Gain matching my dual subs

[meek81]

So I am trying to learn how to gain match my dual PB16 subs, I watched a video on YouTube by Hometheatergurus and although very informative some parts still have me scratching my head.. so I’ve come here for help. I have my subs connected to my minidsp 2x4HD that I just got so I’m trying to learn that while trying to learn gain matching.. I think I’m in a little over my head honestly. So at what volume does my AVR need to be when I am doing the 60Hz tone in REW to adjust the gain to 80db? And also since I have identical subs can I just set the same gain on my second sub without putting it in the position of the first sub to gain match them? Any help will be greatly appreciated, thanks

 

[Kakkadu]

Gain matching is a bad idea. Place your subs asymmetrically to your room and adjust the gain until you get the best frequency response. The subs will smooth out the peaks and nulls otherwise present from room modes.

 

[meek81]

Gain matching is a bad idea. Place your subs asymmetrically to your room and adjust the gain until you get the best frequency response. The subs will smooth out the peaks and nulls otherwise present from room modes.

Dang... I keep hearing that gain matching is good to do now I’m not sure what to think... this rabbit hole just keeps getting deeper and deeper. If you don’t mind me asking, why do you think it is a bad idea?

 

[Kakkadu]

Dang... I keep hearing that gain matching is good to do now I’m not sure what to think... this rabbit hole just keeps getting deeper and deeper. If you don’t mind me asking, why do you think it is a bad idea?

For the reason I just told you... if you match the gain your chances of getting a flat room response are diminished. The best way is to move the subs, measure, adjust gain, measure and repeat until your room response is as flat as you can get it. REW actually has a function for this where you can simulate sub positions and gains and see the calculated response. Then when you get the response flat, you can match the total gain to your main speakers (I assume you weren't talking about integrating the sub to main speakers by matching gain). In any case you need a measurement microphone and REW as minimum if you want to do it right.

 

[meek81]

For the reason I just told you... if you match the gain your chances of getting a flat room response are diminished. The best way is to move the subs, measure, adjust gain, measure and repeat until your room response is as flat as you can get it. REW actually has a function for this where you can simulate sub positions and gains and see the calculated response. Then when you get the response flat, you can match the total gain to your main speakers (I assume you weren't talking about integrating the sub to main speakers by matching gain). In any case you need a measurement microphone and REW as minimum if you want to do it right.

Ok but since I am using a minidsp 2x4HD wouldn’t it make sense to have them gain matched since technically the dsp will be treating my subs as one virtual sub? And I do have a umik-1 and REW but there is so much info out there that I’m not sure what to do at the moment.

 

[JStewart]

Ok but since I am using a minidsp 2x4HD wouldn’t it make sense to have them gain matched since technically the dsp will be treating my subs as one virtual sub?

Gain matched or level matched at the MLP, the dsp will still treat them together as one sub with MiniDSP.

Dang... I keep hearing that gain matching is good to do now I’m not sure what to think... this rabbit hole just keeps getting deeper and deeper

The purpose of gain matching is to avoid a situation of having one sub run out of capacity before the other resulting in a loss of total capacity. The gain matching video you watched tries to explain this at the beginning when he compares gain matching to level matching.

I would reason the importance of gain matching becomes lessened with excess output capacity from the subs.

Then there is the whole business of "soundfield management" where the goal is to have the sub response the same at multiple seats. (See Multi-Sub Optimizer). This concept may intentionally change sub relative gains to achieve the goal.

Then there is sub placement. Its commonly stated that we're not able to localize a sub when the XO is at at 80hz. But consider that with an XO of 80 a sub has output well beyond that. Typically to 160hz. If one sub is near you and one sub not, will you perceive bass as coming from the near one? I've tried it and I believe I can, content depending.

I'm no expert but I think part of the confusion stems from neither gain matching nor level matching concepts are 100% right or 100% wrong and their concepts would not apply in all situations. I think it also depends on your desired outcome, which when starting out is likely also an unknown.

So, my advice will be this... The videos in the home theater guru series present a valid method. So there is no reason not to follow the instructions. If you don't like the outcome try something else. Learning is fun

 

[meek81]

Gain matched or level matched at the MLP, the dsp will still treat them together as one sub with MiniDSP.



The purpose of gain matching is to avoid a situation of having one sub run out of capacity before the other resulting in a loss of total capacity. The gain matching video you watched tries to explain this at the beginning when he compares gain matching to level matching.

I would reason the importance of gain matching becomes lessened with excess output capacity from the subs.

Then there is the whole business of "soundfield management" where the goal is to have the sub response the same at multiple seats. (See Multi-Sub Optimizer). This concept may intentionally change sub relative gains to achieve the goal.

Then there is sub placement. Its commonly stated that we're not able to localize a sub when the XO is at at 80hz. But consider that with an XO of 80 a sub has output well beyond that. Typically to 160hz. If one sub is near you and one sub not, will you perceive bass as coming from the near one? I've tried it and I believe I can, content depending.

I'm no expert but I think part of the confusion stems from neither gain matching nor level matching concepts are 100% right or 100% wrong and their concepts would not apply in all situations. I think it also depends on your desired outcome, which when starting out is likely also an unknown.

So, my advice will be this... The videos in the home theater guru series present a valid method. So there is no reason not to follow the instructions. If you don't like the outcome try something else. Learning is fun

Thank you your input is very much appreciated, this home theatre stuff is fun but man it can get overwhelming real quick The outcome I want is a decent sounding system that I use for movies and gaming, it is a living room set up so no treatments or anything like that so I know the sound won’t be perfect. Thanks again!

 

[Kakkadu]

My recommendation is to use the REW room simulator and experiment using it. Your goal should be to get a flat response in your listening position and worry less about the gain.

 

[JStewart]

My recommendation is to use the REW room simulator and experiment using it. Your goal should be to get a flat response in your listening position and worry less about the gain.

I've found the room simulator to work well in a sealed rectangular room. IIRC the simulator is not designed to work properly without those parameters met. It can be quite handy though in the right situation.
This unfortunately leaves the OP to move subs and measure or put the sub in the listening chair and move the mic to a sub location and measure. Being that its a living room I'd suspect limited placement options available too, but you never know..

 

[meek81]

I've found the room simulator to work well in a sealed rectangular room. IIRC the simulator is not designed to work properly without those parameters met. It can be quite handy though in the right situation.
This unfortunately leaves the OP to move subs and measure or put the sub in the listening chair and move the mic to a sub location and measure. Being that its a living room I'd suspect limited placement options available too, but you never know..

Yeah unfortunately very limited space and locations for my subs.. I’ll try and mess around with that simulator I’m REW, maybe try and move my seating forward or back a bit see if that improves the bass a bit.

 

[Kakkadu]

Of course you have to verify everything with measurements - but try to plaec the subs as asymmetrically as you can. That alone may improve things.

 

[madrac]

There are some guides on-line regarding sub placement but, of course, depends on where you can place in your room.
I keep my subs out of corner - not a fan of corner loading.

Some recommended placements for duals include opposite "corners" (eg Left Front, Right Rear or vice-versa), 1/4 way in on either side of front wall (this is approximately where I have mine, although more like an 1/8 due to space), middle of front/rear walls, middle of side walls.

Drivers don't necessarily have to point at MLP. Play around with different orientations to see how that impacts response. You can even point towards the wall.

As to gain matching, this is what I did with my dual Rythmik E15's, miniDSP 2x4 HD, and REW.
* Set sub volume (or gain, depending on how it's labeled) to 12 o'clock.
* Set AVR volume to -20 (although I don't think it matters where the volume is).
* Set AVR sub crossover (LPF) to max (in my case 250)
* Turn off other parameters in AVR (eg, Cross-overs to flat, distances to 0 (except whichever speaker you are using for timing reference - I chose the Center - input the measured distance), trims to 0, clear all EQ filters that may have been uploaded into the AVR). Use the speaker profile (if you have multiple) that's set this way for REW.
* set up controls on sub amp as per recommendations (eg, defeat any PEQ, max cross-hover, low damping, max extension (in my case 14 hz), rumble filter off, etc.
* Make sure the mini-DSP is set up (eg, no PEQ, no trim, no cross-overs, etc.)
* Set up REW preferences and Windows (or Mac) as per on-line guide (AustinJerry on AVS has an awesome guide).
* Place mic, pointing to ceiling, as close to sub 1 driver as possible. I used 5" from the lowest point of the driver.
* Turn on sub you are measuring in miniDSP, others are off.
* Use pink noise from either AVR or REW to measure sub 1 output. Record.
* Move mic to sub 2, same distance.
* Turn on sub 2 in miniDSP, others off.
* Measure sub 2 output. Record.
* Decrease gain in miniDSP output for sub that reads higher. eg, if sub 1 is 105 DB and sub 2 is 110 DB, apply -5 gain to sub 2 output.
* Repeat each sub, if you want, to confirm measurements.
* Turn on both subs.
* Voila, you have just gain matched your subs (w/o having to move them to a certain position in the room). I believe it's Ep26 (or 26.1) from HomeTheaterGurus where he explains gain matching. This method is explained towards the end of the video. He says there may still be some room boundary effects, but likely negligible (IMO). Much easier than lugging heavy subs back and forth.
* Now you can move to the next step (aligning).
* Measure each sub independently in REW at your MLP.
* Measure both together.
* Add delay in output(s) of sub(s) (or you may need to invert one if behind you) to get the best response of the summed sums (eg, the curve above both individual measurements by ~6 DB if possible. May not have that much output across the frequency response. In my case, there was a spot around 75 HZ where the summed was a bit lower than one of the individual measurements by about 3 db. I could never get it above both playing with delays w/o hurting other parts of the graph. That was fixed with EQing.

 

[linearphase]

I don’t do HT but do spend many hours every day helping people do 2 ch setup and calibration.. Gain matching is not really necessary but if you want to do it the procedure above should work. I would change one thing, trim the sub gain using the sub level pots in the sub itself. The reason being is any DSP unit will clip if the gain from EQ causes it to exceed full scale ( it’s maximun output). This is digital so it creates hard clipping with the ensuing nasty sound.
Also put the mic within 1/2 inch of the woofer cone to measure anechoic response. Just be careful to not drive it so hard it hits the mic. The mic orientation does not matter at sub frequencies.

 

[Gregory Livingstone]

If you do follow up the gain matching with alignment, REW has an alignment tool that has been quite accurate in my experience. It is in the All SPL tab under options. When you enable it a dialog box appears where you supply the individual sub measurements you made via drop down. At the sub closest to you can put in delay values or just use the arrow keys to see how the combined FR looks with different delay setting. You can save the predicted aligned response as a measurement for later comparison with actual measurement.

Set the MiniDSP output channel delay. Use REW to measure the combined subs. The FR should be very close to that predicted by AT. When experimenting with AT there could be multiple settings (delay, polarity) that show acceptable FR. You might want to try all of interest and verify with actual measurements (more efficient to do in single measuring session).

It is as simple as that, but the individual sub measurements need to be accurate in which case they are repeatable (e.g. the same FR and Impulse charts for each measurement of a particular sub). The sub measurements need a timing reference which is easy to do with REW using HDMI connection to AVR and ASIO drivers. You can play the timing reference through the center channel. When I first tried this I found I wasn't able to make repeatable sub measurements (I had to reset subs to factory default).

Note, the MiniDSP HD allows up to 70-80 msec of delay. The regular MiniDSP on 7 msec.
Note2, The AT allows setting delay by phase alignment based on cursor. That can be interesting, for instance see what happens when aligning to dips or peaks or potential Xover.

Have fun!
Greg

 

[linearphase]

Good points. To do it really well is quite challenging. Time reference is one very important reason. Going through that right now with guy from Nashville.
There are virtually infinite combinations of gain, phase and polarity that will produce a reasonably flat response. However, there is only 1 that will produce the best achievable response.
While this arguably more important for two channel it will not hurt HT.
To be very fair though it requires lots of knowledge and patience as well as fair amount of trial and error to get right= lots of time!
For me this is not achievable unless I take lots of measuremnts. Getting those right through is, for me, half the fun.
Good luck with you endeavor.

 

[BJB]

I just have one question. I know I have received outstanding advice from the REW author that for response measurements the AVR LFE crossover should be set to the same as your subs/mains will be, say 80 hz. If using LFE for measurements.

I am wondering about the thought process of setting the LFE crossover to the highest Hz for gain matching measurements? My initial thought is it doesn't matter either way. Just curious.
Thanks,
BJB

 

[linearphase]

As a first approximation that is true. It will remain true as long as the sub or mains have smith response well above or below respectively that frequency. Otherwise you have to measure.

 

[Kakkadu]

The non-HD MiniDSP cannot be recommended, its maximum line out voltage is too low to achieve a full output on regular amplifiers. It requires a booster amp in the chain. I never understood the logic behind that.