Room EQ Comparison

[AJ Soundfield]

Would there be interest in an objective and subjective analysis of Audyssey XT32 and Dirac Live?
Target curves would be made identical. Measurements would be taken before and after.
Goal is to see which achieves smoother results and which gets closest to the desired target.
Then, instant switching to determine if the differences are audible and to relate them to the graphs.

It has the makings of very entertaining reading, since this sort of thing is prevalent on so many forums...and is widely/wildly misunderstood.
Out of curiosity, are you interested in the scientific perspective? This paper is open access and can be read free. It's a useful start for countering the usual anti-science which has already appeared.

cheers

 

[AJ Soundfield]

Since you are probably using two different AVRs, you need first to A/B them in a blind test to see if their differences are, effectively, null.

Premise being no single unit has both Audyssey XT32 and Dirac Live?
Forgive my ignorance, I'm largely oblivious to the whole "EQ for Dummies" series.

cheers

 

[AJ Soundfield]

I think it would be cool to compare them head to head. I've compared them subjectively, but not with instant switching.

That last part is very important. So is this last sentence

cheers

 

[Ethan Winer]

Ethan, very interesting test and article, thanks for posting it here. It does not surprise me that in a room with already-extensive treatment, Dirac can do more harm than good at places other than the sweet spot. I use Dirac myself, in a small and non-ideal room with the speakers too close to walls and corners, mainly for the purpose you say that EQ systems can be useful for: taming bass peaks. But I let it work full range, because I also find the high frequencies to be improved overall in my system.

If you're happy I'm happy. But I also tested Dirac in a small untreated empty room with the same results. Also, my home studio is not extensively treated. It's a very large general purpose room with relatively few bass traps given its size. That EQ makes the response worse in some places is a constant that can't be avoided.

Also, since you have Dirac adjusting mid and high frequencies, I'm curious if you have even minimal absorption at the side-wall reflection points. If not, that will do far more to improve clarity than any EQ setting.

 

[Kal Rubinson]

Premise being no single unit has both Audyssey XT32 and Dirac Live?

Yup.

 

[Eric SVL]

I have a Marantz AVR connected to a MiniDSP 88A. The Marantz has XT32, while the 88A is now running Dirac Live 3.0 (without DLBC). a comparison would use the same speakers in the same room, and the comparison could be based on a one-position calibration for each, without moving the mic. REW comparison measurements could then be taken at the MLP, again without moving the mic. Turn on Audyssey in the Marantz while the 88A is in bypass mode, then turn off Audyssey and turn on Dirac Live in the 88A.

The only think that alludes me would be how to set target curves for both room correction systems that match. No clue how this would be done on my Marantz 8802a, which does not support the new Audyssey app.

Right. You need the capability of the app and in addition, the companion program created by Ratbuddy that basically puts the curve editing abilities on par with Dirac. You would be able to make the same target curve for both and then test the results.

Regarding bass traps, which are not really the topic of this thread, I usually only use EQ in the modal region. It does a miraculous job there, whether it is Audyssey or Dirac – much better than bass traps can. And, it costs less and takes up no space. However used together you can improve on your results further. What is questionable is correcting the higher frequencies, and any attempt at that is only valid if tracing the speaker's natural in-room response. You want to be doing room correction, not speaker correction. Well, you might want to do speaker correction, depending on your speakers. There is nothing wrong with that, but it's not the goal here.

 

[Eric SVL]

Room variance. The tests would have to be done in a number of different listening rooms, otherwise the results would be potentially idiosyncratic and non-indicative. Ideally, also with a number of different types of loudspeaker, at various different positioning locations.

This isn't going to be possible. Testing would be conducted in one controlled environment.

 

[AJ Soundfield]

What is questionable is correcting the higher frequencies, and any attempt at that is only valid if tracing the speaker's natural in-room response. You want to be doing room correction, not speaker correction. Well, you might want to do speaker correction, depending on your speakers. There is nothing wrong with that, but it's not the goal here.

Excellent, someone is paying attention to science! Bravo.
Yes, "speaker" correction is ok IF NOT done from "room response". It is the quasi-anechoic design distance (allow driver integration and some baffle diffraction loss, so 1-2m ish) response that can be massaged, judiciously with care for the global, not just on axis response.
I swore I saw a thread this morning with you doing the opposite...

cheers

 

[Eric SVL]

Excellent, someone is paying attention to science! Bravo.
Yes, "speaker" correction is ok IF NOT done from "room response". It is the quasi-anechoic design distance (allow driver integration and some baffle diffraction loss, so 1-2m ish) response that can be massaged, judiciously with care for the global, not just on axis response.
I swore I saw a thread this morning with you doing the opposite...

cheers

My bass trap results? I'm only using Audyssey to 400 Hz there.

 

[AJ Soundfield]

My bass trap results? I'm only using Audyssey to 400 Hz there.

So the last graph with "Full XT32" is not correcting HF? I misread as.

 

[Eric SVL]

By full I meant it's a full range measurement, since all of the others were only looking at the bass region. I'll make a note to clear up the confusion. Thanks.

 

[AJ Soundfield]

The confusion was mine, you're good! ;-)

 

[Eric SVL]

After seeing Matthew's excellent discussion on the merits of Room EQ last night, I'm less inclined to perform this test. Not because I doubt I can do it, but because I question whether it makes sense to do it. I guess it could still be valid in the bass region.

 

[AJ Soundfield]

You're no fun.

Link to Matt's discussion?

 

[Eric SVL]

I thought about making a new thread, but maybe I'll just continue here.

The discussion I'm referencing with @Matthew J Poes was posted by Audioholics here:

Part 1:

.

Part 2:

(by the way, @Todd Anderson , the Insert Media window has a bug – the OK button displays "Read More", see attached)

I've shared this link on other forums and the reaction has been mixed. Most who follow Floyd Toole's work recognize the overarching theme, which he described in his book and which we discussed with him directly on AVS. It's great to see someone put the theory to the test and measure the listening window with EQ applied. That said, Dirac supporters (and Dirac themselves) claim that the corrections they make lead to audible benefits even if you can't see it in the frequency response. So far, that claim has been made because Dirac can use all-pass filters to adjust the phase response without altering the frequency response. Being skeptical of whether we can hear absolute phase, I am unsure how this benefits us.

I would like to see more discussion on this, and I'll offer some thoughts to get things going:

• Are we measuring the right things? Could the benefits of Dirac Live be shown by measuring a single speaker? Or, do we need to measure the interaction of speaker pairs to see the benefits of the impulse adjustments?
• Or, is this all marketing?

 

[AJ Soundfield]

That said, Dirac supporters (and Dirac themselves) claim that the corrections they make lead to audible benefits even if you can't see it in the frequency response. So far, that claim has been made because Dirac can use all-pass filters to adjust the phase response without altering the frequency response. Being skeptical of whether we can hear absolute phase, I am unsure how this benefits us.

I would like to see more discussion on this

There is nothing to discuss. That is classic "faith based" audio.
With "Toole's work", there are the results of world wide scientific, controlled listening tests..which can be discussed.
The others are "I heard this, I said that", usually under the facade of sciency looking graphs. But zero valid, controlled listening tests.
You might as well discuss folks hearing Santa Claus.

That said, as I mentioned, it still would have been a fun read to see what your real world exercise of these type products came up with. I doubt anyone was expecting an AES submittal. Have some fun

cheers

 

[Todd Anderson]

I thought about making a new thread, but maybe I'll just continue here.

The discussion I'm referencing with @Matthew J Poes was posted by Audioholics here:

Part 1:

.

Part 2:

(by the way, @Todd Anderson , the Insert Media window has a bug – the OK button displays "Read More", see attached)

I've shared this link on other forums and the reaction has been mixed. Most who follow Floyd Toole's work recognize the overarching theme, which he described in his book and which we discussed with him directly on AVS. It's great to see someone put the theory to the test and measure the listening window with EQ applied. That said, Dirac supporters (and Dirac themselves) claim that the corrections they make lead to audible benefits even if you can't see it in the frequency response. So far, that claim has been made because Dirac can use all-pass filters to adjust the phase response without altering the frequency response. Being skeptical of whether we can hear absolute phase, I am unsure how this benefits us.

I would like to see more discussion on this, and I'll offer some thoughts to get things going:

• Are we measuring the right things? Could the benefits of Dirac Live be shown by measuring a single speaker? Or, do we need to measure the interaction of speaker pairs to see the benefits of the impulse adjustments?
• Or, is this all marketing?

Thanks

 

[DanDan]

Hi Ethan, welcome. To provide some balance it think it fair to point out that your Dirac investigation and publications is highly flawed in my professional opinion as an Acoustician and as a daily user of DL in my career as a Legendary Recording Engineer. Some of your generic assertions as to only localised benefit are just simply wrong. In well known reality Eqing the lowest mode softens the peaks, shallows the nulls, and shortens the Decay THROUGHOUT the room. The amount of mitigation you claim as only a few %, I have tested up to 25%, a member here has shown nearly 50%, I will ask him to join us. The spatial robustness of the LM EQ will diminish as frequency increases. BUT tonal shaping over a broad band or shelf is also effective THROUGHOUT the room. e.g. typical 1/2 space LF Eqs and HF tailoring. Anyone with with a little knowledge can make any system fail to deliver, particularly when by asking it to do things it cannot. Similarly anyone could easily test physical treatments and show them to operate in only a limited area too. e.g. RFZ. Or to have too much of a size requirement to address the 30-40Hz region.

EQ Does Improve Bass Decay

EQ Does Improve Bass Decay I wanted to share some examples of how EQ does reduce ringing in the bass, or rather, improves bass decay. Why? Because it's minimum phase and it must. I'm including both simulations of the EQ effect, as generated in REW, as well as an actual measurement using DIRAC...

www.avnirvana.com

 

[sfdoddsy]

I think Ethan is being a bit disingenuous by not being up front about his commercial bias.

That aside, for me the test is quite simple.

The easy bit is whether the Room EQ does what it says it does.

Basic measurements will reveal this.

the tricky bit is whether this is an improvement.

The best example of this is the original Audyssey XT.

Claims were made about working in the time domain, but measurements revealed only an effect on frequency response.

And, because it only worked full range and had an undefeatable target curve dialling in a treble boost, it made most speakers sound worse.

But at least it did something. Measurements I made of competitors such as Yamaha YpAO orMCACC showed no measurable effect at all.

More sophisticated and customisable EQs like ARC, Dirac, Trinnov, Lyngdorf and DEQX allow one to be more selective.

I’ve used ARC since it first came out. The changes are easily measureable, and when used below the Schoeder invariably a subjective improvement. But they are not close to the flat perfection Anthem claim.

I’d love to see rigorous measurements before and after Room EQ to compare to the optimistic projections.

 

[AustinJerry]

I’d love to see rigorous measurements before and after Room EQ to compare to the optimistic projections.

I am running Dirac Live 3.0 room correction. While I am not sure if these measurements fit your definition of "rigorous", I thought I would share a couple of results.

Dirac Live claims to make improvements in the time domain. The REW impulse response shows that there are indeed significant improvements:

And here is the measurement of the combined sub channel:

 

[tesseract]

@sfdoddsy -

//The best example of this is the original Audyssey XT.

Claims were made about working in the time domain, but measurements revealed only an effect on frequency response.

And, because it only worked full range and had an undefeatable target curve dialling in a treble boost, it made most speakers sound worse.

But at least it did something. Measurements I made of competitors such as Yamaha YpAO orMCACC showed no measurable effect at all.

More sophisticated and customisable EQs like ARC, Dirac, Trinnov, Lyngdorf and DEQX allow one to be more selective.//


I have no experience with Yamaha or Pioneer or Anthem DSP, but am onboard 100% with this assessment of the others and came to that conclusion just from listening!

Audyssey has never grabbed me. I wasn't able to take full advantage of Trinnov in my home due to the wonky implementation Sherwood/Newcastle bass management, but still heard soundstage and imaging improvements. Dirac, Lyngdorf and DEQX are top notch, even though DEQX does the "dreaded" speaker correction, THEN room correction.

 

[Eric SVL]

Dennis, have you ever used XT32 with the MultEQ app?

 

[tesseract]

Dennis, have you ever used XT32 with the MultEQ app?

I have never owned an Audyssey unit. I have many hours of flight time in the Captain's Chair of various Audyssey-based systems, in several rooms. Legacy Audyssey DSP, newer XT32 included.

 

[bearr48]

Can somebody help me out here? I've been playing with REW and the Audyssey App, not very experienced yet with either. I know how to make a copy of a curve with the app, but I don't know how to give the copy a new name. When and if I get smarter, I'd like to download and use Ratbuddy, but that's still out of my league. Thanks much!

 

[bearr48]

I finally figured out how to re-name a copy. Never did see any pop-up or dialog box instructing. Wish there was a "Audyssey MultEq App for Dummies" tutorial, cause it seems very unintuitive. Thanks all!