The problem with Audyssey Dynamic EQ

[KMO]

It's your noise floor, not your speakers.

More specifically it will probably be the noise floor combined with the calibration. As soon as noise dominates, you start seeing the shape of the calibration curve - the calibration curve should only be applied to signal.

Same effect in my measurement, with calibration curve visible for comparison:

Clear the calibration and the rise vanishes - the uncalibrated signal is just a very shallow rise to 27dB at 20kHz

 

[OCA]

If anyone is interested, I have posted a correction technique for Audyssey curves in the conversation below:

https://www.audiosciencereview.com/forum/index.php?conversations/love-you-video.21543/#convMessage-150399

 

[Dj7675]

OK, now we're getting somewhere. It's nice that he responded so fast.

That's not the whole story though. If you're doing high frequency corrections, you don't want to try and fix sharp dips and peaks that are highly mic position dependent. The spatial average keeps you from doing that (has a smoothing effect). Here's a single MQX measurement compared with the aggregate (~25 mic positions or so) with all smoothing turned off:

View attachment 228967

View attachment 228968

You can see the dramatic smoothing effect a good spatial average gets you (and a big advantage MQX has by allowing so many measurements). With the lowest level of smoothing applied:

View attachment 228969

View attachment 228970

You can see they do apply smoothing before calculating the correction filter--which is a good thing, you certainly don't want to try and correct all that hash:

View attachment 228971

It looks like their filter is pretty much based on their middle smoothing setting, 1/12 looks about right:

View attachment 228972

And as we've now confirmed, it is based on the steady state response. Here's a comparison of the MQX aggregate to a MMM using a Umik1 for that speaker at the listening position:

View attachment 228974

The deviations at low frequencies are because I modified the speaker which reduced the sensitivity of the woofer at low frequencies. The deviations at high frequencies track well with the differences in calibration between the two mics (which I have measured and can compensate for in the target curve, but didn't for this particular file). Again I'll mention the number of measurements MultEQ-X allows you to take really helps you get a much better, repeatable spatial average.

OK, now that that's out of the way....


It's not that one measurement is "correct" and the other is not, or that one is "true" and the other is not. You're measuring different things. Both are correct. And the Harman curve is based upon the red line, not the green one. If you correct a windowed response to that curve, you aren't really using that curve.

If you read Toole's book, all the papers, all the research from Harman, every time they refer to an "in room response" they are referring to the steady state response--the red line (unless specified otherwise). The Predicted In Room Response from the spin measurements is predicting the steady state response in a typical room. Good anechoic data can not really be obtained from your listening position in a room. I understand there are some newer/high end room correction programs that do some windowing, but I don't know enough about them to comment on how well they do at that. Historically, pretty much all room correction systems have used steady state measurements, and that's what all the various room curves people pass around are based upon.



Generally, that's the way most like to do it. But it's an oversimplification to say you can't do that with RC that uses steady state measurements. It takes some effort, knowledge and anechoic data from the speaker but it can be used for that purpose (it's just a lot more work that buying a JBL SDP-75 which allows you to load "Anechoic EQ"--which from those I've seen amounts to EQing the listening window flat) before you even start your in-room measurements.

Sorry for getting so off-topic, Matt. But I do think it's a useful discussion.

Thanks for posting. In particular, I like seeing the very similar 25 point measurement vs MMM.

 

[OCA]

If anyone is interested, I have posted a correction technique for Audyssey curves in the conversation below:

https://www.audiosciencereview.com/forum/index.php?conversations/love-you-video.21543/#convMessage-150399

Audiophile experience - Audyssey update

I have updated my target curve to a straight line with a fixed slope as explained in this graph (Harman curve is the second best in green below):

sites.google.com

 

[Jon AA]

Getting back to the subject of this thread, the word is that in an upcoming update to MQX, it will give you the option to use DEQ without changing the levels of the surrounds. Finally!

 

[-Matt-]

Getting back to the subject of this thread, the word is that in an upcoming update to MQX, it will give you the option to use DEQ without changing the levels of the surrounds. Finally!

I hope this is true, if so I may finally buy the Windows software. Or do you expect it to be possible via the app also?

 

[krabapple]

Is this rooted in psychoacoustics? Seems too much of a change to be an honest mistake.

Everyone here should read through all the pages of this Audyssey website thread, where Chris K answered myriad questions about DEQ.


Here's one that's relevant to the surround levels:

https://audyssey.zendesk.com/hc/en-us/articles/212347383-Dynamic-EQ-and-Reference-Level?page=5#comments

Audyssey LabsApril 20, 2012 09:34
Hi Zoltan,
Dynamic EQ is designed to raise the level of the surrounds as you turn down the master volume. This is because our research showed that surround envelopment decreases faster from the back as we lower the volume. So, we need to compensate for that to keep constant surround envelopment.
For this to work as intended, the speaker levels must be calibrated to be equal. If you play the internal test noise after Audyssey calibration is finished you will find that it measures the same as the fronts. That's the starting calibration for reference level. So, when you listen to content at 0 dB on the master volume all speakers will play at the same level. But because that is too loud for most people, when you turn the volume down the surround level will come up to compensate.
The internal test noise is not controlled by the volume control so you will not be able to hear this unless you are playing content.

and another a page or so later

https://audyssey.zendesk.com/hc/en-us/articles/212347383-Dynamic-EQ-and-Reference-Level?page=6&sort_by=created_at#comments

Audyssey LabsAugust 18, 2012 21:48
Hi Tim,
If the speakers are not pointing to the mic, then Audyssey may not read their level correctly. The level adjustment of the surround speakers was based on extensive study of film mixer adjustments as we turned the volume down. Our impression of surround envelopment diminishes faster from the back and that's how we came up with the level adjustment rules.

 

[Chromatischism]

Yes, but many user complaints show they overdid the effect.

If they made it tuneable, like with the Reference Level Offset, it would help.

 

[-Matt-]

I suspect that part of the problem may be room configuration related...

Like many others I have the TV and front speakers at the front of the room and the surrounds near the rear corners. My sofa is (unavoidably) near the back wall (which I'm aware isn't ideal and is advised against by Audyssey.

Whilst I trust that Audyssey have done good research on this, I wouldn't be at all surprised if they have not fully taken into account configurations where the surrounds are MUCH closer to the listening position than the front speakers.

I.e. If they have a listening room set up to establish listener preference for surround boost levels, I bet these tests were done with the speakers approximately equidistant from the listener. The results might have been different if the surrounds were a lot nearer.

I wonder if there is a correlation between those who feel the surround boost is excessive and those who sit nearer to the back of the room?

A lot of the comments from Audyssey Labs in the thread linked above relate to correctly configuring the sound to reference level - all very correct. However, the amount of surround boost to apply is a listener preference based choice which has been imposed by Audyssey (not an established standard). If there is an update to the MultEQ-X software that allows users to choose their own preference for the magnitude of surround boost in Dynamic EQ this will be very welcome.


The other factor that may confuse the issue is that whether or not the surround boost seems excessive can be quite content dependant. I wonder if this indicates that the upmixer is also part of the problem. I.e. When upmixing stereo to the surrounds some content may contain loud sounds at frequencies (or with particular phase content) that causes it to get steered to the surrounds.


Note: The minor annoyance of occasional excessive surround levels isn't enough to stop me from using Dynamic EQ, which on the whole is excellent.

 

[Jon AA]

Whilst I trust that Audyssey have done good research on this, I wouldn't be at all surprised if they have not fully taken into account configurations where the surrounds are MUCH closer to the listening position than the front speakers.

Yes, I think that's probably a large part of the problem. Also, I don't think their research reflects real-world reasons people might listen at reduced volume. When somebody turns down a movie so they don't wake up other people in the house, giving up a slight bit of immersion is probably lower on the list of priorities than being able to understand dialog from the center channel without having to turn the volume too high--something cranking up the surrounds really inhibits.

 

[demoncamber]

It appears that another issue now is those of us with height channels, such as Atmos/DTS:x. It appears to also raise the volume of everything but the front height channels as well.

 

[Chromatischism]

It appears that another issue now is those of us with height channels, such as Atmos/DTS:x. It appears to also raise the volume of everything but the front height channels as well.

Yup, side and rear speakers get more boost, including tops/heights.

 

[Wikedlok]

Overall I'm quite happy with what DEQ does to the frequency response (boosting bass and the high end); but I don't like what it does to the surround channel levels. I just made some measurements with different Dynamic EQ settings to confirm my impression that it boosts the surround channels too much.
(Please ignore the generally lumpy response - this isn't about that and I realised that my minidsp sub settings were messed up part way through).

For the centre and front channels the level of the mid frequencies remains fairly constant with different DEQ settings.
View attachment 223961

View attachment 223958

But for the surround channels it also boosts the overall level...
View attachment 223959

The situation is even worse when listening at lower levels...
View attachment 223960

In the kHz range I think the level should stay roughly where the black line (DEQ off) is. With the DEQ set at 0dB offset it boosts the surrounds by at least 6dB, relative to the fronts.

I assume this is how it is intended to operate (and not an implementation error in my Denon AVC-X receiver). But for me, since I sit a lot closer to the surrounds than the front speakers the boosted surrounds are pretty obnoxious.

For quiet listening in the evening the surrounds are close to me and louder. If I set volume based on the surrounds the centre channel is too quiet - reducing dialogue intelligibility (a common problem). Alternatively if I set volume based on the centre channel I get blasted by loud music and effects from the surrounds, eg when the adverts come on.

Edit: Added poll

I can confirm my sides and rears were ~6db higher than frontstage with deq on. Pulled them down by 6 and finally the focus is correct, I'm right in the bubble. Yet to switch heights to surround posts to measure those and adjust xovers properly.

 

[Chromatischism]

I can confirm my sides and rears were ~6db higher than frontstage with deq on. Pulled them down by 6 and finally the focus is correct, I'm right in the bubble. Yet to switch heights to surround posts to measure those and adjust xovers properly.

Just use the REW SPL meter and pink noise. Better than applying EQ to a speaker it wasn't meant for and getting inaccurate readings.

 

[Wikedlok]

Just use the REW SPL meter and pink noise. Better than applying EQ to a speaker it wasn't meant for and getting inaccurate readings

You mean to say eq is a bad thing?

 

[Chromatischism]

You mean to say eq is a bad thing?

If I read your post correctly, you were going to swap speaker wires to sweep the height speakers. All that will do is apply your surround speaker EQ to the height speakers.

 

[Wikedlok]

If I read your post correctly, you were going to swap speaker wires to sweep the height speakers. All that will do is apply your surround speaker EQ to the height speakers.

Good point, but I have turning audyssey off in mind for those 2 pairs, just to check crossovers. I have no problems with eq quality ( for movies, for music I use other devices and Dirac, through same speakers), just that it sets xovers too low and distance usually good for one pair maybe, while others have big dips there.

 

[Chromatischism]

Good point, but I have turning audyssey off in mind for those 2 pairs, just to check crossovers. I have no problems with eq quality ( for movies, for music I use other devices and Dirac, through same speakers), just that it sets xovers too low and distance usually good for one pair maybe, while others have big dips there.

Ok, so crossovers. I thought you meant you would verify levels that way, which obviously wouldn't work. I use a Dolby Atmos test tones video to cycle through all of my speakers. Usually I need to reduce the surrounds by -4 to bring them in line with the fronts with DEQ.

 

[Benedium]

I feel my Polk S20 responds better to Dyn EQ than my DBR62, especially at lower volumes. Not sure if it is cos brighter and more sensitive (88db vs 86db)? Could be my aging ears too of cos.

 

[Trell]

Yes, but many user complaints show they overdid the effect.

If they made it tuneable, like with the Reference Level Offset, it would help.

That would very nice to have: a Surround Reference Level Offset.

I use Dynamic EQ for music that has mixing/mastering of multichannel all over the map, and I find DynEQ to be way too aggressive for surrounds here.