Hello,
I recently obtained second subwoofer for my first "real" home cinama / music reproduction system. Now I would like to get most of it: integrate subwoofers together for both relatively flat response and high efficiency and integrate them with mains. I also intend to play with correcting phase of mains and surrounds etc. I am new to all this, but getting familiar with REW, MSO, Rephase, VoiceMeeter Banana and EqualizerAPO in the last weeks. All very impressive software.
Now I am going to order UMIK 1 and contemplating whether I should pick easy to use but limited MiniDSP 2x4 (HD) together with it or try to use (much more powerful) software DSP with additional hardware.
My components:
- Intel NUC NUC8i3BEH as primary source
- Sony STR-DN1050 AVR (HDMI in)
- 2x Monoprice Monolith 10 THX subs
- B&W DM602 S2 mains + B&W LCR60 S3 center
- 2 surrounds from Infinity HCS series
As a play POC for software solution I used 3.5mm connection out of NUC going to active subwoofers. This could be replaced with dedicated 2-channel DAC. Then l used VoiceMeeter and EqualizerAPO to send sub-80Hz to subwoofers via this DAC and above-80Hz to AVR. Of course immediate problem was that volume was mismatched between AVR and DAC.
So that is my first question: how do you usually tackle volume synchronization when using multiple different DAC/amps? Average AVR does not have enough inputs/outputs to pass everything through it, especially additional subwoofers.
I can see the why the MiniDSP 2x4 (HD) is so popular. In my case it would simply fit after AVR analog sub-out. However I don't like unnecessary A->D->A conversion and hardware limitations, such as low number of FIR taps. With my current understanding, they are not sufficient for linearizing of subwoofer phase linearization for example.
Things could be improved by using digital input of MiniDSP, but in this case I have same problem with mismatched volume as mentioned in first question if my understanding is correct. Digital input would come from NUC and in this case I could instead use dedicated (better) DAC and do DSP in software.
In the second part I am referring to two attached measurement charts. Measurement was done by RoomAcousticsMeter on uncalibrated S10+ microphone. Sweep was done, from which IR was calculated and imported to REW. It is possibly entirely off, but I am using that setup to learn software before I get UMIK 1. But graphs mostly show what I would say I am hearing.
- FRSubs - 4 lines of FR
In listening tests, lowest frequencies (<40Hz) are more powerful if subs are ran in phase (not indicated in plot). However 40-80Hz get all messed up with swings and general lower output than if subs are ran out of phase. If subs are out-of-phase, <40Hz is less pronounced with much higher output in 40-80Hz region. But in this case, room modes are not tackled well and swings are even larger. I'd say output is much higher, but lower quality.
What I would like to play with is with matching the phase of the subs with FIR. 20-40Hz looks nice, but 45-70Hz region looks like it is almost 180° apart, which would explain cancelation. Later I would attempt similar matching with mains at crossover region. Phase matching sounds perfect for getting good efficiency.
Is what I am trying to do sensible at all? Or I am just looking at things way too naively as I am just starting with this? If this is possible in theory, is it possible with limited taps (4K total, 2K if 2 channels are used) in MiniDSP 2x4HD? From what I gathered so far it is not.
I experimented with MSO in the past days. This one uses exclusively IIR filtering. Results I get seem very good from FR flatness perspective, but I don't like apparent efficiency aspect of it. Like 10dB+ of required gain difference between subwoofers and -20dB individual IIR filters. However it seems by using this tool (or IIR filters-only in general) I could simply use regular cheap MiniDSP 2.4 and be done with it.
Sorry for the long post that perhaps tries to address too much at once, but I tried to present whole set-up and thinking. So someone can stop me right there if some ideas are naive or not necessary. Perhaps FIR filtering for subwoofers do not have as many benefits as I am imagining and A-D-A conversion in MiniDSP is good enough for subwoofers, so that I can forget about trying to run everything on PC.
I recently obtained second subwoofer for my first "real" home cinama / music reproduction system. Now I would like to get most of it: integrate subwoofers together for both relatively flat response and high efficiency and integrate them with mains. I also intend to play with correcting phase of mains and surrounds etc. I am new to all this, but getting familiar with REW, MSO, Rephase, VoiceMeeter Banana and EqualizerAPO in the last weeks. All very impressive software.
Now I am going to order UMIK 1 and contemplating whether I should pick easy to use but limited MiniDSP 2x4 (HD) together with it or try to use (much more powerful) software DSP with additional hardware.
My components:
- Intel NUC NUC8i3BEH as primary source
- Sony STR-DN1050 AVR (HDMI in)
- 2x Monoprice Monolith 10 THX subs
- B&W DM602 S2 mains + B&W LCR60 S3 center
- 2 surrounds from Infinity HCS series
As a play POC for software solution I used 3.5mm connection out of NUC going to active subwoofers. This could be replaced with dedicated 2-channel DAC. Then l used VoiceMeeter and EqualizerAPO to send sub-80Hz to subwoofers via this DAC and above-80Hz to AVR. Of course immediate problem was that volume was mismatched between AVR and DAC.
So that is my first question: how do you usually tackle volume synchronization when using multiple different DAC/amps? Average AVR does not have enough inputs/outputs to pass everything through it, especially additional subwoofers.
I can see the why the MiniDSP 2x4 (HD) is so popular. In my case it would simply fit after AVR analog sub-out. However I don't like unnecessary A->D->A conversion and hardware limitations, such as low number of FIR taps. With my current understanding, they are not sufficient for linearizing of subwoofer phase linearization for example.
Things could be improved by using digital input of MiniDSP, but in this case I have same problem with mismatched volume as mentioned in first question if my understanding is correct. Digital input would come from NUC and in this case I could instead use dedicated (better) DAC and do DSP in software.
In the second part I am referring to two attached measurement charts. Measurement was done by RoomAcousticsMeter on uncalibrated S10+ microphone. Sweep was done, from which IR was calculated and imported to REW. It is possibly entirely off, but I am using that setup to learn software before I get UMIK 1. But graphs mostly show what I would say I am hearing.
- FRSubs - 4 lines of FR
- Blue and green solid lines: two individual subwoofers measured separately. They are on opposing sides at the middle of long wall with microphone at seating position at the short wall
- light blue dotted line: both subs running together with same phase setting
- violet dashed line: right sub running with inverted phase
In listening tests, lowest frequencies (<40Hz) are more powerful if subs are ran in phase (not indicated in plot). However 40-80Hz get all messed up with swings and general lower output than if subs are ran out of phase. If subs are out-of-phase, <40Hz is less pronounced with much higher output in 40-80Hz region. But in this case, room modes are not tackled well and swings are even larger. I'd say output is much higher, but lower quality.
What I would like to play with is with matching the phase of the subs with FIR. 20-40Hz looks nice, but 45-70Hz region looks like it is almost 180° apart, which would explain cancelation. Later I would attempt similar matching with mains at crossover region. Phase matching sounds perfect for getting good efficiency.
Is what I am trying to do sensible at all? Or I am just looking at things way too naively as I am just starting with this? If this is possible in theory, is it possible with limited taps (4K total, 2K if 2 channels are used) in MiniDSP 2x4HD? From what I gathered so far it is not.
I experimented with MSO in the past days. This one uses exclusively IIR filtering. Results I get seem very good from FR flatness perspective, but I don't like apparent efficiency aspect of it. Like 10dB+ of required gain difference between subwoofers and -20dB individual IIR filters. However it seems by using this tool (or IIR filters-only in general) I could simply use regular cheap MiniDSP 2.4 and be done with it.
Sorry for the long post that perhaps tries to address too much at once, but I tried to present whole set-up and thinking. So someone can stop me right there if some ideas are naive or not necessary. Perhaps FIR filtering for subwoofers do not have as many benefits as I am imagining and A-D-A conversion in MiniDSP is good enough for subwoofers, so that I can forget about trying to run everything on PC.
