This is a review and detailed measurements of the new NAD T778 Home Theater AVR. It was kindly purchased by a member new and send to me and costs US $3,000. It is only sold through dealers and looks like if you contact them, you may get some discount from that price.
The NAD looks gorgeous and highly differentiated from your typical AVR:
The large high resolution display is a joy to look at. There are even "ear racks" included so you can mount it in an equipment rack!
A very fancy remote with motion detection and back light is included which even has a USB port. So I imagine it may be programmable. More on this a bit later.
The back connections are ultra modern, dispensing with yesterday's video inputs and such:
Despite using (modified?) Hypex class D amplifier modules, the T778 is fairly heavy, weight similarly to other AVRs. The class D amplifiers do bring a major benefit in how cool the unit runs. Only the front warmed up during my testing and even then, barely so. This is a major advantage for those of us who use our AVRs as our main TV Sound and in closed cabinets no less.
After I powered the unit with Ethernet cable plugged in, it prompted me for a firmware update which I allowed to do. This took some 15 minutes. Upon restart: the fan on the back started to go on full power blowing like nobody's business! It was so powerful that I could feel the air being sucked in from the top vents. Alas, there were two problems with this:
1. On full blast the fan is super noisy. I am talking computer server class of noise.
2. There was no reason for fan to be on! The on-screen temperature monitoring did not inspire confidence either:
A power cycle fixed the issue and the fan remained off for the rest of the testing.
The remote control has serious issues. For one, it doesn't have proper "key debounce" and sometimes repeats the button you have pressed. This is stuff that they teach first month new hire embedded programmer, not something from an established audio company who has been supplying remotes for decades.
The second issue is odd use of the four-way arrow and "enter" center button. You would think you would push the "enter" button to select the item but no, it seems to just do what the down arrow does! This is like making a stick-shift car but having the H shape being side-ways or something. Quite maddening until you get used to it.
There is a button called "RTN" next to the four-way arrows. You think that would just be the same as back but it is not. Hitting it while in menus started to select the sleep timer! There is a faint "sleep" label below it but how you select that I don't know.
Researching online there are horrendous reliability issues reported on multiple units suffering from same failures and such. If you buy this, you better use it heavily so that if anything is going to fail, it fails during warranty period.
AVR DAC Audio Measurements
Feeding HDMI input to an AVR and checking the performance of pre-out has become one of the key tests in my reviews as many AVRs pick very low voltages for this, and distort a lot post that. Let's see how the T778 did with volume set to 0 dB:
This is not half-bad! SINAD which is a sum of distortion and noise is "reasonable" by AVR standards:
And a huge step above T758 which brings in the rear. Alas, the indicator was complaining about clipping so I turned the volume down until the indicator went away which was at -8 dB on the volume display:
Usually this improves AVR performance but not here. We actually lose performance due to noise increasing and dragging SINAD down with it. Indeed, you do better to ignore the clipping indicator and drive as much as 4 volts output if you need it:
I think the hypex amps don't suck down the power supply for the DAC like classical class D amps do or else there is independent power supply for the DAC section. Their efficiency means that even if they are driven while you use an external amp, they don't generate much heat so in that case it is "fine" that they are being driven. Of course being able to disconnect their inputs as we see in Denon/Marantz AVRs would be a superior solution.
DAC dynamic range is barely above CD's 16-bit requirement of 96 dB:
Note that in general you want this metric to be 10 dB higher than what you need or else it will degrade the content's signal to noise ratio. For CD then this translates into 106 dB which we miss by fair bit here.
Intermodulation versus signal shows the same high noise floor issue:
Notice that if you run at - 8dB (in green), you are almost as bad as a phone dongle (dashed orange) when it comes to noise! We could argue that distortion is hard to hear but noise is not if you have sensitive speakers.
Multitone test shows what we already know:
Jitter shows lack of good hygiene when it comes to design but is typical of many AVRs and processors:
I discovered a major issue here in that I could not get the Coax input to work. I tried it twice but it would not produce an output. I then switched to Toslink and it worked. This may be operator error but in the span of time I had, it would not work.
DAC filter could use more attenuation but is typical of such implementations:
THD+N versus distortion showed quite high levels due to wide bandwidth of this test unlike the dashboard:
So I made a spectrum analysis with the same bandwidth to see what is above audible band:
We see some DAC filtering issues around 48 kHz but also good bit of other noise spikes which are visible even with the input shut off (in red). A bit of attention in design and measurement could have identified the sources.
Linearity is good:
AVR Amplifier Measurements
When testing AVRs amplifiers, I have a choice of using analog or digital input. The former is preferred since it eliminates the effect of the DAC. But then again if that input is digitized, then it can be worse. To test for that, I run a frequency response test:
Sad to see such poor response truncated just a hair above 20 kHz which tells me the ADC is run at 44.1 kHz. I could not find any "pure" mode in my quick look so if there is one, it may remedy this but as is, performance using analog input is not acceptable:
I switched to HDMI as the input and got improved response:
Still not much to write home about though:
And among all amplifiers tested so far:
A high-end product producing average measurements is a bad combination.
Noise performance is an issue again:
Crosstalk is typical of AV products:
Let's see how much power and distortion we have into 4 ohm load:
I have run that at two different volume levels (0 and -8 dB). Both show very early rise in distortion starting at just 5 watts! The rise is so steep that it is hard to determine where that ends and real clipping occurs. I looked up the hypex UCD102 which I believe is what NAD uses here and it too shows the same issue:
So this is not the Ncore series we love from Hypex.
Good amount of power is available if we allow 1% THD+N though:
Testing for frequency dependency, we don't see much except for 20 kHz where it oddly performed better:
This again agrees with Hypex measurements above.
Switching to 8 ohm gives us the same issue with early distortion:
Finally, let's see how power scales as we turn on amplifier channels one by one:
This is quite good and a testament to efficiency of class D amplification.
Conclusions
NAD gets a lot of things right with T778 with high efficiency and cool running amplifiers, large beautiful display, excellent cooling if needed and good DAC performance (for an AVR). Alas, there just isn't enough attention to detail with respect to noise, bugs, input digitization, etc. Given all of this, I give up and let you all decide if it fits for your purpose. I hope NAD takes this platform and this data and makes a clean up pass for next year to build a truly superior product. As it is, it doesn't get there.
-----------
As always, questions, comments, recommendations, etc. are welcome.
I figured out a great way to save money with panthers. Instead of paying them to pose for every review, I can take their pictures once and photoshop them in for future reviews!!! No more having to buy expensive steaks and clothing for them. Alas, I have a signed contract with them and I need to hire a top-tier lawyer to help me get out of it. And as usual, I lack the funds for that so please donate using : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
The NAD looks gorgeous and highly differentiated from your typical AVR:
The large high resolution display is a joy to look at. There are even "ear racks" included so you can mount it in an equipment rack!
A very fancy remote with motion detection and back light is included which even has a USB port. So I imagine it may be programmable. More on this a bit later.
The back connections are ultra modern, dispensing with yesterday's video inputs and such:
Despite using (modified?) Hypex class D amplifier modules, the T778 is fairly heavy, weight similarly to other AVRs. The class D amplifiers do bring a major benefit in how cool the unit runs. Only the front warmed up during my testing and even then, barely so. This is a major advantage for those of us who use our AVRs as our main TV Sound and in closed cabinets no less.
After I powered the unit with Ethernet cable plugged in, it prompted me for a firmware update which I allowed to do. This took some 15 minutes. Upon restart: the fan on the back started to go on full power blowing like nobody's business! It was so powerful that I could feel the air being sucked in from the top vents. Alas, there were two problems with this:
1. On full blast the fan is super noisy. I am talking computer server class of noise.
2. There was no reason for fan to be on! The on-screen temperature monitoring did not inspire confidence either:
A power cycle fixed the issue and the fan remained off for the rest of the testing.
The remote control has serious issues. For one, it doesn't have proper "key debounce" and sometimes repeats the button you have pressed. This is stuff that they teach first month new hire embedded programmer, not something from an established audio company who has been supplying remotes for decades.
The second issue is odd use of the four-way arrow and "enter" center button. You would think you would push the "enter" button to select the item but no, it seems to just do what the down arrow does! This is like making a stick-shift car but having the H shape being side-ways or something. Quite maddening until you get used to it.
There is a button called "RTN" next to the four-way arrows. You think that would just be the same as back but it is not. Hitting it while in menus started to select the sleep timer! There is a faint "sleep" label below it but how you select that I don't know.
Researching online there are horrendous reliability issues reported on multiple units suffering from same failures and such. If you buy this, you better use it heavily so that if anything is going to fail, it fails during warranty period.
AVR DAC Audio Measurements
Feeding HDMI input to an AVR and checking the performance of pre-out has become one of the key tests in my reviews as many AVRs pick very low voltages for this, and distort a lot post that. Let's see how the T778 did with volume set to 0 dB:
This is not half-bad! SINAD which is a sum of distortion and noise is "reasonable" by AVR standards:
And a huge step above T758 which brings in the rear. Alas, the indicator was complaining about clipping so I turned the volume down until the indicator went away which was at -8 dB on the volume display:
Usually this improves AVR performance but not here. We actually lose performance due to noise increasing and dragging SINAD down with it. Indeed, you do better to ignore the clipping indicator and drive as much as 4 volts output if you need it:
I think the hypex amps don't suck down the power supply for the DAC like classical class D amps do or else there is independent power supply for the DAC section. Their efficiency means that even if they are driven while you use an external amp, they don't generate much heat so in that case it is "fine" that they are being driven. Of course being able to disconnect their inputs as we see in Denon/Marantz AVRs would be a superior solution.
DAC dynamic range is barely above CD's 16-bit requirement of 96 dB:
Note that in general you want this metric to be 10 dB higher than what you need or else it will degrade the content's signal to noise ratio. For CD then this translates into 106 dB which we miss by fair bit here.
Intermodulation versus signal shows the same high noise floor issue:
Notice that if you run at - 8dB (in green), you are almost as bad as a phone dongle (dashed orange) when it comes to noise! We could argue that distortion is hard to hear but noise is not if you have sensitive speakers.
Multitone test shows what we already know:
Jitter shows lack of good hygiene when it comes to design but is typical of many AVRs and processors:
I discovered a major issue here in that I could not get the Coax input to work. I tried it twice but it would not produce an output. I then switched to Toslink and it worked. This may be operator error but in the span of time I had, it would not work.
DAC filter could use more attenuation but is typical of such implementations:
THD+N versus distortion showed quite high levels due to wide bandwidth of this test unlike the dashboard:
So I made a spectrum analysis with the same bandwidth to see what is above audible band:
We see some DAC filtering issues around 48 kHz but also good bit of other noise spikes which are visible even with the input shut off (in red). A bit of attention in design and measurement could have identified the sources.
Linearity is good:
AVR Amplifier Measurements
When testing AVRs amplifiers, I have a choice of using analog or digital input. The former is preferred since it eliminates the effect of the DAC. But then again if that input is digitized, then it can be worse. To test for that, I run a frequency response test:
Sad to see such poor response truncated just a hair above 20 kHz which tells me the ADC is run at 44.1 kHz. I could not find any "pure" mode in my quick look so if there is one, it may remedy this but as is, performance using analog input is not acceptable:
I switched to HDMI as the input and got improved response:
Still not much to write home about though:
And among all amplifiers tested so far:
A high-end product producing average measurements is a bad combination.
Noise performance is an issue again:
Crosstalk is typical of AV products:
Let's see how much power and distortion we have into 4 ohm load:
I have run that at two different volume levels (0 and -8 dB). Both show very early rise in distortion starting at just 5 watts! The rise is so steep that it is hard to determine where that ends and real clipping occurs. I looked up the hypex UCD102 which I believe is what NAD uses here and it too shows the same issue:
So this is not the Ncore series we love from Hypex.
Good amount of power is available if we allow 1% THD+N though:
Testing for frequency dependency, we don't see much except for 20 kHz where it oddly performed better:
This again agrees with Hypex measurements above.
Switching to 8 ohm gives us the same issue with early distortion:
Finally, let's see how power scales as we turn on amplifier channels one by one:
This is quite good and a testament to efficiency of class D amplification.
Conclusions
NAD gets a lot of things right with T778 with high efficiency and cool running amplifiers, large beautiful display, excellent cooling if needed and good DAC performance (for an AVR). Alas, there just isn't enough attention to detail with respect to noise, bugs, input digitization, etc. Given all of this, I give up and let you all decide if it fits for your purpose. I hope NAD takes this platform and this data and makes a clean up pass for next year to build a truly superior product. As it is, it doesn't get there.
-----------
As always, questions, comments, recommendations, etc. are welcome.
I figured out a great way to save money with panthers. Instead of paying them to pose for every review, I can take their pictures once and photoshop them in for future reviews!!! No more having to buy expensive steaks and clothing for them. Alas, I have a signed contract with them and I need to hire a top-tier lawyer to help me get out of it. And as usual, I lack the funds for that so please donate using : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/