Audyssey Room EQ Review

[HarmonicTHD]

That seems reasonable, but if I want to create such a target curve based on Harman's findings on the preferred tilt similar to the one Amir did in one of his reviews, then I would want to consider the slope of the tilt vs my listening level(s). And if I decide not to use DEQ, then I would likely have to do at least two target curves, one has a steeper slope than the others.

Audyssey Room EQ Review

This is a review and detailed measurements of the Audyssey Room Equalization. Audyssey is standard technology in a number of Audio/Video Receivers such as the Denon AVR-X3600H I recently reviewed. The standard version in the AVR is mostly all or nothing so I downloaded the Marantz/Denon...

www.audiosciencereview.com

Given the info you have provided in the link, I fully agree with you on this, but I would say that at least logically speaking, I may (in practice I likely wouldn't bother..) indeed require a modified curve if I typically listen to 70 dB average (vs the 78-80 dB you cited) level or below, unless I can find a way to balance thing off by using DEQ.



Widely? Not on ASR, I hope not..



I don't think anyone has suggested that the so called Harman curve has connection to THX standard. To me, if I want to use such a curve I would definitely consider creating more than one curve, or find a way to use it with DEQ on top effectively and that would require some trial and error, for reasons I explained above. That's just my opinion based on what I know now about the so called Harman curve, Audyssey DEQ, and the logic. It seems to me your "complete red herring" remark is related to something different.

As you most likely know, Audyssey's DEQ algorithm is not based on a fixed "loudness curve".

Audyssey MultEQ Room Correction Interview With Chris Kyriakakis

The Audioholics interview Chris Kyriakakis in order to learn the ins and outs of Audyssey's MultEQ room correction / auto-calibration system. Bass management, calibration, EQ and more.

www.audioholics.com

As for your intended preference curve definition. That is similar to what I keep in the Denons “Reference” storage with the only difference that my tilt starts a bit earlier ( ca 1 or 2 k) but that’s just my room rolloff. I use MultiEQX which gives a bit easier control to the preference curve definition (yes at a price).

 

[HarmonicTHD]

Have you tried lowering the surround channel levels by a couple dB? I have been living with the seemingly elevated (to the point of annoying..) surround channel levels until recently when someone somewhere suggested to lower the surround level trims by 3 dB. I have no idea why I didn't think of such a simple solution. So far it works well, for me anyway. I am quite sure the elevated level effects weren't there prior to 2016 or 2017 so D+M models such as those prior to X4400H, SR7012, AV8802 would not have such issue.

Nah. But I also heard about people (incl Audioholics) saying that the automatic measurement lead to 2 to 3 dB too hot surrounds. I left it unchanged as it was ok in my room. Or maybe Audyssey changed something with one of the several recent firmware updates and made it less hot?

 

[HarmonicTHD]

So coming back to the question from Peng, at what sound levels the Harman studies had been conducted (even if it not relevant here so just out personal interest and not implying anything).

I dug out Toole’s book. The best I could find was Chapter 3.5.1.6:
"Olive et al. (2013) used program at an average level of 82 dBC, slow. This was for an audience of serious listeners doing foreground listening of a demanding kind”

Disclaimer: that is not to say that all the studies were indeed conducted at that level as the quotation leaves room for interpretation and it could be that there are other references in the book which I overlooked and and and … I hope I caught all eventualities ;-)

 

[Sal1950]

I dug out Toole’s book. The best I could find was Chapter 3.5.1.6:
"Olive et al. (2013) used program at an average level of 82 dBC, slow. This was for an audience of serious listeners doing foreground listening of a demanding kind”

Assuming this is correct, then looking at the F-M curves would support the statement that DEQ would have very little effect on levels above 80 db. Again assuming that DEQ somewhat follows F-M guidelines? But then DEQ is kind of in it's own world. I've never understood or agreed with the choice for boosting rear channel levels at low levels. It's effect has been complained by users almost universally?

This is the more modern ISO 226 from the International Organization for Standardization, which are based on a review of modern determinations made in various countries.

 

[HarmonicTHD]

Assuming this is correct, then looking at the F-M curves would support the statement that DEQ would have very little effect on levels above 80 db. Again assuming that DEQ somewhat follows F-M guidelines? But then DEQ is kind of in it's own world. I've never understood or agreed with the choice for boosting rear channel levels at low levels. It's effect has been complained by users almost universally?

This is the more modern ISO 226 from the International Organization for Standardization, which are based on a review of modern determinations made in various countries.

Thx.
Yes ISO seems much “steeper” down low than FM. Mhh - I had read about FM (guess where ;-)) but never looked closer.
Can’t say much about DEQ either.

 

[GalZohar]

As far as I remember at 20Hz, DEQ boosts back bass by 2.6db for each 5db of overall level reduction from reference. So each db reduced at 1KHz reduces 20Hz by about half a db. So yes, the above graph is steeper as it seems to reduce only about 0.3db at 20Hz per each 1db reduced at 1KHz.

 

[Chromatischism]

Assuming this is correct, then looking at the F-M curves would support the statement that DEQ would have very little effect on levels above 80 db. Again assuming that DEQ somewhat follows F-M guidelines? But then DEQ is kind of in it's own world.

It is. Here's what it does. Ignore the scale on the left.

As you can see, it only "stops" altering the curve at MV = 0.

 

[Sal1950]

It is. Here's what it does. Ignore the scale on the left.

Is that a actual Audyssey DEQ graph?
I see it says "Dynamic EQ" on top but?
Just askin,
Thanks! Very Kool to see.

 

[Chromatischism]

Is that a actual Audyssey DEQ graph?
I see it says "Dynamic EQ" on top but?
Just askin,
Thanks! Very Kool to see.

It was reportedly measured from the preouts of an AVR.

 

[markus]

It was reportedly measured from the preouts of an AVR.

That's correct as I did these measurements

 

[FrantzM]

Hi

Coming from a reformed subjectivist... At this point in time, I can't imagine a system without DEQ or something similar. I have not yet purchased the PC APP. IDEQis set with the Audyssey MultiEq mobile APP. The darn thing works and very well ... to my ears .. So much that I haven't cared to measure what it does... I will defer to this graph, on faith . My subjective impressions, exactly

It is. Here's what it does. Ignore the scale on the left.

As you can see, it only "stops" altering the curve at MV = 0.

View attachment 214639

 

[peng]

My whole point is, if you use a fix target curve (similar to those so called Harman curve with the bass tilt) based on say 70 dB based on a test tone (and I am still referring to the bass range only), love the effect at the same volume level, then if you listen at levels significantly different than 70 dB (again based on test tone, or volume position), you may still love the effect of that same target curve vs a flat one, but you may perceive significantly different bass results.

That has nothing to do with THX (never suggested it has either..), or whether Harman did their tests at 78 dB or 85 dB, but if one uses Audyssey without DEQ on and listen to levels well below THX reference level such probably most people do, one will complain Audyssey killing their bass whereas an Anthem (the latest models) user would likely say Anthem's AVRs/AVPs are superior in bass performance vs Denon/Marantz/Yamaha's. That is not to say Anthem AVRs or Anthem ARC are not superior overall, I am just using this point as an example of one reason why I suspect at least in some cases, the often reported Audyssey thin sound might have been due to the flat default curve that can now be fixed easily with the App(s). Again, I know to some or many who are familiar with the topic, I am likely stating the obvious..

 

[GalZohar]

I think anyone using the default Audyssey target curve is not qualified to judge its performance against other DRCs based on his listening experience. It makes no sense to compare one DRC with a bass-flat target curve with another that has elevated bass target curve. Even more so if DEQ was turned off for the comparison.

 

[Sal1950]

I think anyone using the default Audyssey target curve is not qualified to judge its performance against other DRCs based on his listening experience. It makes no sense to compare one DRC with a bass-flat target curve with another that has elevated bass target curve. Even more so if DEQ was turned off for the comparison.

What is it your saying the "default" Audyssey target curve is?

 

[GalZohar]

What is it your saying the "default" Audyssey target curve is?

There are 2 curves: Flat and Reference. Flat curve is flat and reference is still flat in bass, with midrange compensation 3db dip at 2KHz and some high frequency rolloff.
Reference curve can be modified quite extensively with the 20$ mobile app (although interface is crappy and accurate modifications are easier to do with 3rd party apps). Both curves are configurable with what seems to be a good interface with the 200$ PC app.

 

[Sal1950]

There are 2 curves: Flat and Reference. Flat curve is flat and reference is still flat in bass, with midrange compensation 3db dip at 2KHz and some high frequency rolloff.
Reference curve can be modified quite extensively with the 20$ mobile app (although interface is crappy and accurate modifications are easier to do with 3rd party apps). Both curves are configurable with what seems to be a good interface with the 200$ PC app.

Yes, all totally correct.
Don't most other DRC's aim for a flat bass response? That's the part of your post that confused me.

OTOHI don't believe any 2 DRC software's output can be compared at face value. Each uses their own logarithm to compute the final result. Not to mention the variables induced from mic positions, etc.

 

[GalZohar]

Anthem (and I think Dirac as well?) by default use some kind of a bass boost. In Anthem the default bass boost values are determined by the measurements (probably tries to follow your measured room gain, as they describe in their documentation), and you can also manually set tilt and additional deep bass boost for a lower frequency - Refer to Anthem website for more details. Basically the default Anthem curve looks very similar to the Harman style curves, except the tilt seems to need to be set manually and defaults to flat. It also has high frequency rolloff, although by default it doesn't EQ above 5KHz anyway and I'm not sure if the default rolloff is just some fixed value or determined by the measurements.

From several comparison posts on these forums, it seems like the main differences are in the target curve and interfaces. The EQ ability seems actually quite similar, and averaging differences, if significant, probably don't really matter if you focus them around 1 listening position. Of course in the higher end systems you have more features such as Dirac DLBC (seems only available on quite expensive AVRs/AVPs and at an extra cost), not to mention Lyngdorf/Trinnov (which are also very expensive).

 

[Sal1950]

Anthem curve looks very similar to the Harman style curves, except the tilt seems to need to be set manually and defaults to flat.

So Anthem offers optional "house curves", but still will default to flat.
Whatever, in the end we still agree you can't in any way compare the sound of any 2 DRC software since they're all going to be messing with fr, time, etc.. That's what they do. LOL

 

[GalZohar]

So Anthem offers optional "house curves", but still will default to flat.
Whatever, in the end we still agree you can't in any way compare the sound of any 2 DRC software since they're all going to be messing with fr, time, etc.. That's what they do. LOL

I think you can compare but must apply equal target curves or else you are comparing target curves rather than the actual EQ capability.

Anthem default curve includes room gain bass rise which is determined from measurements. Not sure why they don't also auto-determine the default tilt.

 

[amansangar]

Any tips on mic position if listening position is against back wall? Audessey says to measure at least 50cm away from the back wall but I’m planning on buying some acoustic panels for the back wall. Should I still measure 50cm away with and without panels?