REW & High Level Connection

[Moses Magnum]

Hi,
I'm trying to find an idiots guide that can walk me through measuring and correcting 2 subs for a 2.2 setup using the high level connection.

If it's possible I'd like to use REW to help me dial my subs in manually using phase and crossover settings directly on the subs. Most guides I've found are working with an AVR and a MiniDSP.
Any pointers would be greatly appreciated.

Thanks in advance.

 

[John Mulcahy]

Easiest is probably to use REW's pink periodic noise signal and watch the response on the RTA from a mic at the listening position while you adjust the sub settings. Do left and right separately, so signal from REW using the left channel as output for setting up the left, right for right.

 

[Moses Magnum]

Thanks John.
I'll have a fiddle and see how I get on.

 

[Moses Magnum]

dic noise signal and watch the response on the RTA from a mic at the listening position while you adjust the sub settings. Do left and right separately, so signal from REW using the

Is there an idiots guide for this procedure?

 

[John Mulcahy]

There isn't a lot to it. Play the pink periodic noise signal linked above, start the RTA with the settings shown in the help for the signal. Change subwoofer settings and watch the effect they have on the RTA plot.

 

[Moses Magnum]

Thank you.
I have the settings for the Pink Noise as per the image in the help file, same for the RTA.
My issue though is I have no idea as to what I should be gleaning from the data and what I should be aiming for...

 

[John Mulcahy]

Ideally the response would be fairly level through the crossover region between sub and mains. One approach to getting the phase right is to adjust the subs polarity and (if available) phase settings to get the biggest dip in the crossover region, which should correspond to sub and main being 180 degrees out of phase with each other, then flip the sub polarity which should get them back in phase. You can save the current RTA display as a measurement using the button above the RTA graph, if you do that a few times with settings you are considering you can attach the mdat file (produced from using "Save all" on the main REW window) for comment.

 

[Moses Magnum]

No idea if what I've uploaded is of any use...
I set the high pass filter in the amp at 80hz, crossover setting on the sub at 80hz and took 3 readings from the left channel. Speaker, sub and combined.
I don't understand what I'm supposed to be looking for at the crossover point on the readings.
My apologies for being a pain.

 

[John Mulcahy]

How did you have the pink periodic noise set up? Full range would be usual but it looks like it has a very limited range. The speaker only reading doesn't look like it has an 80 Hz high pass applied to it.

With the speaker and sub playing, watch the trace, then flip the sub polarity and see what changes. If you want to capture the response from both polarities you can attach them. Also on the RTA appearance settings uncheck "Use bars on RTA".

 

[Moses Magnum]

Hopefully this is more like it's supposed to look.

Set at full range pink noise.
Thank you for your time.

 

[John Mulcahy]

Looks better. I've applied a little smoothing to the traces in these plots, 1/24 octave.

Looking at the low end, 10 to 200 Hz, the first measurement in red has more output throughout the region the sub is active, which indicates it is better aligned to the main than the second measurement:

Unusually, the difference persists right up to 2 kHz, which suggests the sub output does not have a low pass filter applied or its low pass is set to the maximum frequency. You probably want the sub low pass set to 80 Hz or so. The sub's low pass has an effect on its phase through the crossover region, so you'll need to measure again after setting the low pass filter to see how the results look after that.

Something you can do when measuring is to save the RTA response with your starting settings, which adds that saved trace on the RTA. It then provides a handy reference against which you can compare the live result as you adjust the sub's controls. You can also try coarser RTA settings, such as 1/12 or 1/6 octave, which reduce the detail (because the bands of the RTA are wider) but preserve the overall level differences, making a little easier to make the comparison.

 

[Moses Magnum]

Thank you for that, hopefully this next file looks better. Set sub at 80hz.

 

[John Mulcahy]

Yes, phase 0 looks the better option, though the sub volume seems to have been turned right up so it is very loud compared to the main. It can be subjectively preferable to have the sub a bit higher than the main, but by a few dB, it looks to be about 12 dB too loud. Hopefully you have got the hang of changing things and watching what happens to the response. If the sub has a variable phase control you could try adjusting it to see if the transition from sub to mains gets more even, but fix the volume first.

 

[Moses Magnum]

Thank you!
I think I know what I'm looking at now, I'll have another play around after work. If I'm understanding this correctly the aim is to get a flat a line as possible up to around 200hz?
Would measuring the speaker on it's own so I can see exactly what the introduction of the sub does be wise? This would give me something to aim for once I start adjusting the sub.

 

[John Mulcahy]

Yes, that's right. It will never be particularly flat, the room pushes the response all over the place at low frequencies though smoothing (or a lower octave fraction on the RTA) helps make the underlying trend easier to see. A speaker measurement helps show where it is contributing, but it is the combined response that is most relevant for your setup.

 

[Moses Magnum]

I've has another play around, would you mind looking at these and let me know what you think please?

 

[John Mulcahy]

MainL and MainR look the better pair to me, levels are good. An alternative to the 1/6 octave RTA is to leave the RTA on 1/48 and use 1/6 octave smoothing (box below the RTA mode), may be easier to read when you are making changes.