• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Anthem AVM60 Review (a second test)

Joined
Apr 8, 2020
Messages
15
Likes
45
Location
Italy
#1
Intrigued by Amirm's review of Anthem AVM60 A/V processor and having one unit available, I tried my hand at making quickly some measurements, to see where they fit together or not.
I state that these results are very good, even if they are obtained in a less stressful situation. Amirm’s measurement results are much worse, maybe due to a unit defect or a set up not fully correct (a modern A/V is very complex to configure).


AUDIO SPECIFICATIONS
11.2 Channels Preamp; XLR/RCA output
DAC: AKM AK4458 Differential Output
ADC: AKM AK5381 Delta-Sigma
DSP: Quad-Core Cirrus Logic CS49844A

Frequency response (2Vrms output): 10Hz–35kHz +0, –0.25dB
THD+N: -95dB at 2V output
S/N: Ratio: 110dB at 2V output


MEASUREMENTS
All following measurements have been made with this set up:
- Input: HDMI L+R at 24bit/192KHz PCM, via a blu-ray player (only audio, no video);
- Output: XLR Front L+R;
- ADC: via RME ADI-2 PRO FS at 24bit/192KHz (Sharp filter);
- Load: 18 Kohm;
- ARC disabled, no bass management, no dolby volume;
- Volume: -3dB (2Vrms output at -1.8dBFS input);


FREQUENCY RESPONSE

Module L+R: -1.8dBFS Tone sweep 10Hz-96KHz; 2Vrms output (-3dB)
Amp LR 2V 0dB.jpg


The frequency response is very good, flat beyond the audible band, with a little difference between channels, < 0.2dB.

Phase L+R: -1.8dBFS Tone sweep 10Hz-96KHz; 2Vrms output (-3dB)
Phase LR 2V 0dB.jpg

Small distortion in low frequency; pronounced in the mids and highs: this is due to the D/A filter, as can be seen in the impulse response.


IMPULSE RESPONSE

Impulse Response, band limited @96KHz: 2Vrms output (-3dB)
Impulse 2V 0dB.jpg

The curve is not symmetric: It seems that the AVM60 uses IIR filters for D/A conversion, not phase-linear.


CROSSTALK

Crosstalk RL+LR: -1.8dBFS Tone sweep 10Hz-96KHz; Volume -3dB; 1/3 Smoothing
Crosstalk LR 2V 0dB.jpg

Crosstalk figure show very good values, below -110dB at 1KHz, with modest increase up to -100dB at 20KHz; small dB the difference between the Front L and the Front R.


HARMONIC DISTORTION

Sine -1.8dBFS @ 1KHz R: 96KHz bandwidth, 1M samples, 2Vrms output (-3dB)
Sin 1KHz R 2V 0dB.jpg

At 2Vrms output, THD+N is -100dB (20Hz-20KHz bandwidth) for a 1KHz tone, with a noise floor of -100.4dB; overall, good values for an A/V processor. However, harmonic distortion is high in the third harmonic (!); Higher harmonics are very low. The other peaks are present also without the input signal:

Noise (no input signal) L+R: 96KHz bandwidth, 1M samples, Volume -3dB
Noise LR 2V.jpg

Without any input signal, the noise floor does not show alarming artefacts. Some (low) peaks are due to power supply (50Hz + harmonics); some at 60Hz and multiples (?); left channel has some additionals (still low) peaks around 12KHz, probably due to internal interference.


HARMONIC DISTORTION PER FREQUENCY

THD per frequency R: -1.8dBFS Tone sweep 10Hz-96KHz; 2Vrms output (-3dB)
THD R 2V 0dB.jpg

THD is mainly due to the second harmonic, "flat" for all frequencies. But the third one (less “euphonic”) increases with the frequency, overcoming just before 2KHz the second harmonic, contributing to increase the THD, in any case low.


INTERMODULATION

CCIF Test R: 96KHz bandwidth, 1M samples; Volume -3dB
CCIF R 1.6V.jpg

SMPTE Test R: 96KHz bandwidth, 1M samples; Volume -3dB
SMPTE R 2V.jpg

Good both tests: IMD is -93.2dB for CCIF and -89.5dB for SMPTE (20Hz-20KHz band limited). Symmetrical peaks around high frequency signals are due to jitter.

Multitone 25Hz-20KHz 1/3 Oct. R: 96KHz bandwidth, 1M samples; Volume -3dB
Multone 25-10K 2V.jpg

From this graph we’ve about 18 bit of distortion free range (see next).


DAC LINEARITY

Relative Level of a tone @1KHz from 0dBFS to -130dBFS; Volume -3dB
1614207952464.png

It seems that conversion linearity is maintained at least up to 18bit; not very good for an audio DAC, but “normal” for a good A/V processor.


JITTER NOISE

J-Test: 48KHz sample rate, 0.5M samples; Volume -3dB
J-Test 48K R.jpg

J-test reveals a modest amount of low frequency spurious jitter tones. However, they are likely not audible at these levels.


CONCLUSIONS

Based on above (not complete) measurements, I would say that the AVM60 behaviour is good, even if from a device of this class I would expect more. Considering that an A/V processor is used normally for home cinema, with a DSP elaborating heavily the audio signal for the room acoustic correction (ARC in this case, very effective), it fulfils well this need with all its features (not addressed here). As a stereo preamplifier used for music, as usual for all A/V, there is space for improvements.

Comments and suggestions are welcome.
 
Last edited:

peng

Major Contributor
Forum Donor
Joined
May 12, 2019
Messages
1,775
Likes
1,393
#2
Intrigued by Amirm's review of Anthem AVM60 A/V processor and having one unit available, I tried my hand at making quickly some measurements, to see where they fit together or not.
I anticipate that these ones result a bit better: Maybe the unit tested by Amirm has some defects or it is not set correctly.


AUDIO SPECIFICATIONS
11.2 Channels Preamp; XLR/RCA output
DAC: AKM AK4458 Differential Output
ADC: AKM AK5381 Delta-Sigma
DSP: Quad-Core Cirrus Logic CS49844A

Frequency response (2Vrms output): 10Hz–35kHz +0, –0.25dB
THD+N: -95dB at 2V output
S/N: Ratio: 110dB at 2V output


MEASUREMENTS
All following measurements have been made with this set up:
- Input: HDMI L+R at 24bit/192KHz PCM, via a blu-ray player (only audio, no video);
- Output: XLR Front L+R;
- ADC: via RME ADI-2 PRO FS at 24bit/192KHz (Sharp filter);
- Load: 18 Kohm;
- ARC disabled, no bass management, no dolby volume;
- Volume: -3dB (2Vrms output at -1.8dBFS input);


FREQUENCY RESPONSE

Module L+R: -1.8dBFS Tone sweep 10Hz-96KHz; 2Vrms output (-3dB)
View attachment 114680

The frequency response is very good, flat beyond the audible band, with a little difference between channels, < 0.2dB.

Phase L+R: -1.8dBFS Tone sweep 10Hz-96KHz; 2Vrms output (-3dB)
View attachment 114683

Small distortion in low frequency; pronounced in the mids and highs: this is due to the D/A filter, as can be seen in the impulse response.


IMPULSE RESPONSE

Impulse Response, band limited @96KHz: 2Vrms output (-3dB)
View attachment 114684
The curve is not symmetric: It seems that the AVM60 uses IIR filters for D/A conversion, not phase-linear.


CROSSTALK

Crosstalk RL+LR: -1.8dBFS Tone sweep 10Hz-96KHz; Volume -3dB; 1/3 Smoothing
View attachment 114686

Crosstalk figure show very good values, below -110dB at 1KHz, with modest increase up to -100dB at 20KHz; small dB the difference between the Front L and the Front R.


HARMONIC DISTORTION

Sine -1.8dBFS @ 1KHz R: 96KHz bandwidth, 1M samples, 2Vrms output (-3dB)
View attachment 114689

At 2Vrms output, THD+N is -100dB (20Hz-20KHz bandwidth) for a 1KHz tone, with a noise floor of -100.4dB; overall, good values for an A/V processor. However, harmonic distortion is high in the third harmonic (!); Higher harmonics are very low. The other peaks are present also without the input signal:

Noise (no input signal) L+R: 96KHz bandwidth, 1M samples, Volume -3dB
View attachment 114690

Without any input signal, the noise floor does not show alarming artefacts. Some (low) peaks are due to power supply (50Hz + harmonics); some at 60Hz and multiples (?); left channel has some additionals (still low) peaks around 12KHz, probably due to internal interference.


HARMONIC DISTORTION PER FREQUENCY

THD per frequency R: -1.8dBFS Tone sweep 10Hz-96KHz; 2Vrms output (-3dB)
View attachment 114691

THD is mainly due to the second harmonic, "flat" for all frequencies. But the third one (less “euphonic”) increases with the frequency, overcoming just before 2KHz the second harmonic, contributing to increase the THD, in any case low.


INTERMODULATION

CCIF Test R: 96KHz bandwidth, 1M samples; Volume -3dB
View attachment 114694

SMPTE Test R: 96KHz bandwidth, 1M samples; Volume -3dB
View attachment 114695

Good both tests: IMD is -93.2dB for CCIF and -89.5dB for SMPTE (20Hz-20KHz band limited). Symmetrical peaks around high frequency signals are due to jitter.

Multitone 25Hz-20KHz 1/3 Oct. R: 96KHz bandwidth, 1M samples; Volume -3dB
View attachment 114696

From this graph we’ve about 18 bit of distortion free range (see next).


DAC LINEARITY

Relative Level of a tone @1KHz from 0dBFS to -130dBFS; Volume -3dB
View attachment 114692

It seems that conversion linearity is maintained at least up to 18bit; not very good for an audio DAC, but “normal” for a good A/V processor.


JITTER NOISE

J-Test: 48KHz sample rate, 0.5M samples; Volume -3dB
View attachment 114697

J-test reveals a modest amount of low frequency spurious jitter tones. However, they are likely not audible at these levels.


CONCLUSIONS

Based on above (not complete) measurements, I would say that the AVM60 behaviour is good, even if from a device of this class I would expect more. Considering that an A/V processor is used normally for home cinema, with a DSP elaborating heavily the audio signal for the room acoustic correction (ARC in this case, very effective), it fulfils well this need with all its features (not addressed here). As a stereo preamplifier used for music, as usual for all A/V, there is space for improvements.

Comments and suggestions are welcome.
Thank you for the interesting tests. Just a few questions if you don't mind.

1) Input signal -1.8 dBFS, and you only get 2 V from the XLR? Looking at Amir's graph, the best result appeared to be at about 4 V output and I think he normally used 0 dBFS (unless specified otherwise). If you do the same, do you think THD+N would even been better in your test?

2) From the full test results I highly doubt Amir's test sample would be defective, I can't imagine something would be slightly defective to have that kind of results, without showing something more obvious and I would also expect some erratic behavior in terms of the test result such as inconsistencies, fluctuation etc., so it could be the way he hooked everything up resulting in more noise? As a control, do you have another unit to measure, one that Amir also measured, such as one of the D+M's?

3) What kind of test instruments would I need to do the same kind of tests?
 
OP
Pinox67
Joined
Apr 8, 2020
Messages
15
Likes
45
Location
Italy
Thread Starter #3
Thank you for the interesting tests. Just a few questions if you don't mind.

1) Input signal -1.8 dBFS, and you only get 2 V from the XLR? Looking at Amir's graph, the best result appeared to be at about 4 V output and I think he normally used 0 dBFS (unless specified otherwise). If you do the same, do you think THD+N would even been better in your test?

2) From the full test results I highly doubt Amir's test sample would be defective, I can't imagine something would be slightly defective to have that kind of results, without showing something more obvious and I would also expect some erratic behavior in terms of the test result such as inconsistencies, fluctuation etc., so it could be the way he hooked everything up resulting in more noise? As a control, do you have another unit to measure, one that Amir also measured, such as one of the D+M's?

3) What kind of test instruments would I need to do the same kind of tests?
Hi Peng,
Below my answers.

1) Using a -1.8dBFS input signal you obtain 2Vrms in the XLR output, setting the A/V volume to -3dB (checked with a multimeter). But this set up is specific for this A/V, it depends essentially by its gain. This is a less stressing situation respect to 4Vrms, of course, but:
- it is closer to real working situations (power amps usually have input sensibility around 2Vrms);
- it is closer to brand's specification.
I suppose that increasing the volume will give worst results.

2) About connections/fluctuations: I don't know. An A/V has a multitude of settings... also in my initial measurements the data was very strange, and the reason was a wrong A/V setting. I don't have another unit to measure.

3) You need: a PC with a measurement SW, a multimeter, and mostly an audio interface with high precision. I use RME ADI-2 Pro FS with REW and Audition, used also by Archimago for its measurements. In his site you can find all details about this set up.
 
Last edited:
Joined
Jan 3, 2021
Messages
19
Likes
7
#4
@Pinox67
I think it is awesome that you did this for a round 2. I'm not an EE, but my engineering background allows me to read these and appreciate much of it, but not quite all. And I pay attention to other reviews and try to get a feel of where a product comes down on its specs.

Its much better than my..."Well I hooked me up my Radio Shack Volt Meter, powered it on, and cranked that sucker up. Sounded gooood."...analysis.
 
OP
Pinox67
Joined
Apr 8, 2020
Messages
15
Likes
45
Location
Italy
Thread Starter #5
@Pinox67
I think it is awesome that you did this for a round 2. I'm not an EE, but my engineering background allows me to read these and appreciate much of it, but not quite all. And I pay attention to other reviews and try to get a feel of where a product comes down on its specs.

Its much better than my..."Well I hooked me up my Radio Shack Volt Meter, powered it on, and cranked that sucker up. Sounded gooood."...analysis.
I'm glad you appreciate my work.
Like you, to understand the performance of a device, I collect information from different sources. But if you have the device available, you can test it directly in your laboratory or audo/video set-up: this is the best, because you can try to understand its strengths/weaknesses yourself and compare them to other opinions.
I have reported here only the results (reproducible under the same conditions) of my experience with the measurements, with few comments.
If there is any aspect that is not clear or incomplete let me know, I would be happy to investigate them.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
32,923
Likes
110,687
Location
Seattle Area
#7
1) Using a -1.8dBFS input signal you obtain 2Vrms in the XLR output, setting the A/V volume to -3dB (checked with a multimeter). But this set up is specific for this A/V, it depends essentially by its gain. This is a less stressing situation respect to 4Vrms, of course, but:
- it is closer to real working situations (power amps usually have input sensibility around 2Vrms);
- it is closer to brand's specification.
I suppose that increasing the volume will give worst results.
Well, why not find out? If you are going to counter my measurements and declare that my unit was broken, you need to use the same test conditions. As noted, that should be 0 dBFS and 4 volt output.
 

peng

Major Contributor
Forum Donor
Joined
May 12, 2019
Messages
1,775
Likes
1,393
#9
Hi Amirm,
good point. I’ll try to do some measurements in the next days.
Thank you, that would be really interesting to see, and I think the results may even be better at higher output voltage. It would answer the question that whether the results get better with -1.8 dBFS and higher volume setting, or 0 dBFS and a lower volume setting, for the same 4 V output voltage.

So in each test please include the volume settings.
 

Spocko

Addicted to Fun and Learning
Forum Donor
Joined
Sep 27, 2019
Messages
672
Likes
1,187
Location
Southern California
#10
As Amir will be reviewing the Trinnov Altitude 16 shortly, I welcome you to post questions and measurements requests you'd like to see! I am collecting those questions HERE for @amirm to consider.
 

danzilla31

Member
Forum Donor
Joined
Feb 14, 2021
Messages
53
Likes
57
#11
As Amir will be reviewing the Trinnov Altitude 16 shortly, I welcome you to post questions and measurements requests you'd like to see! I am collecting those questions HERE for @amirm to consider.
This will be an interesting review. Boy if he finds anything off with the Trinnov well let's just bring out the popcorn fellas.
 

Anton S

New Member
Joined
Mar 9, 2021
Messages
1
Likes
2
#12
Bench measurements presented in a Dec 2017 review of the AVM 60, beginning about 2/3 of the way down THIS page, confirm an increase in distortion for measurements made at 4V RMS out vs. 2V RMS out. Note the results for an HDMI input signal at -5 dBFS (2V out) vs. 0 dBFS (4V out) XLR. However, as previously noted by Pinox67, most power amps will reach rated power with a 2V RMS signal in.
 

Similar threads

Top Bottom