Please help me integrate my subwoofer

[kifeep]

I'm going a little crazy trying to integrate my subwoofer. I managed to find a pretty good location for it, and here is my LF response after EQ:

LF response from the mains is also not too bad:

The trouble is that the FR gets worse no matter where I set the crossover:

I imagine this is a a timing issue, but I set the subwoofer timing as accurately as I could:

Also, as a side note, I see impulse variation between runs, with no changes on my part. Is this normal?:

Can you help me understand what's going on here? I'd also appreciate any help fixing the problem.

Here is a link to the mdat file: https://www.dropbox.com/scl/fi/ytoh...ion.mdat?rlkey=fvlnr7tfi5jfyqc8o8u0r2s7h&dl=0

Thanks.

 

[GaryY]

Have you tried to reduce the gain of sub a little bit or change the slope of cut-off?

 

[Chrispy]

What gear is involved and how is eq being applied? Are measurements a single point or multi or moving mic or ?

 

[kifeep]

Thanks for weighing in.

Have you tried to reduce the gain of sub a little bit or change the slope of cut-off?

Sort of. The sub gain was too high at first and I adjusted it down to better match both the target and the SPL of the mains. Before adjusting, the FR was even worse when crossed over. I suppose I could lower it more, but I don't see how I'll be able to hit my target curve if I do. I don't have control over the slopes.

What gear is involved and how is eq being applied? Are measurements a single point or multi or moving mic or ?

Denon AVR-X3700H, 3x Philharmonic BMR monitors (F3 around 34hz), Rythmik F12SE.

EQ was applied by @OCA's Audyssey hack (essentially generate MP corrective target curves in REW and load them into an ady file).

The FR graphs above all include both a single point sweep and a MMM measurement for each setting.

 

[Chrispy]

Might be easier to add a sub or two What is your goal here?

 

[terryforsythe]

Can you help me understand what's going on here? I'd also appreciate any help fixing the problem.

Are you adjusting the time delay/phase of the subwoofer to ensure that it is in phase with your speakers at the crossover frequency?

To me it looks like they are not in phase.

 

[kifeep]

Might be easier to add a sub or two What is your goal here?

I figured that I should try to get one sub integrated as well as possible before adding another set of variables. I mean, I know additional subs can help even out responses, but the FR on the one sub isn't that bad. I assume adding another wouldn't just magically make the crossover work as expected, but maybe it would. I'm still new at this.

I'd like to understand why crossing over between two decent FR curves makes things worse. I'd also like it to make it not happen.

 

[kifeep]

Are you adjusting the time delay/phase of the woofer to ensure that it is in phase with your speakers at the crossover frequency?

No. I'm not sure how to accomplish that. But I can pull some phase graphs from the measurements here in a minute.

 

[Chrispy]

I figured that I should try to get one sub integrated as well as possible before adding another set of variables. I mean, I know additional subs can help even out responses, but the FR on the one sub isn't that bad. I assume adding another wouldn't just magically make the crossover work as expected, but maybe it would. I'm still new at this.

I'd like to understand why crossing over between two decent FR curves makes things worse. I'd also like it to make it not happen.

What does this mean particularly ? "I managed to find a pretty good location for it"

 

[kifeep]

Here's the phase graph of the uncrossed right and sub channels.

And here's the phase graph at 3 different crossover points

 

[kifeep]

What does this mean particularly ? "I managed to find a pretty good location for it"

Just that there no serious dips in the FR, making EQ easy and effective. A prior location had a bad dip at 40hz and could never get a good EQed response.

 

[terryforsythe]

No. I'm not sure how to accomplish that. But I can pull some phase graphs from the measurements here in a minute.

The easiest solution is to adjust the subwoofer's phase/time delay.

Another option is to select a higher crossover frequency, and play with the distance of the subwoofer from the listening position until you get the speakers and subwoofer in phase at the listening position.

 

[kifeep]

The sub has a phase/delay dial that is disabled when using the LFE input. Adjusting the delay in the AVR is simple, but I don't know how to determine the correct value.

 

[Chrispy]

Just that there no serious dips in the FR, making EQ easy and effective. A prior location had a bad dip at 40hz and could never get a good EQed response.

I was more thinking you really need to explore sub positions, particularly for a single, as the room will dictate so much. Old saying, the sub more needs to go where it needs to go more than where you want.....

 

[terryforsythe]

The sub has a phase/delay dial that is disabled when using the LFE input. Adjusting the delay in the AVR is simple, but I don't know how to determine the correct value.

REW has a feature for that.

1. Set the crossover point for both the speaker and subwoofer.
2. Run a frequency response sweep (SPL & Phase) for the speaker, and then for the subwoofer.
3. Select the "All SPL" tab, then place the cursor at the crossover frequency.
4. Then select "Actions", then select "Alignment Tool".
4. In that tool, select the two graphs and then select either "Align phase at cursor" or "Align phase slopes at cursor". That will compute the appropriate time delay.

Alternatively, you can output a single tone at the crossover frequency to the subwoofer and speakers, measure the SPL at the listening position, and adjust the time delay on the subwoofer until you find the time delay that results in the highest SPL at the crossover frequency.

 

[terryforsythe]

Also, you may want to play with different crossover slopes/topology if your AVR provides various options for that.

 

[kifeep]

REW has a feature for that.

1. Set the crossover point for both the speaker and subwoofer.
2. Run a frequency response sweep (SPL & Phase) for the speaker, and then for the woofer.
3. Select the "All SPL" tab, then place the cursor at the crossover frequency.
4. Then select "Actions", then select "Alignment Tool".
4. In that tool, select the two graphs and then select either "Align phase at cursor" or "Align phase slopes at cursor". That will compute the appropriate time delay.

Alternatively, you can output a single tone at the crossover frequency to the subwoofer and speakers, measure the SPL at the listening position, and adjust the time delay on the subwoofer until you find the time delay that results in the highest SPL at the crossover frequency.

Ok, cool. I'll give that a try tomorrow. I already have full suite of sweeps so it should be simple.

Also, you may want to play with different crossover slopes/topology if your AVR provides various options for that.

The AVR is controlling the crossover. If I don't use the LFE input on the sub, then I can adjust the LP slope, but I have no control over the HP side.

 

[kifeep]

The AVR is controlling the crossover. If I don't use the LFE input on the sub, then I can adjust the LP slope, but I have no control over the HP side.

Or to be more accurate, I could add the subwoofer's LP filter on top of the LP'd signal the AVR is already sending to the sub. Full control of the crossover topology would involve ditching the AVR for some other crossover solution. Not really willing to do that at this point.

 

[terryforsythe]

Ok, cool. I'll give that a try tomorrow. I already have full suite of sweeps so it should be simple.

Make sure that the sweeps you choose for the subwoofer and speakers are ones that were performed with the crossover enabled and set to your desired crossover frequency, and with the speakers and subwoofer placed where you want them.

As a side note, too often people choose too low of a crossover frequency for their speakers, which makes getting the transition smooth more difficult. Generally speaking, I usually choose a frequency that is around 1 octave higher than the resonant frequency of the speaker's woofer. You can determine that by running an impedance sweep on the speaker.

Also, if the speakers are ported, I have achieved better subwoofer integration by plugging the ports of the speakers. I put dense foam in mine. You can try it both ways and see which one works best for your speakers.

 

[terryforsythe]

Or to be more accurate, I could add the subwoofer's LP filter on top of the LP'd signal the AVR is already sending to the sub. Full control of the crossover topology would involve ditching the AVR for some other crossover solution. Not really willing to do that at this point.

I would try using the AVR first and see how it turns out once you get the time delay properly set.