• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Big news coming from Sound United in 2023!

Now SU, pulls out the brasso, and periodically polishes it with new app interfaces.... as it gradually becomes less relevant.

But the real question is if it’s audibly different from Dirac.

Because the other argument is that while everyone else is leasing their car, or renting their home, SU was able to just buy it and amortize the costs. You have to imagine that XT32 can trickle down into the very lowest priced products as CPU power improves, esp. when there is no licensing fee to worry about.
 
But the real question is if it’s audibly different from Dirac.

Because the other argument is that while everyone else is leasing their car, or renting their home, SU was able to just buy it and amortize the costs. You have to imagine that XT32 can trickle down into the very lowest priced products as CPU power improves, esp. when there is no licensing fee to worry about.

Switching from 2013 version XT32 (Integra DTR 70.4) to 2022 version Dirac Live (Integra DRX 3.4) - the difference was immediately audible and noticeable.

But, as I have said before - the MRC in Audyssey is probably the main issue (there was no way to disable MRC without the "Pro kit" on the Audyssey AVR) - and then there would also be the target curve / voicing - which also differ.

Having said that - voices and midrange details were muffled, imaging poor, with Audyssey, with Dirac the details suddenly "popped" - like a camera image coming into focus - all the midrange stuff became clear, imaging was now clearly discernible, and vocals on many movies, which had previously been indistinguishable mumbles, became audible & clear with Dirac. - I think it likely that with the latest versions and the app which can disable MRC - this might no longer be an issue - but was not willing to but a couple of grand of my money on the line to find out! - preferred to try Dirac... and was not disappointed.

It seems that the most optimistic thing that can be said about Audyssey XT32, is that it matches Dirac Live
Audyssey's include SubEQ, seems to only do very basic time/level latching, and not the sophisticated integration that Dirac DLBC does. (still, better than noting, and the basic Dirac Live AVR's don't have DLBC at all... so no level of multi sub integration)


I still think that the reasoning behind adding Dirac, is based on something new that Dirac is about to launch, and which may well be a game changer.
 
You forgot the MRC "BBC Dip" debacle - which I believe is still the default.... without shelling out for the app, you cannot get Audyssey to work properly!
Midrange Compensation is not the BBC dip nor is it a debacle - it's still recommended for many speakers that have a dip in that range to prevent the EQ from boosting it when there's a directivity mismatch. The fact that Dirac pays no attention to this makes it flawed.
 
Switching from 2013 version XT32 (Integra DTR 70.4) to 2022 version Dirac Live (Integra DRX 3.4) - the difference was immediately audible and noticeable.
Curious. Did you run the exact same target curve on both systems? Same measurement points? Very few people have done this.

Having said that - voices and midrange details were muffled, imaging poor, with Audyssey, with Dirac the details suddenly "popped" - like a camera image coming into focus - all the midrange stuff became clear, imaging was now clearly discernible, and vocals on many movies, which had previously been indistinguishable mumbles, became audible & clear with Dirac. - I think it likely that with the latest versions and the app which can disable MRC - this might no longer be an issue - but was not willing to but a couple of grand of my money on the line to find out! - preferred to try Dirac... and was not disappointed.
Yeah, I just don't have any of those issues and neither does anyone else I know. It's possible you are benefiting more from "speaker correction" rather than room correction. Which is fine and I'm glad it's working for you. :cool:
 
Last edited:
Curious. Did you run the exact same target curve on both systems? Same measurement points? Very few people have done this.


Yeah, I just don't have any of those issues and neither does anyone else I know. It's possible you are benefiting more from "speaker correction" rather than room correction.
Was it possible to adjust the target curve on a 2013 Integra AVR.... No, not without the Pro Kit... costing about half the price of the AVR over again!

So, no I did not adjust the target curve - the base comparison was the default from Audyssey, vs the Default from Dirac, on my speakers, in my room, with my amps.

Previous attempts had used Quad ESL63's and Quad ESL989's (in my previous home) - they sounded better with Audyssey OFF
Later attempts were on Gallo Reference 3.2 (in both my previous and current home)- they too sounded better with Audyssey OFF

Now with Dirac - whether running the default curve, a curve designed to match my speakers natural voicing, or a custom room voicing - Dirac sounds better ON - note that with Dirac I am now able to set various curves and have experimented with the target curves... I am not a fan of most "Harman" curves, and find them too heavy on the bass ... but yes I have tried a range of target curves, and in most cases (unless something went badly wrong with the target curve) - the system sounds better with Dirac enabled.

The Audyssey based AVR has now moved on - so further testing is not possible.

To me, for my purposes, Dirac was the right way to go, confirmed by my listening.
As stated earlier, it is possible that the current audyssey with the app adjustments, would match Dirac... ie: at the most basic: be able to sound better with it ON than with it OFF

But fundamentally the default from Audyssey failed to achieve that with my 2008 AVR Audyssey XT, and with my 2013 AVR Audyssey XT32 - my 2022 Dirac Live AVR achieves it easily.

Yes it is possible that some of my specific observations relate to Speaker EQ... But my disappointment with Audyssey also extended to my time running Quad ESL's - it just never sounded as good in my room, as the speakers/room did without Audyssey.
 
But the real question is if it’s audibly different from Dirac.
Oh, it is definitely audibly different. Dramatically so. Everyone in the room that experienced the
MWAVE Room Correction Shootout
put Audyssey dead last. Yes, every single person. I know because I was there.
 
Last edited:
Midrange Compensation is not the BBC dip nor is it a debacle - it's still recommended for many speakers that have a dip in that range to prevent the EQ from boosting it when there's a directivity mismatch. The fact that Dirac pays no attention to this makes it flawed.
This only mattered in 2012.
If you own any of those many speakers, you need to toss them on the bonfire.
There is no need today for a room correction software that creates a dip just because 70% of people still own crap speakers.
95% of people today who have Audyssey running would already know better than to use that kind of crap in their systems.
 
Last edited:
Midrange Compensation is not the BBC dip nor is it a debacle - it's still recommended for many speakers that have a dip in that range to prevent the EQ from boosting it when there's a directivity mismatch. The fact that Dirac pays no attention to this makes it flawed.

I think I agree with your assumption that the room correction target curve should generally track the measured sound power, though the Olive blind tests using a speaker with bad sound power problems (B&W N802) provides evidence to the contrary. Anyone who has actually used Dirac (your comments suggest you have not) knows that their curve drawing tool makes those adjustments much easier than on Audyssey’s iOS app or the dumbass faux PEQ thing on MultEQ X (or RoomPerfect). That’s IMO the main difference between pre-DLBC Dirac and all flavors of Audyssey. However, one should avoid speakers with midrange sound power problems regardless of room correction employed. It’s easy enough to eliminate many bad speaker designs by sight. Well designed speakers always should be the priority over audio electronics or software. Otherwise it’s a lipstick-on-a-pig situation.

Curious. Did you run the exact same…measurement points? Very few people have done this.

Unless the systems are unstable and unsuitable for use, that is not necessary. The whole point is that the samples provide a good representation of the sound power. If they don’t, then the sampling process is by definition a failure.
 
Last edited:
But the real question is if it’s audibly different from Dirac.
Who can ever really say? Every single run of the software with either product will produce a different tuning? LOL
Funny but none the less true. The final results will always be in the hands of the operator.

That’s IMO the main difference between pre-DLBC Dirac and all flavors of Audyssey.
Nope, the biggest difference was Audyssey's ability to tune 2 separate woofers for 5 years before Dirac.
It and Atmos was the number one reason I upgraded from my Marantz AV7701 to a 7703.
I know what your going to say next, Audyssey doesn't do it right, blah, blah blah.
Right or wrong, something was WAY better than nothing. ;)
 
Who can ever really say? Every single run of the software with either product will produce a different tuning? LOL
Funny but none the less true.

That shouldn’t be the case, except in the case of software bug or user error. Look at my Marantz 7703 and Denon X4100 reviews to see how closely different runs on different products with different microphones track in the hands of a competent operator. Also, a software bug :)


Nope, the biggest difference was Audyssey's ability to tune 2 separate woofers for 5 years before Dirac.
It and Atmos was the number one reason I upgraded from my Marantz AV7701 to a 7703.
I know what your going to say next, Audyssey doesn't do it right, blah, blah blah.

That, and what little Audyssey does for two subs can be done manually in just a few minutes, with REW and the Audyssey mic if you don’t have a different one.

Right or wrong, something was WAY better than nothing. ;)

Maybe, maybe not. It depends. I agree that someone who doesn’t measure will likely imagine an improvement whether or not one exists just because the product says it will.
 
What the hell are you talking about? Please point out ONE “misstatement” I’ve made regarding Audyssey.
I already did, but you apparently ignored it.
Audyssey does not control bass management at all. That is handled by the AVR.
That is a false statement. It obviously doesn't control 100% of the bass management (you're stuck with a 4th order low pass on the sub, but I'm not sure why you'd want anything else), but it makes a significant contribution to it on the speaker side and that contribution is now adjustable with MQX in the ways I described above.
If you look at my Denon AVR and Marantz AVP reviews you’ll see I have thoroughly documented the performance of the Audyssey suite, frankly probably more so than anyone else in their whole multi-decade history.
That's nice, but I hate to tell you, without any experience with MQX, much less extensive experience with it, those reviews are largely now obsolete for any high end user of Audyssey. A short list of things MQX allows that wasn't possible with Audyssey before, even with the app:

Already discussed are the adjustments to bass management mentioned above, along with customizing the low frequency cutoff for those running speakers full range.

While it's still not as easy as I'd like (I wish they'd allow the import of a frequency response curve) creating the target curve is dramatically easier and more accurate than the app. In the app, you could only make adjustments relative to the Reference curve which obviously wasn't a flat line and not an exact science. With MQX you can make a real curve with a few PEQ's in REW and then you have an identical target curve for REW to verify the corrections (and make after corrections/adjustments afterward if you like).

Two separate target curves in the same file. This allows for near instant switching between the two for listening tests. It may only be one single PEQ you want to test (such as filling in a crossover dip or not) or a completely different curve. Or you could use this for BEQ on your subs, along with the appropriate level adjustment.

Speaking of PEQ's, people were rightly excited when the HTP-1 came out and allowed 16 PEQs along with Dirac. MQX allows an infinite number of PEQs. You can easily import these PEQs directly from REW.

After adjustments. This goes along with the PEQs. With MQX correcting to an identical curve you have in REW, it's really easy to check its work and make corrections to any deviations you find objectionable. The more mic positions you use, the fewer there will be, but there might be some (especially if using an uncalibrated mic). Simply import the filters from REW. I typically only bother with the LCR and subs with a calibrated mic.

32 possible mic positions. This provides a substantially better spatial average which will result in better and more repeatable corrections. I've found MQX measurements to be very, very, close to MMM measurements.

Individually calibrated mic. They have them now and you can use them with MQX. The mic I was using prior had a pretty large high frequency rolloff. Alternatively, any DIYer worth his salt can use MQX to calibrate whatever Audyssey mic they have to a reference mic, create a correction file and easily apply that to all his target curves.

You can see the actual filter for each speaker. This goes hand in hand for setting the frequency EQ limits. With MQX, you can set the lower limit (impossible before MQX) as well as the upper limit. For example, if you run a ported speaker full range and want a nice bass boost (who doesn't?) but obviously don't want to boost below the tuning frequency, you can drag the lower EQ limit down until you see it boosting below that frequency. The same with a high frequency rolloff--you can drag the upper limit upward until you see it boosting the crap out of the tweeter at a frequency you probably can't hear anyway but will eat into the headroom of the tweeter and stop before it does that.

Individual measurements. If you move a speaker, replace a speaker or pair of speakers, add a sub, move a sub, replace a sub or whatever, you can measure just that speaker and incorporate the change into your calibration without having to re-calibrate all 12, 14, 16, etc channels and start from scratch. This is an immense time saver and eliminates the discouraging feeling many have about making changes in their system because they dread re-doing a full calibration.

A more obscure one--different sweet spots. Some people for some reasons may want more than one. One for his favorite chair when he's alone, and maybe on for a more central position for when he has visitors but doesn't want to go through the effort if doing a full calibration for both locations. If they aren't too far apart and he has good speakers, simply doing a single measurement at the secondary sweet spot may get him 90% there with virtually no time/effort. Simply "Exclude" that position in the primary file, then recalculate all the trims and delays based on that measurement in MQX for the secondary file. It works surprisingly well.

Those are all very real, meaningful things MQX allows that wasn't possible before off the top of my head. I'm sure I've forgotten a few. They make a real difference to a high end user and would hope any "reviews" by anyone who speaks with authority on the capabilities of Audyssey in the year 2022 would be very familiar with them.
The “SubEQ” approach to combining two subs is crude (level match over some frequency and “time align,” then EQ the sum, as opposed to exploiting level and time as additional variables to improve performance) is best described as “better than nothing, probably,”
Yes, it is a notch down from a MSO-type correction, no doubt. But until very, very recently, that was as good as it got for room correction below the Trinov/JBL Synthesis cost level. People who wanted better did it themselves. There are still people with high end processors waiting for a functioning version of DLBC to show up for their processors. It'll be interesting to see what they offer for the new units--keep in mind, the Denon/Marantz hardware was not capable of doing anything more in the past. We have no idea what Audyssey will do with the newer hardware.
 
Was it possible to adjust the target curve on a 2013 Integra AVR....
That's a really nice story. But it has no relevance to this thread. Please find some threads about 10 year old AVR's and post your thoughts there where they might be relevant.
 
Oh, it is definitely audibly different. Dramatically so. Everyone in the room that experienced the
MWAVE Room Correction Shootout
put Audyssey dead last. Yes, every single person. I know because I was there.
Shootouts are fun get-togethers, but often that's about it. I was unable to find any info on test procedures on the above site. If one wants to use such a shootout as any sort of evidence, actual procedures must be fully documented and available. Getting the most out of Audyssey requires some knowledge (which is a legit criticism of it). Without detailed descriptions of how it was set up, such anecdotal results are no more meaningful than dlaloum's above. Useless.
 
Oh, it is definitely audibly different. Dramatically so. Everyone in the room that experienced the
MWAVE Room Correction Shootout
put Audyssey dead last. Yes, every single person. I know because I was there.
Was the same target curved used with the same measurement points? Did they know how to use all of the options and features?
 
Personally I feel AudysseyX is the equal Dirac and they're both always playing the upgrade game so "what's best" can change next week. Like so much else in HiFi the winds are controlled by a lot of myth and perception.
D-M was kind of forced into the current move by Dirac being the current fav posterboy product.
The average receiver buyer will want no part of ether's TOTL, too complicated, the learning curve is too steep. Only the enthusiasts like us here even understand the most of it.
Then there's also better, even more complicated stuff out there if you want a real headache. LOL
I just got my new Denon X4700H but I'm staying with the $20 Editor app, at least for now.
Same. I plan on waiting for whatever the newest version of Dirac is (comes after DLBC), then about another 6 months for the bugs to be worked out. I've used Dirac for years and it's incredibly buggy for a while.

Followed by making sure that DEQ is available. Then I'll have to see what I like more between MQX and Dirac (it will be contigent upon me being able to remove the subs from calibration then modifying the delay for the mains after calibration) and either way I'll be spending more money because I dumb licensing... I have some time with this X4700. DEQ is a massive upgrade from not having it before (I have a large surround sound setup).
 
That is a false statement. It obviously doesn't control 100% of the bass management

You can’t have it both ways. My statement was clearly true as written by your above admission.

Unless the software keeps each channel together (I.e. the target curve for each channel encompasses the whole channel, not just a limited bandwidth of it with the rest of it stuck elsewhere.

BTW one could add highpass and low pass filters to channel curve components as early as Dirac 1.

That's nice, but I hate to tell you, without any experience with MQX, much less extensive experience with it, those reviews are largely now obsolete for any high end user of Audyssey.

They remain accurate to the core functionality, which hasn’t changed. To the extent they were obsoleted, it was by the iOS app that allow one to use the core good parts of Audyssey and ditch the terrible default target curves.

While it's still not as easy as I'd like (I wish they'd allow the import of a frequency response curve) creating the target curve is dramatically easier and more accurate than the app.

Correct me if I’m wrong, but MultEQ X doesn’t even allow you to shape the target curve directly. You have to use the dumbass abstraction of faux PEQ bands. Therefore, it’s actually a step backwards from the $20 iOS app, which admittedly does need a “copy curve” button.

Speaking of PEQ's, people were rightly excited when the HTP-1 came out and allowed 16 PEQs along with Dirac.

For different reasons though. The utility of PEQ on HTP-1 (along with Storm, Datasat, etc) is to “precondition” prior to running room correction. One use case is in place of a separate DSP for assisted speakers. Another use case is to take down a a big peak that’s throwing off level/volume calibration before EQ.

Note that I’m writing about actual utility, not what excites people on the internet.

32 possible mic positions. This provides a substantially better spatial average which will result in better and more repeatable corrections.

Does it? Per the Geddes and Blind AES paper 5 points generally suffice. If you want to take 32 different measurements of 15 different speakers for one calibration good for you I guess.

I don’t have time or inclination to address the rest point by point. Someone may find them useful and even worth the bizzaroworld consumer-hating license terms. Every product likely finds some fans. However, to me the story of MultEQ X is one of obvious omissions, far too little too late functionality, with unacceptable license terms being the nail in the coffin.
 
Shootouts are fun get-togethers, but often that's about it. I was unable to find any info on test procedures on the above site. If one wants to use such a shootout as any sort of evidence, actual procedures must be fully documented and available. Getting the most out of Audyssey requires some knowledge (which is a legit criticism of it). Without detailed descriptions of how it was set up, such anecdotal results are no more meaningful than dlaloum's above. Useless.
I remember a shootout some time ago where Audy came in well behind Dirac, but IIRC these shootouts tend to be running each RC on Full Auto, Full Range. As Toole has explained in some detail, that’s a sure-fire way to end up with mid-fi at best, no matter which product we discuss, and talking about relative ‘pop’ to the sound etc is only of interest to those who are not interested enough in really great sound to move beyond Full Auto, Full Range.

And anyone who thinks that ‘tuning’ the target curve to map the sound power of the speakers is ideal, has lost the plot.
 
You can’t have it both ways. My statement was clearly true as written by your above admission.

Unless the software keeps each channel together (I.e. the target curve for each channel encompasses the whole channel, not just a limited bandwidth of it with the rest of it stuck elsewhere.

BTW one could add highpass and low pass filters to channel curve components as early as Dirac 1.



They remain accurate to the core functionality, which hasn’t changed. To the extent they were obsoleted, it was by the iOS app that allow one to use the core good parts of Audyssey and ditch the terrible default target curves.



Correct me if I’m wrong, but MultEQ X doesn’t even allow you to shape the target curve directly. You have to use the dumbass abstraction of faux PEQ bands. Therefore, it’s actually a step backwards from the $20 iOS app, which admittedly does need a “copy curve” button.



For different reasons though. The utility of PEQ on HTP-1 (along with Storm, Datasat, etc) is to “precondition” prior to running room correction. One use case is in place of a separate DSP for assisted speakers. Another use case is to take down a a big peak that’s throwing off level/volume calibration before EQ.

Note that I’m writing about actual utility, not what excites people on the internet.



Does it? Per the Geddes and Blind AES paper 5 points generally suffice. If you want to take 32 different measurements of 15 different speakers for one calibration good for you I guess.

I don’t have time or inclination to address the rest point by point. Someone may find them useful and even worth the bizzaroworld consumer-hating license terms. Every product likely finds some fans. However, to me the story of MultEQ X is one of obvious omissions, far too little too late functionality, with unacceptable license terms being the nail in the coffin.
Correct.

Just to add info for MultiEQX. Yes it has that indirect (dumbass) target curve editor but since about two months approximately allows to import filters generated by REW.
 
That, and what little Audyssey does for two subs can be done manually in just a few minutes, with REW and the Audyssey mic if you don’t have a different one.
Oh sure. All those thousands of owners need to get REW and learn it's steep operation curve :facepalm:
Your just plain wrong. The Audyssey method of tuning dual subs may not be perfect, specially in your "not so" humble opinion, but does work very well all by it self. It sounds excellent to me and post measurements with REW look fine.
Maybe, maybe not. It depends. I agree that someone who doesn’t measure will likely imagine an improvement whether or not one exists just because the product says it will.
Nice strawman there.
Jay, your totally over the top Dirac fanboy approach here is getting to the point of being sic. (past it IMHO)
As a reviewer for hometheaterhifi I would expect a more balanced approach but maybe that's part of the problem? No matter what D-M/Audyssey does with the software you always find ways to call it schitt next to your beloved Dirac. I'm honestly beginning to question your objectivity. :mad:
Audyssey is an excellent piece of semi-automated DRC software, not the worlds best but then neither is Dirac, there are much better approaches if you chose to go down those hands-on manual modes. Or call my friend Mitch at Accurate Sound for something really excellent.

In closing this post I would like to remind our members and guests what Amir's opinion on Audyssey using the $20 Editor app was a little over 2 years ago. The bolding is his, not mine.
"Anyway, I told it to upload the data to the AVR which took a bit. Once there, the results were superb! The sound out of my system was now reference quality, pinning me down to my seat listening to track after track from my "audiophile" playlist. Room modes were gone and the deep, floor shaking bass that I expect from my Salon 2 speakers were back with the new target curve. Detail was to die for. Bass was clean. On and on."
 
Back
Top Bottom