This is a review and detailed measurements of the NAD M33 Streaming Player with Purifi Eigentakt Amplification. It is kindly sent for testing from a member and has a retail cost of US $4,999.
The NAD is part of their "Masters" series and has one of the best industrial designs of any audio gear I have seen:
It is both modern and luxurious. The touchscreen is not as fast and accurate as your phone or tablet but ages ahead of many other displays on audio gear. On the other hand, as I have noted before, I hate the touch power button on top which seemingly requires multiple touches to actually turn the unit off.
Notice the cool instrumentation panel telling you the internal temperatures. Then again the unit runs so cool that such monitor is not needed in practice.
The back panel shows decent connectivity although I was very disappointed to not see balanced XLR pre-out:
There is balanced input but why not output?
Dirac Room EQ is provided putting it on par with home theater AVRs in that regard but obviously for only two channels.
I was pleased to see network firmware update work and work a lot faster than a lot of AVRs and processors. It indicates a clean architecture than a boatload of licensed chunks/subsystems each of which has its own kludgy update mechanism (I am talking to you Arcam).
No USB DAC functionality is provided which is a big miss. And Roon endpoint lacks certification which means it is useless with newer versions of Roon that I have. Roon now refuses to use these uncertified implementations. Strange to see a company the size of NAD has not finished the certification. Ariplay streaming works but it limits streaming to 16 bits.
NAD M33 DAC Measurements
As usual we start with our dashboard of feeding the unit digital content and measuring what analog output we have out of the pre-out. First measurement was disappointing:
Distortion was extremely low but that rise in low frequency noise was not something I had seen. I then realized the amps were on so I turned them off and performance much improved:
This is much better performance than we typically see in this class of product. Comparing it to AVRs and AV Processors for example, the M33 is way above the class:
Whoever said you can't combine EQ and DACs and get good performance was just proven wrong!
Output can go up to 4 volts so let's measure the full response range:
Nice! You have a broad range of usability up to 3.2 volts.
Dynamic range is very good as well:
Linearity is nailed:
Jitter over Coax input is problematic from measurement point of view but not audibility:
Since jitter amplitude is frequency dependent (gets worse the higher the frequency is), it actually pollutes the higher frequency tones in multitone:
Otherwise performance is very good.
IMD distortion relative to level is excellent sans our famous "ESS DAC IMD Hump:"
Fascinating that small companies in China have solved this problem yet a major company like NAD has not. Shame on ESS for not providing a solution for all of their OEMs to benefit.
Reconstruction filter could be better:
This impacts the default measurement of THD+N versus frequency due to allowance for more ultrasonic noise:
NAD M33 Amplifier Measurement
The Purifi amplifier used in NAD is one of our favorite, top performing class D amps. My previous review of the NAD M28 Amplifier showed that NAD had compromised performance some, likely in their buffer stage. Let's see how the M33 performs using analog input first:
Ah, looks like we take another hit again with a SINAD of 93 dB versus 98 dB for M28 which is a shame. Looking at the frequency response however shows that the analog input is digitized first and then converted back to analog for amplification:
The pass to and from digital domain adds distortion and noise so degrades performance some. Let's use digital input then and eliminate the effect of the ADC:
We have good and bad news here. Good news is that the ADC is essentially transparent which is a first in any digital device/AVR/AVP. Bad news is that eliminating it didn't get us much. So for the rest of the tests I will be using analog input for easy of measurements. The ranking is still very good though:
And comparison to AVRs:
Likely the amplifier is as good as what is in M28. It is just that we don't have a transparent path to it. The above AVRs were mostly measured with digital path to the amplifier so a bit unfair. Still the M33 outperforms them all.
I love to see 16 bits of dynamic range at 5 watt output but rarely see it. The M33 delivers on that which is great:
Crosstalk is excellent and near state of the art:
Let's see the power into 4 ohm:
Ah, this is not as impressive as I was expecting especially in the one channel. Lots of power is available though at 421 watts.
I could not get burst performance because the amp's protection would kick in. But here is the power at 1% THD:
Same into 8 ohm:
Let's switch to digital input to see if there is improvement:
We are doing much better than AVRs but the channel inconsistency is still there.
Finally let's test for frequency dependency:
If you took away the writing on this graph I would tell you this is a measurement of class AB amp, not switching class D! Most class D amps have all kinds of strange non-linearities but not this Purifi amp. There is no change in response with power and only modest increase in distortion with frequency. The clean heart of Purifi still beats inside the NAD M33!
NAD M33 Streaming DAC Measurements
As I noted in the intro, I was out of business streaming lossless 24 bit test signals I need for our dashboard. Roon would refuse to talk to it putting up a "shame you NAD" dialog box. So I resorted to Airplay which due to conversion to 16 bit has a limit of 93 dB SINAD:
So we are as good as the transport allows.
Conclusions
There is a lot of good news and some bad news in review of NAD M33. As an integrated product it shows for the first time that quality does not need to be compromised. Performance across the board beats any integrated product like an AVR or even AV Processor on DAC performance. For the first time, we don't have to apologize or make excuses.
If you were however expecting the pure performance of best in class stereo DAC and Purifi amplification you would be disappointed. We can get SINAD of 120 dB in desktop DACs and Purifi performance of 98 dB. And at much lower prices but you would have add a streamer of sorts. And none will have the pretty UI of the NAD M33. And you would have to find one where you can use Dirac (e.g. a PC platform).
As compared to an AVR, the M33 is a home run albeit at much higher price and with just two channels.
All in all, NAD shows the path to superb performance in integrated products. It however stops a bit short of what it could be.
Overall, I am happy to recommend the NAD M33. Just miss the bit of performance it left on the table.
-----------
As always, questions, comments, recommendations, etc. are welcome.
Planted a bunch of watermelon and cantaloupe against my better judgement as we just don't have enough sunny days for them to fully develop. Still, the plants did their best producing these cute little melons:
Alas, inside was mostly white and barely sweet. I have a much larger one to test so hoping that one will be better.
Appreciate any kind donations using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
The NAD is part of their "Masters" series and has one of the best industrial designs of any audio gear I have seen:
It is both modern and luxurious. The touchscreen is not as fast and accurate as your phone or tablet but ages ahead of many other displays on audio gear. On the other hand, as I have noted before, I hate the touch power button on top which seemingly requires multiple touches to actually turn the unit off.
Notice the cool instrumentation panel telling you the internal temperatures. Then again the unit runs so cool that such monitor is not needed in practice.
The back panel shows decent connectivity although I was very disappointed to not see balanced XLR pre-out:
There is balanced input but why not output?
Dirac Room EQ is provided putting it on par with home theater AVRs in that regard but obviously for only two channels.
I was pleased to see network firmware update work and work a lot faster than a lot of AVRs and processors. It indicates a clean architecture than a boatload of licensed chunks/subsystems each of which has its own kludgy update mechanism (I am talking to you Arcam).
No USB DAC functionality is provided which is a big miss. And Roon endpoint lacks certification which means it is useless with newer versions of Roon that I have. Roon now refuses to use these uncertified implementations. Strange to see a company the size of NAD has not finished the certification. Ariplay streaming works but it limits streaming to 16 bits.
NAD M33 DAC Measurements
As usual we start with our dashboard of feeding the unit digital content and measuring what analog output we have out of the pre-out. First measurement was disappointing:
Distortion was extremely low but that rise in low frequency noise was not something I had seen. I then realized the amps were on so I turned them off and performance much improved:
This is much better performance than we typically see in this class of product. Comparing it to AVRs and AV Processors for example, the M33 is way above the class:
Whoever said you can't combine EQ and DACs and get good performance was just proven wrong!
Output can go up to 4 volts so let's measure the full response range:
Nice! You have a broad range of usability up to 3.2 volts.
Dynamic range is very good as well:
Linearity is nailed:
Jitter over Coax input is problematic from measurement point of view but not audibility:
Since jitter amplitude is frequency dependent (gets worse the higher the frequency is), it actually pollutes the higher frequency tones in multitone:
Otherwise performance is very good.
IMD distortion relative to level is excellent sans our famous "ESS DAC IMD Hump:"
Fascinating that small companies in China have solved this problem yet a major company like NAD has not. Shame on ESS for not providing a solution for all of their OEMs to benefit.
Reconstruction filter could be better:
This impacts the default measurement of THD+N versus frequency due to allowance for more ultrasonic noise:
NAD M33 Amplifier Measurement
The Purifi amplifier used in NAD is one of our favorite, top performing class D amps. My previous review of the NAD M28 Amplifier showed that NAD had compromised performance some, likely in their buffer stage. Let's see how the M33 performs using analog input first:
Ah, looks like we take another hit again with a SINAD of 93 dB versus 98 dB for M28 which is a shame. Looking at the frequency response however shows that the analog input is digitized first and then converted back to analog for amplification:
The pass to and from digital domain adds distortion and noise so degrades performance some. Let's use digital input then and eliminate the effect of the ADC:
We have good and bad news here. Good news is that the ADC is essentially transparent which is a first in any digital device/AVR/AVP. Bad news is that eliminating it didn't get us much. So for the rest of the tests I will be using analog input for easy of measurements. The ranking is still very good though:
And comparison to AVRs:
Likely the amplifier is as good as what is in M28. It is just that we don't have a transparent path to it. The above AVRs were mostly measured with digital path to the amplifier so a bit unfair. Still the M33 outperforms them all.
I love to see 16 bits of dynamic range at 5 watt output but rarely see it. The M33 delivers on that which is great:
Crosstalk is excellent and near state of the art:
Let's see the power into 4 ohm:
Ah, this is not as impressive as I was expecting especially in the one channel. Lots of power is available though at 421 watts.
I could not get burst performance because the amp's protection would kick in. But here is the power at 1% THD:
Same into 8 ohm:
Let's switch to digital input to see if there is improvement:
We are doing much better than AVRs but the channel inconsistency is still there.
Finally let's test for frequency dependency:
If you took away the writing on this graph I would tell you this is a measurement of class AB amp, not switching class D! Most class D amps have all kinds of strange non-linearities but not this Purifi amp. There is no change in response with power and only modest increase in distortion with frequency. The clean heart of Purifi still beats inside the NAD M33!
NAD M33 Streaming DAC Measurements
As I noted in the intro, I was out of business streaming lossless 24 bit test signals I need for our dashboard. Roon would refuse to talk to it putting up a "shame you NAD" dialog box. So I resorted to Airplay which due to conversion to 16 bit has a limit of 93 dB SINAD:
So we are as good as the transport allows.
Conclusions
There is a lot of good news and some bad news in review of NAD M33. As an integrated product it shows for the first time that quality does not need to be compromised. Performance across the board beats any integrated product like an AVR or even AV Processor on DAC performance. For the first time, we don't have to apologize or make excuses.
If you were however expecting the pure performance of best in class stereo DAC and Purifi amplification you would be disappointed. We can get SINAD of 120 dB in desktop DACs and Purifi performance of 98 dB. And at much lower prices but you would have add a streamer of sorts. And none will have the pretty UI of the NAD M33. And you would have to find one where you can use Dirac (e.g. a PC platform).
As compared to an AVR, the M33 is a home run albeit at much higher price and with just two channels.
All in all, NAD shows the path to superb performance in integrated products. It however stops a bit short of what it could be.
Overall, I am happy to recommend the NAD M33. Just miss the bit of performance it left on the table.
-----------
As always, questions, comments, recommendations, etc. are welcome.
Planted a bunch of watermelon and cantaloupe against my better judgement as we just don't have enough sunny days for them to fully develop. Still, the plants did their best producing these cute little melons:
Alas, inside was mostly white and barely sweet. I have a much larger one to test so hoping that one will be better.
Appreciate any kind donations using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
