Does it matter where I put the EQ?

[Boomzilla]

Dramatis Personai:

Emotiva PT1 preamp with sub out
Parts-Express DSP-LF
Emotiva Virtual Cu transmitter/receiver

I can put the DSP-LF unit either:

1. Between the preamp and the transmitter or
2. Between the receiver and the sub

Either way, there’ll be an additional A-D / D-A conversion for the room correction, but with the inherent distortion of the sub, I’m thinking it will be inaudible. My questions are:

Does it matter where the DSP-LP unit goes and why?

Thanks - Boom

[ttocs]

Shouldn't matter where in the chain you put the LF DSP.

There's an argument for placing it at the subwoofer end. Case in point, my subwoofers have built-in ARC. When I run the ARC app on the laptop it communicates wirelessly to each sub. Then, there are some subs in the system which are using VSUB transmitter/receiver "connections" between the processor and subs. So your DSP unit, if connected directly to the subwoofer, would effectively be the same arrangement as with each of my subs that are using the VSUB wireless copper for "connection" to the processor.

Then there's the argument for placing it near the beginning of the chain. Case in point, my miniDSP's are all connected directly to the processor, and some of the outputs are then connected to a VSUB transmitter.

Plus there's DSP stuff going on in the processor.

So there is DSP stuff happening in the processor, miniDSP, and every Martin Logan subwoofer. I run ARC on every ML product I have first, then tweak with the miniDSP's, then run Dirac on the processor. Truth be told however, I don't do much with the miniDSP's, just some delay, level, and in at least one of them there's a bit of PEQ but I try not to do much with miniDSP PEQ if possible.

So in my case, it's all over the place, both wired and wireless.

[marcl]

Feb 12, 2022 16:51:28 GMT -5 ttocs said:

Shouldn't matter where in the chain you put the LF DSP.

There's an argument for placing it at the subwoofer end. Case in point, my subwoofers have built-in ARC. When I run the ARC app on the laptop it communicates wirelessly to each sub. Then, there are some subs in the system which are using VSUB transmitter/receiver "connections" between the processor and subs. So your DSP unit, if connected directly to the subwoofer, would effectively be the same arrangement as with each of my subs that are using the VSUB wireless copper for "connection" to the processor.

Then there's the argument for placing it near the beginning of the chain. Case in point, my miniDSP's are all connected directly to the processor, and some of the outputs are then connected to a VSUB transmitter.

Plus there's DSP stuff going on in the processor.

So there is DSP stuff happening in the processor, miniDSP, and every Martin Logan subwoofer. I run ARC on every ML product I have first, then tweak with the miniDSP's, then run Dirac on the processor. Truth be told however, I don't do much with the miniDSP's, just some delay, level, and in at least one of them there's a bit of PEQ but I try not to do much with miniDSP PEQ if possible.

So in my case, it's all over the place, both wired and wireless.

Ideally we'd want to stay in the digital domain until the last moment when we go DAC out to the amps.  Wouldn't it be nice to have digital processing loop outputs/inputs so we could step out to the device of our choice and back inside without any A/D/A cycles.

[Boomzilla]

Feb 12, 2022 17:12:51 GMT -5 marcl said:

Ideally we'd want to stay in the digital domain until the last moment when we go DAC out to the amps.  Wouldn't it be nice to have digital processing loop outputs/inputs so we could step out to the device of our choice and back inside without any A/D/A cycles.

In theory, that would be great. In practice, it would require the adoption of a universal digital standard that all digital devices would adhere to. Good luck with that...

As to ttocs' comment, If I understand you correctly, location (at least in your system) would depend on whether one wanted the DSP-LF to affect only one sub, some of them, or all of them. That's not a decision I'll need to make since (at least, at the moment) I'm using only a single sub. It sounds nice to have the flexibility, though.

As to ideal theoretical solutions: A processor with multiple, assignable, sub outputs would allow the user to select:

* what channel(s) to assign to each sub output
* what crossover frequency & slope to assign to each sub output
* what equalization (if any) to assign to each sub output
* what room correction (if any) to assign to each sub output
* what sub outputs would share the +1 movie soundtrack sub channel

and do all the above without additional A-D / D-A conversions. Since the original signals in the processor are digital anyway, analog conversion wouldn't have to be applied until the actual, individual sub output. A processor like this would minimize distortion in the bass.

[marcl]

Feb 13, 2022 5:43:41 GMT -5 Boomzilla said:

Feb 12, 2022 17:12:51 GMT -5 marcl said:

Ideally we'd want to stay in the digital domain until the last moment when we go DAC out to the amps.  Wouldn't it be nice to have digital processing loop outputs/inputs so we could step out to the device of our choice and back inside without any A/D/A cycles.

In theory, that would be great. In practice, it would require the adoption of a universal digital standard that all digital devices would adhere to. Good luck with that...

As to ttocs ' comment, If I understand you correctly, location (at least in your system) would depend on whether one wanted the DSP-LF to affect only one sub, some of them, or all of them. That's not a decision I'll need to make since (at least, at the moment) I'm using only a single sub. It sounds nice to have the flexibility, though.

As to ideal theoretical solutions: A processor with multiple, assignable, sub outputs would allow the user to select:

* what channel(s) to assign to each sub output
* what crossover frequency & slope to assign to each sub output
* what equalization (if any) to assign to each sub output
* what room correction (if any) to assign to each sub output
* what sub outputs would share the +1 movie soundtrack sub channel

and do all the above without additional A-D / D-A conversions. Since the original signals in the processor are digital anyway, analog conversion wouldn't have to be applied until the actual, individual sub output. A processor like this would minimize distortion in the bass.

How about HDMI?  Sources come in multichannel to the processor on HDMI, so they can go out ... PCM of course.  Sell a breakout box that takes one HDMI input from a processor's HDMI Audio Loop Output, break it to multiple digital channel output pairs (Toslink or coax) which can go into processors like miniDSP, digital outputs back to the breakout box digital inputs to pack them back into HDMI and return to the HDMI Audio Loop Input.  Or people like miniDSP make HDMI multichannel Input/Output versions.

[Boomzilla]

HDMI could work, but the likelihood of getting manufacturers to support it seems slim...

[ttocs]

Feb 13, 2022 5:43:41 GMT -5 Boomzilla said:

As to ttocs ' comment, If I understand you correctly, location (at least in your system) would depend on whether one wanted the DSP-LF to affect only one sub, some of them, or all of them. That's not a decision I'll need to make since (at least, at the moment) I'm using only a single sub. It sounds nice to have the flexibility, though.

It's simpler to show the flow, so here's a diagram for only the Large Fronts in the system. The ML Speakers and ML Subs all have Anthem ARC built-in, which for the speakers only helps the woofers. So this DSP is always active. Each front speaker has a couple subs connected via speaker level, so no DSP between them, just speaker wire.

Then, there's a sub in the rear of the room for smoothing the bass response for the fronts, between 60-70Hz mostly, and miniDSP3 is in charge of this.

So all of this is active all the time for everything, two channel, and HT.  After being output from the XMC-2, there is triple ADA for only the rear sub including the VSUB wireless. Fronts are Large, so the only time processor DSP is added is when using a Dirac filter, which is only for HT at the moment. I still use Reference Stereo for music.


For HT, BM and LFE use three of the same five subs. So in this case ADA appears twice for the front subs and thrice for the rear sub after being output from the XMC-2. And Sub5 has two VSUB receivers connected to different inputs, one for BM/LFE, and one for the Left/Right signal.