Where to process the signal for room correction and subwoofer performance?

[skypickle]

So I am asking about using a sub for augmenting a stereo setup. I am NOT asking about subs for home theater because most multichannel processors for HT have dedicated sub outputs and a way to adjust the crossover frequency. Some processors also offer timing adjustment as well as polarity,etc. I have a 'simple' setup:
Windows 10 pro using foobar2k -> PPang USB card driven by a separate low noise linear power supply -> USB cable to Gustard x20 DAC -> schitt headphone amp -> analog balanced line outs to amplifier -> magneplanar speakers (maggie 3.7i). There is no sub presently in this chain. Adding a sub complicates things a bit.

1) which frequencies should the sub play and how do i shunt those frequencies away from my main speakers? Another reason to add a sub would be to make maggie's job easier. The big planar membrane isnt stressed with low frequencies that just add bad harmonics.

2) how do I make sure that the sub is synchronized with my main speakers?

The solution seems to be having a processor before the signal reaches the amplifier. I know REL recommends taking the hi output signal from the speaker terminals (for synchronization purposes?), but that won't make maggie's job easier. So then the question becomes - should i put the crossover between the preamp and the amp or between the DAC and the preamp? Or should I get a sub with a built in crossover and eq like the JL Fathom2?

The question of processing also rears its head. If I have a room anomaly that i cannot fix by adding 'stuffing on the walls', can i add a miniDSP in the chain to correct aberrant hot spots or dead spots? These devices however do an analog to digital conversion, then process and then do another digital to analog conversion.

I am thinking tho that perhaps i should just get software that i can run on the windows box that i can calibrate with a UMIK microphone to generate an eq curve which foobar could then use. This way all processing is done in the digital domain. Then i could just use a splitter on the balanced line out of the headphone amp to feed the main speaker amp as well as the JL fathom2 or the REL s/5.

 

[shadyJ]

The place to implement a crossover in your setup is between the (unfortunately named) schitt headphone and the amplifier. You can use something like a miniDSP, but if you really want to avoid adding another layer of digital processing, look at the Hsu crossovers (this or this) that I mentioned before.

As for the subwoofer frequency range, I would try a few different crossover points. Start with 80 Hz. If you are getting two subwoofers, then you can try higher crossovers points without having to worry about localization. Since the subwoofers will be far more capable in bass frequencies than magnapans, I would try to offload as much work from the Maggies as possible. But set it to where it sounds good to you.

As for synchronizing the sub with your mains, to be really sure you are doing this correctly you will need something to measure the frequency response, so you need a measurement mic like the UMIK-1 and software like Room EQ Wizard. What you do is take a measurement of the bass frequencies, say 20 Hz to 200 Hz, and if you see a big dip in the response at the crossover point, that means the sub and speakers are out of phase. Flip the phase on the subwoofer by 180 degrees and it should be in phase with the speakers. I suppose this is doable by ear, but it is best done by measurement, and besides, if you want great bass you will want to measure the response for a bunch of other reasons as well.

By the way, almost all subs have built-in low pass filters for themselves. Very few subs have high-passed line-outs. These subs will not offload the signals sent to the maggies. You will need a separate crossover.

 

[WaynePflughaupt]

You need an electronic crossover in front of the Maggie’s amplifier. Personally I don't think AD/DA conversions are a big deal, as long as you're using quality equipment. I'm using digital pro audio parametric EQs in my system, and they sound better than the high-end analog EQs I was using before.

But - if you want to avoid additional conversions, get a vintage analog pro-audio model from a company like Ashly, TDM, Klark Teknik, BSS, Furman, Rane, etc. Most of these can be had pretty cheap these days, as can an analog parametric EQ you could use to address your room issues.

I wouldn’t worry about time alignment unless your sub is a considerable distance from your main speakers.

However, you can get some adjustment there with a sub that has a continuously variable phase control, which is essentially an analog delay control. It’s not as precise as a digital delay, but you could do a good job adjusting it with measurements from REW, as suggested.

If the delay needed is for the main speakers (i.e. they are significantly closer to you than the sub), and you really feel it’s necessary to compensate, then you’re going to have to endure the AD/DA conversions from a digital processor like the miniDSP or a pro-audio speaker processor.

Regards,
Wayne A. Pflughaupt

 

[skypickle]

Thank you both for your clear replies. MiniDSP has a new product coming out that operates on s/pdif input , the nanoDIGI. Their other product uses hdmi input. Since I am using USB output from a pc, neither of these will work. I guess I could try Dirac live software on the pc. Staying in the digital domain seems preferable for noise reasons. I really enjoy the sonic 'blackness' of my system. Silence and quiet passages are pristine and you feel like you are in a forest after a snowstorm.

I will also look into the analog processors. My systems' components use xlr inputs/outputs so a pro audio interface should be no problem.