Audyssey's Next Generation of Room Correction (MultEQ-X)

[ChrisB]

Has the feature for disabling the rear speaker level boosting with Dynamic EQ been implemented yet? They promised it years ago.. they don't seem to publish good changelogs so gotta ask here.

No, it is not implemented yet.

 

[tjcinnamon]

How many filters would you recommend for the bass - up to 300hz?
I correct the frequency response with REW and import the filters into MultEQ-X. It could be ten filters or 30 filters, dependent how I choose the settings in REW.

I do something similar but I’ve found it to be way easier. I exclude all of the measurements for the speaker I’m correcting. In REW, I do Create Measurment from filters. I save that measurement, then import that to a target curve in MqX. You go from 20 filters to 1 per speaker. It’s much easier to manage.

 

[ChrisB]

I do something similar but I’ve found it to be way easier. I exclude all of the measurements for the speaker I’m correcting. In REW, I do Create Measurment from filters. I save that measurement, then import that to a target curve in MqX. You go from 20 filters to 1 per speaker. It’s much easier to manage.

Seems to be a good way either. The overview in the filter tab in multeq-x is way more structured. But does it create the same results as using filters? Think I will check it with additional REW measurements.

 

[tjcinnamon]

Seems to be a good way either. The overview in the filter tab in multeq-x is way more structured. But does it create the same results as using filters? Think I will check it with additional REW measurements.

I was thinking that myself. I'm wondering what their FIR filter creates relative to the REW filters. I wonder if the REW filters have some "magic" to them where it provides and doesn't provide changes?

 

[Acerun]

I just set up a pair of Revel F226Be speakers with two 15" sealed Rythmik subs and performed a new configuration of MultiEQ-X and then measured and tweaked a little bit with readings from REW. Here are the readings below. I am only applying multieq-x at 500 and below as that is where my room issues reside. On the other hand I thought there was going to be more of a fall off on the higher end of the spectrum but I'm not seeing that. Does anyone have any suggestions for tweaking the filters to provide even better results?

 

[theREALdotnet]

Does anyone have any suggestions for tweaking the filters to provide even better results?
View attachment 373505

The subwoofer integration looks fine, at least the amplitude part of it. I’d be investigating the 8dB difference between 500Hz and 1kHz, that can’t be good.

 

[Acerun]

The subwoofer integration looks fine, at least the amplitude part of it. I’d be investigating the 8dB difference between 500Hz and 1kHz, that can’t be good.

Hmm. Thanks. I suppose I could extend the EQ to 1k or above and see what that does. Would you say that's most likely a room mode that is affecting the frequency response?

 

[theREALdotnet]

Would you say that's most likely a room mode that is affecting the frequency response?

Hard to tell with the smoothing applied. Look at it with 1/12 octave or less smoothing. The group delay graph should also give a clue.

 

[tjcinnamon]

Hard to tell with the smoothing applied. Look at it with 1/12 octave or less smoothing. The group delay graph should also give a clue.

How would group delay show that? Not questioning; trying to learn on my end. I don’t understand group delay.

 

[theREALdotnet]

How would group delay show that? Not questioning; trying to learn on my end. I don’t understand group delay.

Spikes and excursions in the group delay chart usually indicate regions you shouldn’t try to EQ.

 

[tjcinnamon]

Spikes and excursions in the group delay chart usually indicate regions you shouldn’t try to EQ.

What is the reason for that?

 

[theREALdotnet]

What is the reason for that?

See here, section 16 “Minimum Phase”. That explains it much better than I could.

 

[KenMasters]

@Acerun - I've noticed when curtaining with MQX, the corrected region can be raised in relation to the rest of the speaker range.

Open MQX and reconnect, then shift the range of the correction to see how it's affecting the volume level, try checking or unchecking "Disable Auto-Leveling" to see if that improves anything if there is a problem, or see if you can't find a curtain point that better maintains the volume balance.

This for me was an issue and partly because of it I ended up sticking with manual correction through MQX only.

 

[Acerun]

@Acerun - I've noticed when curtaining with MQX, the corrected region can be raised in relation to the rest of the speaker range.

Open MQX and reconnect, then shift the range of the correction to see how it's affecting the volume level, try checking or unchecking "Disable Auto-Leveling" to see if that improves anything if there is a problem, or see if you can't find a curtain point that better maintains the volume balance.

This for me was an issue and partly because of it I ended up sticking with manual correction through MQX only.

By "curtain" you mean the EQ'd frequencies, correct?

 

[KenMasters]

By "curtain" you mean the EQ'd frequencies, correct?

Yes, EQ'd frequency range.

 

[Acerun]

Yes, EQ'd frequency range.

I have found in the past, in my room, that I if I use tilt or a bass boost, I run into trouble if I include the subwoofer for that boost in MultiEQX. I "think" my subs are so good + room effects that I simply don't need the boost there and adding it seems to add bloat. I am playing with that now to try to tame things and find the right crossover and room response. Unclicking the Sub button for the extra EQ, but keeping the LRCS.

 

[Acerun]

Yes, EQ'd frequency range.

Btw, I always disable auto leveling, and add headroom. Looking more carefully at some of my previous measurements there may have been a little bit of user error so I am back and retest. Quick and dirty is looking like 60 may be the best cut over. That was where I crossed my m126be's as well.

 

[KenMasters]

Btw, I always disable auto leveling, and add headroom. Looking more carefully at some of my previous measurements there may have been a little bit of user error so I am back and retest. Quick and dirty is looking like 60 may be the best cut over. That was where I crossed my m126be's as well.

I wasn't referring to your subwoofer or crossover settings, but to the lift from about 1K down. I've noticed MQX can introduce this when curtaining the response, so what I was suggesting is that you look at what the effect is of dragging the curtain across the range.

 

[KenMasters]

Would anyone be able to clarify Atmos enabled speaker distance settings for me.

In the Denon menu you can set distances for individual speakers, but in the case of Atmos enabled speakers there is an additional setting for distance to ceiling. Am I right in assuming this distance is then summed with the physical distance for appropriate delay?

 

[Reverend Slim]

Would anyone be able to clarify Atmos enabled speaker distance settings for me.

In the Denon menu you can set distances for individual speakers, but in the case of Atmos enabled speakers there is an additional setting for distance to ceiling. Am I right in assuming this distance is then summed with the physical distance for appropriate delay?

View attachment 381039

Not summed with the distance from MLP to speaker. They do a basic calculation of what the third length of that triangle (the point sound hits ceiling to the MLP) would be given the assumption of the typical upfiring speaker's baffle angle per Dolby, then use that distance plus the distance to ceiling as the expected arrival time.

Whether they get it right is questionable. It's definitely closer than the first round of Atmos receivers that just assumed 9 foot ceilings and the customary "seated ear level" as a makeshift average, but I would still recommend tweaking that parameter with demo material on a loop to listen for which setting snaps the imaging into place... or better yet, using the Speaker Pairs tests on the Spatial Audio Calibration Toolkit to isolate each upfirer and its corresponding ear-level speaker so you can nail the delays. Upfirers can actually work pretty well if you get them lined up right, but most auto-cal doesn't quite get there.