REW and "Sub" Integration - what's your take?

[bogart]

LR4 not LR2.

And usually you want to do similar measurements for the subwoofer so the (anechoic) volume of the two is equal. I discuss that in the thread i linked. You don’t want the speaker to play louder than the subwoofer or vice versa at the crossover frequency.

Got it. So, applying the electrical BU2 on top of the measured acoustic BU2 gives us an LR4 effective crossover.

Restating my next steps for my own clarity, and to help this thread be a more helpful reference:

We will now do the same approach with close-mic measurements of the subwoofer to first establish its anechoic response, correct for baffle step, and use these determine what its appropriate crossover shape for the subwoofer would be. Since the subwoofer is ported, we will need to measure its port and driver then sum it as in napilopez's guide since its port is used, unlike for the desktop speakers.

 

[mdsimon2]

Facilement - here we are. Looks a little tighter in the crossover range but a little more overshoot in the lower frequencies. Probably a better match?

View attachment 172290

There is a good chance that you will not have a BW2 (Q = 0.7) roll off. Eye balling that response it looks to me like you are probably at a slightly higher Q (maybe 0.8-0.9) at a slightly higher Fc than 93.5 Hz. It is better to get an accurate model of the response and then use a Linkwitz Transform to achieve a BW2 response.

Michael

 

[abdo123]

There is a good chance that you will not have a BW2 (Q = 0.7) roll off. Eye balling that response it looks to me like you are probably at a slightly higher Q (maybe 0.8-0.9) at a slightly higher Fc than 93.5 Hz. It is better to get an accurate model of the response and then use a Linkwitz Transform to achieve a BW2 response.

Michael

This is indeed the optimal way to do it, but I wouldn’t do it unless i want to do the crossover higher or lower than the F3 (-3dB) of the speaker.

It wouldn’t make that much of a difference otherwise.

 

[bogart]

@mdsimon2 thank you for the suggestion. I think I will first measure my subwoofer's wide-open response as I may end up needing to do the modeling exercise twice and will be easier if I do that in succession.

I've uploaded my mdat file here in case I get out of my depth I'll update it again as I get my subwoofer sweeps in.

Microsoft OneDrive

1drv.ms

 

[abdo123]

You don’t need to do the whole shablang for the subwoofer just make sure that the volume at the crossover frequency is equal anechoically (gain matched).

So just the subwoofer driver with baffle step. Lower the gain or the subwoofer or increase it till the speaker and subwoofer response meet at the crossover.

 

[bogart]

I'm quiet right now as the measured woofer response with BW 12dB/Octave crossover did not appear to have an effect - woofer response looked unchanged. I expect I had an issue with the XO settings in my miniDSP, so I'm planning to re-measure it for confidence!

 

[abdo123]

I'm quiet right now as the measured woofer response with BW 12dB/Octave crossover did not appear to have an effect - woofer response looked unchanged. I expect I had an issue with the XO settings in my miniDSP, so I'm planning to re-measure it for confidence!View attachment 172455

Make sure that you’re using a low pass for the subwoofer, it lets the low frequency through.

Also i would limit the measurement or sweep to 500Hz, no meaningful data above that.

 

[bogart]

Make sure that you’re using a low pass for the subwoofer, it lets the low frequency through.

Also i would limit the measurement or sweep to 500Hz, no meaningful data above that.

Got that one - LR4 on the subwoofer seems to have taken just fine. Just trying to be sure the woofer gets its slope right before I get to gain matching.

 

[abdo123]

Got that one - LR4 on the subwoofer seems to have taken just fine. Just trying to be sure the woofer gets its slope right before I get to gain matching.

View attachment 172456

Both of them should be at -6 dB on the crossover frequency. So LR4 behavior on both side.

 

[bogart]

Both of them should be at -6 dB on the crossover frequency. So LR4 behavior on both side.

Was thinking I might want to PEQ down from 40-100 Hz in the subwoofer to get it closer to line with the 20-30Hz response. I think I need to do that first before gain-matching?

 

[abdo123]

Was thinking I might want to PEQ down from 40-100 Hz in the subwoofer to get it closer to line with the 20-30Hz response. I think I need to do that first before gain-matching?

The port response will even that out, only do global / input level EQ to manage the response at the listening position.

 

[bogart]

Nearly there. Speaker LF appears to now have an LR4 behavior, so it must have been the problem earlier. Subwoofer level is still a tiny bit too hot since we're crossing at 100Hz instead of 93, but a quick listen indicates it's actually quite close with a fairly smooth transition between speakers and no obvious "subwoofer" sound.

Close mic driver measurements with baffle step adjustments and active crossovers for each speaker below!

 

[abdo123]

Nearly there. Speaker LF appears to now have an LR4 behavior, so it must have been the problem earlier. Subwoofer level is still a tiny bit too hot since we're crossing at 100Hz instead of 93, but a quick listen indicates it's actually quite close with a fairly smooth transition between speakers and no obvious "subwoofer" sound.

Close mic driver measurements with baffle step adjustments and active crossovers for each speaker below!
View attachment 172476

Looking good!

 

[D!sco]

Looks great! Just remember that the electrical crossover point, aka what you enter, could be different from the acoustic crossover points. If the subwoofer doesn’t blend, adjusting its crossover frequency and slope can be useful, such as a hump around 100hz appearing due to SPL mismatch could be resolved by moving the sub LPF down to, say, 95-90hz. This kind of fine tuning takes me weeks and probably doesn’t matter all that much, but there are a lot of ways to skin a cat here.

 

[abdo123]

Looks great! Just remember that the electrical crossover point, aka what you enter, could be different from the acoustic crossover points. If the subwoofer doesn’t blend, adjusting its crossover frequency and slope can be useful, such as a hump around 100hz appearing due to SPL mismatch could be resolved by moving the sub LPF down to, say, 95-90hz. This kind of fine tuning takes me weeks and probably doesn’t matter all that much, but there are a lot of ways to skin a cat here.

He used the accoustic roll-off to design the crossover.