Potentially stupid questions about Audyssey and target curves

[Jbrunwa]

Ok, I've tried to read a lot and watch videos about Audyssey room correction. There are still some things I don't understand about Multi EQ and Audyssey and I hope that someone more knowledgeable can explain. If the answer is read RTFM, and there is such a thing that I just haven't come across, then I apologize:

1. The BBC dip - if I understand what they say, supposedly to remove phase distortion at a crossover centered around 2k. Don't good speakers already design this correctly, and I'm sure that not all speakers have the same crossover, so why is this applied across the board as an assumption?

2. The target curve - to the best of my knowledge, musicians and producers mix their music at the source to produce a good sounding curve that rolls off high frequencies. Then Audyssey builds in a curve that rolls off high frequencies. So doesn't that depress the high frequencies twice? Why not adjust the EQ to be completely flat, and then the speakers will reproduce whatever curve the producer created? What am I missing?

3. Forgetting the mid/high frequencies, and focusing on bass correction, at least in my room, there are some low frequency dips/dead spots, like for example in my room 38 Hz and 70 Hz. Multi EQ shows that these have been magically corrected in the target curve. Yet when I measure the response with REW the dips are still there in the same place and amplitude that Audyssey originally measured. So the target curve is just a target, but it never shows the actual improvements it thinks it can make. Is there a way to find out what EQ has been applied like a parametric EQ would use?

4. Dynamic equalization - since Multi EQ only measures the frequency response at a single volume level, how does it know how much equalization to apply at different volume levels. Do all speakers effectively have the same difference in response by sound pressure level or is this somehow based on people's hearing at different sound levels and variation of speaker brands and models not important?

Thanks in advance

 

[Melvinne]

This thread is one of the best on dialing in custom target curves for Audyssey, I used it to apply a custom curve to my setup. I used the spreadsheet linked in the thread (DeLub?) to apply one of the house curves to the output. Get an idea of how it works, tweak it (I use iCloud to share the files between my Mac and iPad) and you’ll find it to be great.

If you have REW, use the subwoofer distance to adjust your subwoofer phase with the main speakers, that fixed the dip I had right at the crossover frequency, and significantly evened out my frequency response.

1. I disable MRC on all curves, it’s pointless imho.
2. See the thread I linked
3. You have REW, so find where you are getting nulls/cancellation from your crossover points.
4. I have it enabled, I find it gives a nice umph on lower listening levels.

By the way, my in-room response via REW w the USB mic everyone uses (can’t recall the name - miniDSP?). That’s using only Audyssey, after applying my custom target curve.

 

[Jbrunwa]

This thread is one of the best on dialing in custom target curves for Audyssey, I used it to apply a custom curve to my setup. I used the spreadsheet linked in the thread (DeLub?) to apply one of the house curves to the output. Get an idea of how it works, tweak it (I use iCloud to share the files between my Mac and iPad) and you’ll find it to be great.

If you have REW, use the subwoofer distance to adjust your subwoofer phase with the main speakers, that fixed the dip I had right at the crossover frequency, and significantly evened out my frequency response.

1. I disable MRC on all curves, it’s pointless imho.
2. See the thread I linked
3. You have REW, so find where you are getting nulls/cancellation from your crossover points.
4. I have it enabled, I find it gives a nice umph on lower listening levels.

Thanks for sharing the link and tips. I also have iPad and Mac and using iCloud to save the files works great.

When using the subwoofer distance to adjust the phase, can you explain in more detail? Are you actually moving the subwoofer, or just adjusting the distance to the subwoofer in Multi EQ or manually editing the file? My subwoofer amp only has 0 or 180 degree settings

 

[Melvinne]

Thanks for sharing the link and tips. I also have iPad and Mac and using iCloud to save the files works great.

When using the subwoofer distance to adjust the phase, can you explain in more detail? Are you actually moving the subwoofer, or just adjusting the distance to the subwoofer in Multi EQ or manually editing the file? My subwoofer amp only has 0 or 180 degree settings

I just adjusted the speaker distance in manual settings on the AVR, though you can do it in MultEQ and resend the setup each time if you want. It’s easier to do with the remote while rerunning the sweeps in REW. You aren’t moving the subwoofer at all, it’s just changing the distance which means the AVR reduces the delay to the sub which alters the phase. It’s in the pages and pages of detail but a simple adjustment vs using the phase knob/toggle on the subwoofer. And some would say more accurate since it’s adjusting the delay.

 

[Jbrunwa]

I just adjusted the speaker distance in manual settings on the AVR, though you can do it in MultEQ and resend the setup each time if you want. It’s easier to do with the remote while rerunning the sweeps in REW. You aren’t moving the subwoofer at all, it’s just changing the distance which means the AVR reduces the delay to the sub which alters the phase. It’s in the pages and pages of detail but a simple adjustment vs using the phase knob/toggle on the subwoofer. And some would say more accurate since it’s adjusting the delay.

Thanks for explaining

 

[rdenney]

As I understand it, the target curve is not a target at all, but an outcome of speakers with flat on-axis frequency response combining with early reflections that have a downward spectral tilt (because highs are never as widely disbursed as lows, and are thus weaker in the large angles off-axis related to early reflections). If that’s the case, you have to have a sense of those early reflections before adopting the target. Rooms with few early reflections will provide less downward tilt, but that’s as it should be.

Assuming the Harman target suggests that the “typical” in-room reflections are like your room, and that may not be.

I think the best strategy is to buy speakers with flat on-axis anechoic frequency response and good directivity, and then don’t mess with EQ above a few hundred Hz unless you have reliable anechoic data.

Rick “a gozoutta, not a gozinta” Denney

 

[Jbrunwa]

As I understand it, the target curve is not a target at all, but an outcome of speakers with flat on-axis frequency response combining with early reflections that have a downward spectral tilt (because highs are never as widely disbursed as lows, and are thus weaker in the large angles off-axis related to early reflections). If that’s the case, you have to have a sense of those early reflections before adopting the target. Rooms with few early reflections will provide less downward tilt, but that’s as it should be.

Assuming the Harman target suggests that the “typical” in-room reflections are like your room, and that may not be.

I think the best strategy is to buy speakers with flat on-axis anechoic frequency response and good directivity, and then don’t mess with EQ above a few hundred Hz unless you have reliable anechoic data.

Rick “a gozoutta, not a gozinta” Denney

Thanks Rick,

I have been weighing whether to upgrade my LCR speakers or to find a subwoofer or sub amp with EQ, but not if I can get the current equipment to my liking.

Here is what I am seeing with REW with the minidsp mike before and after Audyssey for the subwoofer and right front speaker. I wouldn’t say that my speakers have high directivity, they design attempted to minimize off axis hf drop. My room has a vaulted ceiling nd is open on one side, so not ideal. I have Multi EQ filter frequency set to about 600 hz and turned off mid range correction. I only measured 20-200 for the LFE before Audyssey but measured 0-200 after. LFE crossover is at 60 for sub and LCR.
I would be interested to hear any additional pointers.

EDIT - Fixed incorrect graph image

 

[dasdoing]

2. The target curve - to the best of my knowledge, musicians and producers mix their music at the source to produce a good sounding curve that rolls off high frequencies. Then Audyssey builds in a curve that rolls off high frequencies. So doesn't that depress the high frequencies twice? Why not adjust the EQ to be completely flat, and then the speakers will reproduce whatever curve the producer created? What am I missing?

the thing is, a perfectly flat speaker desn't produce a flat meassurement in a room.
now, the generalized Harman curve is actualy only valid in the Harman room.
the best way to go around this problem is to EQ only direct sound above 1000Hz-ish, and EQ the room below this.
check out this power point http://www.aes-media.org/sections/pnw/ppt/jj/room_correction.ppt

 

[hex168]

My understanding is that you can use REW appropriately gated to see your speaker's anechoic response, above a few hundred Hz, anyway. Any difficulty with measuring anechoic in room arises at lower frequencies, where we want to EQ in-room response anyway. Is that an oversimplification?

 

[Melvinne]

I would return the subwoofer LFE setting at 120 Hz in the AVR, as that is the LFE channel (.1) in the soundtrack. The Small/Hz setting for each speaker will _additionally_ direct any signal below the crossover to the subwoofer, regardless of the LFE channel setting. The LFE channel is not mixed into your L/R speakers when a subwoofer is present.

And your REW data shows some pretty significant humps (10db swings) in the bass region, I'd run through it a few more times and maybe adjust the subwoofer position to avoid room modes from causing an uneven response.

Regardless of all the statements on response, in-room, etc. if you're measuring it with calibrated devices you're on the right path. What happens in your room with your speakers is what matters.

Also, if you find yourself trying to EQ your response curve into a better place, and don't see a resulting +3db in response with a +3db in EQ, you're likely dealing with cancellations, nulls, or room modes and should try to adjust positioning to compensate.

 

[Jbrunwa]

the thing is, a perfectly flat speaker desn't produce a flat meassurement in a room.
now, the generalized Harman curve is actualy only valid in the Harman room.
the best way to go around this problem is to EQ only direct sound above 1000Hz-ish, and EQ the room below this.
check out this power point http://www.aes-media.org/sections/pnw/ppt/jj/room_correction.ppt

Thanks. This powerpoint is written at a good level to help me understand how it works and what I need to do

 

[Jbrunwa]

I would return the subwoofer LFE setting at 120 Hz in the AVR, as that is the LFE channel (.1) in the soundtrack. The Small/Hz setting for each speaker will _additionally_ direct any signal below the crossover to the subwoofer, regardless of the LFE channel setting. The LFE channel is not mixed into your L/R speakers when a subwoofer is present.

And your REW data shows some pretty significant humps (10db swings) in the bass region, I'd run through it a few more times and maybe adjust the subwoofer position to avoid room modes from causing an uneven response.

Regardless of all the statements on response, in-room, etc. if you're measuring it with calibrated devices you're on the right path. What happens in your room with your speakers is what matters.

Also, if you find yourself trying to EQ your response curve into a better place, and don't see a resulting +3db in response with a +3db in EQ, you're likely dealing with cancellations, nulls, or room modes and should try to adjust positioning to compensate.

Thank you. I mis-spoke on the LFE crossover. Its set to 120 with the LCR crossover set to 60. I will need to do some more experimentation on the subwoofer. Also I do have an extra old Polk Audio subwoofer that I’m not using, I may try adding that to the mix to see if it helps or hurts.

 

[Jbrunwa]

I realized that Dynamic EQ was on. I know that some recommend using it, but im getting much better sound with it turned off. So far I haven’t been able to solve the sharp dips in bass response so I guess they are room nulls.

 

[pedrob]

I asked this in Denon AVR-X8500H AVR Review, but here might be better.

Dolby Atmos Response Curve