This is a review and detailed measurements of the NAD T758 V3 Home Theater Audio/Video Receiver (AVR). It was kindly sent to me by a member. The T758 appears to have been released last year and costs US $1,399.
The T758 is quite a bit heavier than recently measured AVRs and has a serious, business-like look that I appreciate:
As you see, there are hardly any buttons on the front. To change settings, you push the Menu button and then navigate using the 4-way control. I am used to such controls have a center "enter" button but this one does not. Instead you had to push the right button to go into that setting. And left to get out. I could not get used to this and kept pushing the thing in.
The 4-way control is almost too responsive. It can double select easily so it would be nice if it had a bit more "key debounce" control. Nothing major though.
The menus are all text and continue the business-like feel of the unit which I was fine with. No need to wait 5 seconds for the slow CPU in these units to draw a picture of a living room for you.
The back panel shows a modular architecture where all the digital AV switching is on a vertical card to the left:
Given the thin margins on these products, I have yet to see a company provide any upgrades of such modules in the future that makes economic sense. I did like the organization though of having them grouped in one place, and the amplifier/analog ins and outs elsewhere.
The speaker terminals are a little less flimsy than some of the other AVRs I have tested but still won't give you a feeling of luxury.
I get a kick out of IEC inputs that lack safety ground. I guess they give you modularity and ability to change to a different power cord compared to the cheap units of the past with the wires coming out of the unit. Similarly it is cute to see the switched outlet in there. I can see an old-timer at NAD insisting on having this connector even though the consumer will likely never use it as he has a lot more than one other device to connect (cable box, etc.).
I could not find any "pure" or "direct" mode to bypass internal processing. There is such a mode though but it is hidden as you will see in testing.
There are hefty, albeit small heatsinks inside. They are quite far from the top of the unit so can't feel them for how hot they get. The top of the unit though was hardly warm even through my testing. I am assuming some kind of power supply rail switching (i.e. Class G) is used in the amplification to keep power consumption low.
The NAD T758 comes with Dirac Lite room equalization which if it works right, should make it audibly perform better than many AVRs that use lesser systems.
Despite my complaints here and there, my overall impression of the NAD T578 was a positive one. It definitely sets its apart from mass-market brand AVRs.
AVR DAC Testing
For testing, I started with S/PDIF coax input since my analyzer can control it without issue unlike HDMI. HDMI results will come in the next section. For output, I used the convenient pre-amp output. Here is what the "DAC" portion performs:
No, your eyes are not deceiving you. When set to 0 dB volume on the unit, you get massive harmonic distortion with the third harmonic dominating. As a result, the NAD T758 takes the dubious crown of worst measured DAC of any product I have tested:
We can get much better results by dialing the input level or volume control by 6 dB:
But now, we have an unfair comparison as the output level has dropped to just 1 volt. But even going with this, the T758 cannot escape the fourth and worst quadrant SINAD of any DAC tested.
Notice that even at this lowered volume, we have issues with noise. The FFT noise floor is quite high and SINAD is determined by that, rather than distortion products. A problem that we see when we measure its dynamic range:
The problem with rising distortion with level shows up clearly in intermodulation distortion versus level:
As you see, you best avoid the last 6 dB of your dynamic range. Even putting that side, the high noise level lowers performance significantly at lower input levels.
Linearity test shows improper signal processing in conversion of 24 bit samples in the input signal to something lower:
You are barely good enough for 16 bit content.
Distortion products are quite visible in 32 tone intermodulation test:
If you want your signal to be free of distortion products, you only have 70 dB or 11.6 bits to work with! I have seen tube products with less distortion than this.
To see what is going on, I set the sampling rate to 192 kHz and ran a frequency response sweep. Ideally we would see 96 kHz of flat response:
But we don't. Response ends at 48 kHz or so, telling me perhaps the input is resampled to 96 kHz. If so, this is twice as good as the recent Pioneer VSX-LX504 I tested so by itself is not bad. Then again, someone sticking a S/PDIF cable into this receiver doesn't expect to see it resampled by half.
The resampling process seems to also reduce low frequency response as we are down 0.7 dB at the low end, and 0.9 dB at 20 kHz. Small numbers to be sure but in this day and age, I better get flat response from 20 Hz to 20 kHz when I running at 192 kHz sampling rate.
THD+N versus frequency shows what we already know:
Except that now we see a rise in high frequency distortion to boot! And no, this is not caused by ultrasonics. It is the clipping behavior I showed in the dashboard.
Could it commit any more sins? It can in the form of jitter:
OK, this is not end of the world but clean it isn't. Thankfully, likely it is not "audioble."
DAC filter response is fine but plateaued at just -60 dB:
HDMI DAC Performance
I routed the output of my PC through the T758 so that I could then send audio to use using WASAPI exclusive mode in my Roon media player. Here is our dashboard again, this time with HDMI as the input to the unit:
Overall distortion are there and set the SINAD limit to similar value. But in addition, we have a ton of "junk" thrown in there for good measure.
Jitter test generates similar outcome to S/PDIF input:
This says that we don't have input specific jitter but rather, just garbage leaking into the output of the DAC from various sources. There is a pattern in there which may indicate data dependent jitter (which is worse that random).
Based on these results, we know that our S/PDIF input is representative of HDMI performance.
AVR Amplifier Audio Measurements
The amplifier in these products is wideband, giving us an opportunity to diagnose further how the audio pipeline works. Let's start by feeding the unit analog input through CD jack and measuring the frequency response. And the compare the same to using multi-channel input:
Ah, we see that the roll off is gone when you use multichannel input. So it seems that the NAD's notion of "pure" or "pure direct" is to use the analogy 7.1 input -- a relief for people who have sources such as SACD players for this type of use.
The digitization of the CD input is quite harsh at just 44.1 kHz. Why not humor us with something higher than this?
Alas, we are limited by the performance of the power amplifier so using multi-channel input versus CD doesn't make much of a difference:
This gives pretty poor scores in ranking of the amplifiers we have tested:
Multichannel input has lower noise floor though:
Most important test is power versus distortion+noise. Here it is at 4 ohm:
That is healthy amount of power so putting aside the high noise floor, this is not too bad.
Power drops as expected into 8 ohm:
Peak power is much more as expected:
But far short of the spec although they use a different standard than what is used in my testing.
Stability was excellent at 50 watts into 4 ohm. Unlike the Pioneer amplifiers I have tested, there was no massive pull-back after a few seconds.
Conclusions
The overall fit and finish of the NAD T758 V3 was to my liking as I reported earlier. Alas, objective performance is quite shameful. It appears that no testing was performed for the digital processing or even simple distortion measurements. No wonder that outside of the amplifier, NAD doesn't provide any specifications. It appears that even after you spend $1,400, you get distortion and noise factory in these AVRs. A $9 Apple dongle easily outperforms what these devices produce.
These companies need to have a reset, go back to the drawing board and have proper architectural review and design verification. It won't cost any money to get better performance. Despite being a 20 year old part, the DAC chip they use is spec'ed at 96 dB SINAD for example. Toiling that down to just 55 dB is insulting.
As much as I hate to say it given the good looks and feeling of the NAD T758 V3, I cannot recommend it.
------------
As always, questions, comments, recommendations, etc. are welcome.
The panthers started school this week and are complaining that every kid has a cell phone but they do not! I don't know how I can afford to buy and pay the monthly fee for half a dozen phones for them. Please help by donating generously using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
The T758 is quite a bit heavier than recently measured AVRs and has a serious, business-like look that I appreciate:
As you see, there are hardly any buttons on the front. To change settings, you push the Menu button and then navigate using the 4-way control. I am used to such controls have a center "enter" button but this one does not. Instead you had to push the right button to go into that setting. And left to get out. I could not get used to this and kept pushing the thing in.
The 4-way control is almost too responsive. It can double select easily so it would be nice if it had a bit more "key debounce" control. Nothing major though.
The menus are all text and continue the business-like feel of the unit which I was fine with. No need to wait 5 seconds for the slow CPU in these units to draw a picture of a living room for you.
The back panel shows a modular architecture where all the digital AV switching is on a vertical card to the left:
Given the thin margins on these products, I have yet to see a company provide any upgrades of such modules in the future that makes economic sense. I did like the organization though of having them grouped in one place, and the amplifier/analog ins and outs elsewhere.
The speaker terminals are a little less flimsy than some of the other AVRs I have tested but still won't give you a feeling of luxury.
I get a kick out of IEC inputs that lack safety ground. I guess they give you modularity and ability to change to a different power cord compared to the cheap units of the past with the wires coming out of the unit. Similarly it is cute to see the switched outlet in there. I can see an old-timer at NAD insisting on having this connector even though the consumer will likely never use it as he has a lot more than one other device to connect (cable box, etc.).
I could not find any "pure" or "direct" mode to bypass internal processing. There is such a mode though but it is hidden as you will see in testing.
There are hefty, albeit small heatsinks inside. They are quite far from the top of the unit so can't feel them for how hot they get. The top of the unit though was hardly warm even through my testing. I am assuming some kind of power supply rail switching (i.e. Class G) is used in the amplification to keep power consumption low.
The NAD T758 comes with Dirac Lite room equalization which if it works right, should make it audibly perform better than many AVRs that use lesser systems.
Despite my complaints here and there, my overall impression of the NAD T578 was a positive one. It definitely sets its apart from mass-market brand AVRs.
AVR DAC Testing
For testing, I started with S/PDIF coax input since my analyzer can control it without issue unlike HDMI. HDMI results will come in the next section. For output, I used the convenient pre-amp output. Here is what the "DAC" portion performs:
No, your eyes are not deceiving you. When set to 0 dB volume on the unit, you get massive harmonic distortion with the third harmonic dominating. As a result, the NAD T758 takes the dubious crown of worst measured DAC of any product I have tested:
We can get much better results by dialing the input level or volume control by 6 dB:
But now, we have an unfair comparison as the output level has dropped to just 1 volt. But even going with this, the T758 cannot escape the fourth and worst quadrant SINAD of any DAC tested.
Notice that even at this lowered volume, we have issues with noise. The FFT noise floor is quite high and SINAD is determined by that, rather than distortion products. A problem that we see when we measure its dynamic range:
The problem with rising distortion with level shows up clearly in intermodulation distortion versus level:
As you see, you best avoid the last 6 dB of your dynamic range. Even putting that side, the high noise level lowers performance significantly at lower input levels.
Linearity test shows improper signal processing in conversion of 24 bit samples in the input signal to something lower:
You are barely good enough for 16 bit content.
Distortion products are quite visible in 32 tone intermodulation test:
If you want your signal to be free of distortion products, you only have 70 dB or 11.6 bits to work with! I have seen tube products with less distortion than this.
To see what is going on, I set the sampling rate to 192 kHz and ran a frequency response sweep. Ideally we would see 96 kHz of flat response:
But we don't. Response ends at 48 kHz or so, telling me perhaps the input is resampled to 96 kHz. If so, this is twice as good as the recent Pioneer VSX-LX504 I tested so by itself is not bad. Then again, someone sticking a S/PDIF cable into this receiver doesn't expect to see it resampled by half.
The resampling process seems to also reduce low frequency response as we are down 0.7 dB at the low end, and 0.9 dB at 20 kHz. Small numbers to be sure but in this day and age, I better get flat response from 20 Hz to 20 kHz when I running at 192 kHz sampling rate.
THD+N versus frequency shows what we already know:
Except that now we see a rise in high frequency distortion to boot! And no, this is not caused by ultrasonics. It is the clipping behavior I showed in the dashboard.
Could it commit any more sins? It can in the form of jitter:
OK, this is not end of the world but clean it isn't. Thankfully, likely it is not "audioble."
DAC filter response is fine but plateaued at just -60 dB:
HDMI DAC Performance
I routed the output of my PC through the T758 so that I could then send audio to use using WASAPI exclusive mode in my Roon media player. Here is our dashboard again, this time with HDMI as the input to the unit:
Overall distortion are there and set the SINAD limit to similar value. But in addition, we have a ton of "junk" thrown in there for good measure.
Jitter test generates similar outcome to S/PDIF input:
This says that we don't have input specific jitter but rather, just garbage leaking into the output of the DAC from various sources. There is a pattern in there which may indicate data dependent jitter (which is worse that random).
Based on these results, we know that our S/PDIF input is representative of HDMI performance.
AVR Amplifier Audio Measurements
The amplifier in these products is wideband, giving us an opportunity to diagnose further how the audio pipeline works. Let's start by feeding the unit analog input through CD jack and measuring the frequency response. And the compare the same to using multi-channel input:
Ah, we see that the roll off is gone when you use multichannel input. So it seems that the NAD's notion of "pure" or "pure direct" is to use the analogy 7.1 input -- a relief for people who have sources such as SACD players for this type of use.
The digitization of the CD input is quite harsh at just 44.1 kHz. Why not humor us with something higher than this?
Alas, we are limited by the performance of the power amplifier so using multi-channel input versus CD doesn't make much of a difference:
This gives pretty poor scores in ranking of the amplifiers we have tested:
Multichannel input has lower noise floor though:
Most important test is power versus distortion+noise. Here it is at 4 ohm:
That is healthy amount of power so putting aside the high noise floor, this is not too bad.
Power drops as expected into 8 ohm:
Peak power is much more as expected:
But far short of the spec although they use a different standard than what is used in my testing.
Stability was excellent at 50 watts into 4 ohm. Unlike the Pioneer amplifiers I have tested, there was no massive pull-back after a few seconds.
Conclusions
The overall fit and finish of the NAD T758 V3 was to my liking as I reported earlier. Alas, objective performance is quite shameful. It appears that no testing was performed for the digital processing or even simple distortion measurements. No wonder that outside of the amplifier, NAD doesn't provide any specifications. It appears that even after you spend $1,400, you get distortion and noise factory in these AVRs. A $9 Apple dongle easily outperforms what these devices produce.
These companies need to have a reset, go back to the drawing board and have proper architectural review and design verification. It won't cost any money to get better performance. Despite being a 20 year old part, the DAC chip they use is spec'ed at 96 dB SINAD for example. Toiling that down to just 55 dB is insulting.
As much as I hate to say it given the good looks and feeling of the NAD T758 V3, I cannot recommend it.
------------
As always, questions, comments, recommendations, etc. are welcome.
The panthers started school this week and are complaining that every kid has a cell phone but they do not! I don't know how I can afford to buy and pay the monthly fee for half a dozen phones for them. Please help by donating generously using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/