminds of similar problem dating back 2014 when I tested the original Anthem MRX series AVR. Shame there is no commitment to engineering excellence to find the sources of interference and nailing it.
All of these AVRs seem to have secretly settled on some middle-of-the-road performance which is likely based on dirt-cheap components they similarly buy. At least that is my guess. It is uncanny how they fall in similar range that way. Sadly such performance is below that of even headphone dongles on phones these days.
I was almost tempted to give the decapitated Pink Panther award to the Anthem but eventually decided to give it the "I don't know what is going on" Pink Panther. Hopefully that measure of goodwill on my part (oh yeah
), will implore Anthem to do better. A bit of scrubbing inside the unit could result in a much cleaner audio product.
For now, none of these AVRs are coming close to the performance of even budget desktop products on DACs or even amps. Please excuse me as I go to jump into the bay to drown myself.....
------------
As always, questions, comments, recommendations, etc. are welcome.
Given how poor these AVRs have been performing, I am starting to get awfully depressed. Thinking about going to a psychiatrist to get better. They charge $350/hour and I am going to need many hours. So please donate generously using:
https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
