Review and Measurements of NAD T758 V3 AVR

[Chromatischism]

Which Dirac target curves did you try?

I didn't stick with the default, though I did listen to it. I definitely didn't like Harman and I don't think it's intended to be a set-it-and-forget-it preset, either. I adjusted it to be as close to sounding "right" as I could, which was very close to what I end up with using Audyssey and DEQ's gradual bass slope. It probably sounded great to the uninitiated but I'm picky and spend a lot of time on this system so I can hear the differences. I can't tell you where the issue was causing the lack of impact at the bottom or lack of integration up top – I spent a lot of time chasing it down and eventually concluded I had tried everything with my 2 subs short of a MiniDSP. My subs have delay adjustments up to 16 ms each which should be enough in my 16-foot room, and while the resulting frequency response looked nearly on the money, the sound wasn't. Aside from those two things, I found the quality of the bass in the rest of the subwoofer range to sound similar to Audyssey. There was another user at AVS who noted the same issue with lack of sub bass impact with Dirac and I don't think he ever solved it, either.

The default does not have a lot of bass, and I don't find that commonly suggested Harman curves work well as in-room targets either because they are flat besides the house curve, unless they are modified to roll off mids and treble. And Dirac's IR correction cleans up the bass significantly so I find it is capable of running strong house curves as long as the system can handle it. After a lot of experimentation I am running the below custom curve which is a kind of hybrid between Harman 10 dB and Dirac's default curve, but with even more bass boost and treble rolloff - it is a 24 dB spread in total. It may look excessive but was finely tuned for most realism on my system which which was confirmed by several users in a series of blind tests with a variety of movies and music. I couldn't say how well it would work on other systems...

Since I have experience with this I can say that although your bass is a little high, I don't find it excessive. I bet you have sealed subs? In my small room and with vented subs I'm only needing a 6-10 dB rise (depends on volume level). It was higher when I had sealed. I like the shape of the sub bass curve for the most part – if I had to compare to mine, I wouldn't have a knee at 70 Hz but more of a slope that starts at maybe 50Hz, and I would have it blend with the speakers by 120 Hz at the latest. Every time I've had a bass boost extend to 150 Hz or higher it just sounds too thick and I lose clarity in the lower mids.

I always kept DEQ off because it also boosted satellites which was undesirable to me. Does it still work that way?

Yes it does, and I hope they address it. Still, it's not worth ditching DEQ for that. Just calibrate your surround levels with DEQ on after the fact with Dolby test tones or similar. That usually means they need to be reduced by anywhere from 3-5 in the Levels menu.

Still, from my own testing on several systems, Audyssey with the EQ editor was unable to produce the same clean level of bass, precise imaging, and live sound of Dirac.

I am not sure what constitutes "live sound", but I definitely don't think Audyssey lacks in bass quality. It uses a huge amount of taps with its FIR filters to achieve great resolution in its corrections. I can't say regarding imaging since I don't EQ outside the bass region and have a nice triangular setup with equidistant walls and waveguided speakers so imaging is great – no issues to correct there. However, I would believe if you had differences from left to right due to an uneven room setup that imaging could be improved by room correction.

What I did hear from Dirac when I ran it full range was that the speakers sounded more tame, smooth, and composed. Probably the opposite of live sound which can be more brash due to the amplification and spaces used. But you have to take that with a grain of salt because the results were specific to that speaker. I don't have the hardware anymore or I could test with a much better speaker. But there's really not much point equalizing my current speakers beyond maybe 1 kHz at the absolute most.

When I used it two years ago, Audyssey only independently set levels and delays, and then EQs the subs as one, and that is consistent with everything I've ever seen about this topic from Audyssey and its users. It would be a big deal if Audyssey could do fully independent subwoofer EQ and they would certainly market it.

That is the correct way to do it. In a room, the subwoofers will essentially act as one below a certain frequency and above that, it depends on location and phase. That's where placement, delays, and EQ come in. But you want to EQ them as one, not individually. Applying individual EQ will be a waste because the results will change when they are summed as a single LFE channel.

 

[Sal1950]

That doesn't address my point. Sure you can take REW plots of multiple positions but how do you interpret which has more precise imaging or better "live" sound (whatever that means)? "Proof" based on a complete set of measurements are not going be the most helpful in answering that. Blind listening tests are more reliable for this and that's where Audyssey routinely comes up short.

Now you've gone into the area of preference. How do you know what "imaging " is right? Did you have the engineer come by and listen for you? Same for your "live sound" claim. The measurements will tell you which is getting the flatter responce, beyond that your playing with preferences, what YOU like. Sighted or blind it's you judging the results, your not comparing to a known value. Same for room curves, even Harman's was based on listener preference.

in a controlled listening test between Dirac Live and XT32, both tweaked with the curves approaching flat between 20 and 300 Hz or so. I would bet few people could tell a difference, let alone concluding one is "better" than the other.

That's right on point, they're both capable of doing very good DRC and if you got very close measured results I also highly doubt few if any could hear a difference in a blind test.

When I used it two years ago, Audyssey only independently set levels and delays, and then EQs the subs as one, and that is consistent with everything I've ever seen about this topic from Audyssey and its users.

As has already been pointed out to you, that is the correct way to do it. The sum of two EQ'd independently would no longer give the desired result.

On last question you still haven't answered, how much did you have to pay for your DLBC upgrade?

 

[Flak]

Now you've gone into the area of preference. How do you know what "imaging " is right? Did you have the engineer come by and listen for you? Same for your "live sound" claim. The measurements will tell you which is getting the flatter responce, beyond that your playing with preferences, what YOU like.

I think it's worth noting that measurements showing that you're getting the flatter response (I'd say the closest to the target instead) does not necessarily mean that you'll always get the best listening results... actually Dirac Live 3 might compromise a bit on that aspect in order to optimize the phase coherence between speakers pairs which is important for imaging.
And no... we don't send engineers for listening instead of users

 

[Sal1950]

I think it's worth noting that measurements showing that you're getting the flatter response (I'd say the closest to the target instead) does not necessarily mean that you'll always get the best listening results...

Yes that's correct, the "target" would have been the more appropriate term to use in my sentence.

does not necessarily mean that you'll always get the best listening results... actually Dirac Live 3 might compromise a bit on that aspect in order to optimize the phase coherence between speakers pairs which is important for imaging.
And no... we don't send engineers for listening instead of users

Once again we fall back to a "preference" either designed by some luminary, engineer, or the owner-installer himself, there isn't any absolute correct "target" unless you aim strictly for flat, but that is really outside this discussion.
I've not made any attacks on Dirac's design, you guys make an excellent product, one of the best currently on the market.
This whole BruHaHa was stirred up by fanboy attitudes and claims of "all else is junk" which is miles far from the truth.
In any case this BS has taken this thread much too far off topic and I will no longer contribute to this mess.
Y'all have a good day.

 

[Bear123]

Blind listening tests are more reliable for this and that's where Audyssey routinely comes up short.

Can you post a link to these blind tests? Thanks!!

 

[peng]

Can you post a link to these blind tests? Thanks!!

Thank you for asking, hopefully we might see at least a link or two of those where "....where Audyssey routinely comes up short......." . The only blind test of that sort I could find is the one Harman did more than 10 years ago (2009 I think). No names in the ranking chart were identified but Audysssey probably ranked bottom in that one, but in that test they did not adjust the default target curve so Audyssey's flat bass would have been put at disadvantage.

Aside from that Harman test, all I could find were many subjective ones done before the Audyssey Editor app was launched. Without DBT comparison listening, we couldn't even assume the listener actually heard a difference, let alone he/she had a preference if both use the same or very similar target curve.

 

[SynthesisCinema]

I wonder did @Matthew J Poes use receiver with Mult EQ XT on this as we see worse measurements with Audussey on than off? XT32 with app shouldn`t make similar mess.

 

[Flak]

A blind comparison, together with a carefully level matched listening, isn't that easy to set up but one user took the time and effort,
he also smoothed the response curve to 1/3rd octave and lined them up so that they overlapped as much as possible:

https://www.avsforum.com/threads/blind-comparisons-of-ssps.1699697/

Dirac is the winner of that comparison but when everything else is equal we are left with the differences in the time domain and listening tests of SS&I (sound stage & imaging) are somehow subjective... we can objectively measure the impulse response and the phase coherence between speakers pairs though.

 

[peng]

A blind comparison, together with a carefully level matched listening, isn't that easy to set up but one user took the time and effort,
he also smoothed the response curve to 1/3rd octave and lined them up so that they overlapped as much as possible:

https://www.avsforum.com/threads/blind-comparisons-of-ssps.1699697/

Dirac is the winner of that comparison but when everything else is equal we are left with the differences in the time domain and listening tests of SS&I (sound stage & imaging) are somehow subjective... we can objectively measure the impulse response and the phase coherence between speakers pairs though.

Thank you for the link, I don't know what others such as Bear123 would think about those tests. I actually did find and read that one long time ago, it was hardly what I would consider "controlled" DBT like Harman's were but you are right, it is still a lot of work to set it up so I do appreciate his efforts.

It was done in 2014/15?, well before the Audyssey App became available and Dirac Live's latest DLBC obviously wasn't even developed yet I suppose.. audioguy obviously would conclude DL the winner in his experiments but if you read the fine print, he seemed very careful to not make his/their findings very conclusive either. For example, he also said "The Integra with Audyssey X32 is an amazing value. Amazing. I would, however, never consider any of their newer products since they have abandoned Audyssey in favor of a home grown room correction technology which, according to the Onkyo web site, IGNORES CORRECTING THE MAINS AND SUBS!! Seriously??" That, in my opinion, is a lot more reasonable and perhaps believable than some of the more seemingly exaggerated remarks made by some..

That's nothing "...routine..." either as claimed by database (post#860), so I am still waiting for links..

 

[peng]

I wonder did @Matthew J Poes use receiver with Mult EQ XT on this as we see worse measurements with Audussey on than off? XT32 with app shouldn`t make similar mess.


View attachment 157124

I wonder about that too. And I don't know if he just happened to pick that one particular incident that may not be easily reproduced either. Depending on how the reviewers measurement protocols (e.g. mic positions, how many, where etc..), yes you can run into some situations where it may appear worse on paper but I am sure that can happen with any EQ or no EQ. I just wish Dr. Toole and Olive would do one like they did back in 2009 but do it in a) Just run the thing following the manufacturer's instructions without adjusting, and b) As in a) but make necessary adjustments with the apps for the flattest possible FR and whatever else that contribute to neutrality and transparency. The tiled "Harman" curve thing is something that the individual user can decide on so that's a second test after the first one.

 

[Flak]

That's nothing "...routine..." either as claimed by database (post#860), so I am still waiting for links..

You don't need links if you trust measurements more than listening tests... you can compare the impulse responses yourself with REW, it's much simpler and much less costly than measuring the NAD T758 SINAD as accurately as Amir did in this thread

 

[peng]

You don't need links if you trust measurements more than listening tests... you can compare the impulse responses yourself with REW, it's much simpler and much less costly than measuring the NAD T758 SINAD as accurately as Amir did in this thread

I do trust my measurements, my comments on the blind tests referred to database post#860, and those who stated their subjective views as though the were facts. As I said before, DL definetely did better in terms of impule measurements of my own even when I was using the trial PC version.

It seems to me DBTs (not on REQ systems)were often done to prove or disapprove subjective measurments and validate or invalidate certain objective objective measurements.

 

[database]

Since I have experience with this I can say that although your bass is a little high, I don't find it excessive. I bet you have sealed subs? In my small room and with vented subs I'm only needing a 6-10 dB rise (depends on volume level). It was higher when I had sealed. I like the shape of the sub bass curve for the most part – if I had to compare to mine, I wouldn't have a knee at 70 Hz but more of a slope that starts at maybe 50Hz, and I would have it blend with the speakers by 120 Hz at the latest. Every time I've had a bass boost extend to 150 Hz or higher it just sounds too thick and I lose clarity in the lower mids.

I have a pair of Rythmik FV18s. They're ported but they almost sound sealed. I've experienced what you're talking about with the bass boost extending too high and my previous curve didn't extend as high, but its deep bass response was weaker. This curve was born from trying to fix that, but I may have to revisit this curve yet again...

Yes it does, and I hope they address it. Still, it's not worth ditching DEQ for that. Just calibrate your surround levels with DEQ on after the fact with Dolby test tones or similar. That usually means they need to be reduced by anywhere from 3-5 in the Levels menu.

I turned off DEQ first because it bloated the bass too much, and second because of the satellite levels issue. I only really listen at one particular volume, but different material is mastered at different volumes, so I figured DEQ just added unwanted complexity.

I am not sure what constitutes "live sound", but I definitely don't think Audyssey lacks in bass quality. It uses a huge amount of taps with its FIR filters to achieve great resolution in its corrections. I can't say regarding imaging since I don't EQ outside the bass region and have a nice triangular setup with equidistant walls and waveguided speakers so imaging is great – no issues to correct there. However, I would believe if you had differences from left to right due to an uneven room setup that imaging could be improved by room correction.

What I did hear from Dirac when I ran it full range was that the speakers sounded more tame, smooth, and composed. Probably the opposite of live sound which can be more brash due to the amplification and spaces used. But you have to take that with a grain of salt because the results were specific to that speaker. I don't have the hardware anymore or I could test with a much better speaker. But there's really not much point equalizing my current speakers beyond maybe 1 kHz at the absolute most.

I'm not sure how to define "live sound" either, or what contributes to creating a live sound. But I know I only got it after moving to Dirac Live (that was the first time stereo music brought me any excitement). I'm sure it's the result of several factors. Maybe Dirac's impulse response correction that Audyssey doesn't have is one factor. Maybe making sure you don't overcorrect the frequency response in a small area is another factor - if you space the measurements too close together with Dirac, the results sound dull and the soundstage is collapsed.

That is the correct way to do it. In a room, the subwoofers will essentially act as one below a certain frequency and above that, it depends on location and phase. That's where placement, delays, and EQ come in. But you want to EQ them as one, not individually. Applying individual EQ will be a waste because the results will change when they are summed as a single LFE channel.

No, not quite. To be clear:

• Standard Dirac treats each subwoofer as one, and corrects individual subwoofers. This is incorrect, and you can work around it by using a splitter when connecting the subwoofers to the processor. For many subs you can control levels and delays individually by using the sub amp controls.
• Audyssey can individually set levels and delays on levels on subwoofers, but then EQs them together. This is OK.
• Dirac Live Bass Control (DLBC) and Multi-Sub Optimizer (MSO) can individually set levels, delays, and EQs on subwoofers, with the objective of optimizing their combined response. This is the best approach, and successful implementations have been thoroughly documented on AVS.

Now you've gone into the area of preference. How do you know what "imaging " is right? Did you have the engineer come by and listen for you? Same for your "live sound" claim. The measurements will tell you which is getting the flatter responce, beyond that your playing with preferences, what YOU like. Sighted or blind it's you judging the results, your not comparing to a known value. Same for room curves, even Harman's was based on listener preference.

That's right on point, they're both capable of doing very good DRC and if you got very close measured results I also highly doubt few if any could hear a difference in a blind test.

As has already been pointed out to you, that is the correct way to do it. The sum of two EQ'd independently would no longer give the desired result.

On last question you still haven't answered, how much did you have to pay for your DLBC upgrade?

I actually did have a small time engineer listen to one of the songs he mixed, and he was blown away by the results, saying it sounded better than what he heard in his studio. He was able to hear the layers of tracks he added come through with a level of separation that he didn't have in his studio. So my goal isn't what the engineer heard when they mixed the song. Anyway, that's just a story. This is getting into Toole's circle of confusion, I don't think there's much point in discussing this angle.

The difference between Audyssey's approach and DLBC is explained above.

DLBC is sold in tiers. Some processors include the top tier of Bass Control without an additional purchase. The NAD T758v3 doesn't support Bass Control. For the HTP-1 it's a $500 upgrade for the top tier for multiple subwoofers (https://www.dirac.com/online-store/monolith-htp-1/). I bought the HTP-1 as B-Stock so I put the savings towards full Bass Control.

Yes that's correct, the "target" would have been the more appropriate term to use in my sentence.

Once again we fall back to a "preference" either designed by some luminary, engineer, or the owner-installer himself, there isn't any absolute correct "target" unless you aim strictly for flat, but that is really outside this discussion.
I've not made any attacks on Dirac's design, you guys make an excellent product, one of the best currently on the market.
This whole BruHaHa was stirred up by fanboy attitudes and claims of "all else is junk" which is miles far from the truth.
In any case this BS has taken this thread much too far off topic and I will no longer contribute to this mess.
Y'all have a good day.

You're right about this all being stirred up with the fanboy attitude in this post: https://www.audiosciencereview.com/...s-of-nad-t758-v3-avr.8912/page-42#post-920195

Thank you for asking, hopefully we might see at least a link or two of those where "....where Audyssey routinely comes up short......." . The only blind test of that sort I could find is the one Harman did more than 10 years ago (2009 I think). No names in the ranking chart were identified but Audysssey probably ranked bottom in that one, but in that test they did not adjust the default target curve so Audyssey's flat bass would have been put at disadvantage.

Aside from that Harman test, all I could find were many subjective ones done before the Audyssey Editor app was launched. Without DBT comparison listening, we couldn't even assume the listener actually heard a difference, let alone he/she had a preference if both use the same or very similar target curve.

I'm not referring to anything new - only links I've already posted to Home Theater Shack, AVS, and my own tests. I'm not aware of any tests done in scientific publications. If you want that, someone on ASR will probably have to volunteer their time and money and do it themselves.

I do trust my measurements, my comments on the blind tests referred to database post#860, and those who stated their subjective views as though the were facts. As I said before, DL definetely did better in terms of impule measurements of my own even when I was using the trial PC version.

It seems to me DBTs (not on REQ systems)were often done to prove or disapprove subjective measurments and validate or invalidate certain objective objective measurements.

I didn't realize you already verified superior IR measurements from Dirac. You said earlier you were looking for objective proof with measurements. Don't you already have them, then? Here's another set of IR measurements where Dirac performs best: https://www.avsforum.com/threads/ro...-dirac-2-0-accueq-2018.3037658/#post-57324882

 

[Chromatischism]

No, not quite. To be clear:

• Standard Dirac treats each subwoofer as one, and corrects individual subwoofers. This is incorrect, and you can work around it by using a splitter when connecting the subwoofers to the processor. For many subs you can control levels and delays individually by using the sub amp controls.
• Audyssey can individually set levels and delays on levels on subwoofers, but then EQs them together. This is OK.
• Dirac Live Bass Control (DLBC) and Multi-Sub Optimizer (MSO) can individually set levels, delays, and EQs on subwoofers, with the objective of optimizing their combined response. This is the best approach, and successful implementations have been thoroughly documented on AVS.

I'm not sure what the "no, not quite" means because I don't disagree with any of this.

I turned off DEQ first because it bloated the bass too much

I definitely don't have this issue and I'm very sensitive to it. What kind of speakers and what is your crossover? Is your response smooth?

 

[Flak]

• Standard Dirac treats each subwoofer as one, and corrects individual subwoofers. This is incorrect, and you can work around it by using a splitter when connecting the subwoofers to the processor. For many subs you can control levels and delays individually by using the sub amp controls.

I'd not say it's incorrect ... as you mention multiple subs should be connected, measured, and corrected as if they were one so that Dirac will "see" their combined response, and commonly AVRs have two sub outputs that are equivalent to a Y splitter.
Even better results can be obtained when fully independent channels are available for each subwoofer and in that case, Dirac Bass Control can be used (provided that it's supported by the AVR)

 

[Dj7675]

I'd not say it's incorrect ... as you mention multiple subs should be connected, measured, and corrected as if they were one so that Dirac will "see" their combined response, and commonly AVRs have two sub outputs that are equivalent to a Y splitter.
Even better results can be obtained when fully independent channels are available for each subwoofer and in that case, Dirac Bass Control can be used (provided that it's supported by the AVR)

Is their a list of all receivers/AVPs that support DLBC?

 

[toddsdonald]

I stumbled on this while doing a Google search and was really surprised with the horrible review of the NAD T 758 V3.

I'm definitely not a scientist or electronics whiz, but this seems like the polar opposite of the real world performance of the NAD receivers. Have you spoken to NAD to discuss your measurements and findings?

This should get me laughed right off the forum, but I've heard tons of sound systems (stereo and surround), and have to say I've never heard anything sound as good as an NAD for AV. My ears aren't objective like your tests, but from listening to an NAD powered system I can't imagine anyone walking away thinking about poor measurements.

Out of curiosity, (and please if you reply put it in layman's terms), why does my 10 year old used NAD T757 V2 I bought on eBay for $75 plus shipping, absolutely smoke my Sony DH790 and puts my Sony ZA810ES to shame? I run my NAD in direct, (which is simple you just turn off all the various surround modes to "none"), and run it in tone defeat so it's as raw as I can get it.

I would think that in terms of electronic wizardry and measurements Sony would be off the charts in terms of electronic perfection, (heck don't they engineer movie sound)?... Anyway I've read elsewhere that comparing an NAD to a Sony is like comparing a Rolls Royce to a Kia. As a car guy I can definitely appreciate the analogy.

I suppose I'm just confused, if the NAD is so awful why does it sound so good?

My next question would be, if an NAD receiver is as bad as you say, yet sounds excellent in use, what's better and how big of a difference could I expect?

Thanks!

 

[peng]

I stumbled on this while doing a Google search and was really surprised with the horrible review of the NAD T 758 V3.

I'm definitely not a scientist or electronics whiz, but this seems like the polar opposite of the real world performance of the NAD receivers. Have you spoken to NAD to discuss your measurements and findings?

This should get me laughed right off the forum, but I've heard tons of sound systems (stereo and surround), and have to say I've never heard anything sound as good as an NAD for AV. My ears aren't objective like your tests, but from listening to an NAD powered system I can't imagine anyone walking away thinking about poor measurements.

Out of curiosity, (and please if you reply put it in layman's terms), why does my 10 year old used NAD T757 V2 I bought on eBay for $75 plus shipping, absolutely smoke my Sony DH790 and puts my Sony ZA810ES to shame? I run my NAD in direct, (which is simple you just turn off all the various surround modes to "none"), and run it in tone defeat so it's as raw as I can get it.

I would think that in terms of electronic wizardry and measurements Sony would be off the charts in terms of electronic perfection, (heck don't they engineer movie sound)?... Anyway I've read elsewhere that comparing an NAD to a Sony is like comparing a Rolls Royce to a Kia. As a car guy I can definitely appreciate the analogy.

I suppose I'm just confused, if the NAD is so awful why does it sound so good?

My next question would be, if an NAD receiver is as bad as you say, yet sounds excellent in use, what's better and how big of a difference could I expect?

Thanks!

I think you can answer your own questions by a) find a way to do a comparison listening tests that are properly controlled including the use of direct mode free from any dsp/tune control functions to make sure it is only the amps are being compared, and do it in DBT form, b) same as a) but let Dirac Live does it job, though this may not be relevant because the T757 V2 does not have DL, just Audyssey.

Short of doing the above, no matter how anyone may (not me, I won't..) try to convince you amps at such level simply should, and most often do sound transparent, and that it boils down to the quality of the source contents, source players, and the speaker's in-room performance that determine the final sound quality, the rest of it would be left to one's imagination blablabla.. But then I am sure such talks would only "..get you laughed right off the forum...." right?

 

[SMc]

I suppose I'm just confused, if the NAD is so awful why does it sound so good?

My next question would be, if an NAD receiver is as bad as you say, yet sounds excellent in use, what's better and how big of a difference could I expect?

First, you are operating it within its limits. Second, if you don't have a problem now, there's nothing to improve upon.

OTOH, Dirac should make an audible difference.

 

[Chromatischism]

It also depends what input is being used