Acceptable Group Delay - Subwoofers

[gnarly]

i'm struggling with a few things:

* i can use massive FIRs (in equalizer APO) to make the step response look amazing, but the pre-ringing is bonkers. i believe this is just a fundamental constraint of this sort of correction?

The massive FIR correction, is it correction of the entire spectrum? the sub and KH-120 II combined?
If so, it's pretty much bound to show a step response with huge pre-ring.

* ... music sounded pretty great though - really prominent visceral "chest kick" with bassdrums, in particular, and a sense of extra clarity, but i suspect i was just feeling the pre-delay bass come through before the real kick attack. it was a pretty impressive sensation though.

Can you describe what you mean by feeling the pre-delay? Audible or tactile, or both? And if audible, how the pre-delay sounded?

anyway, i settled with using a 1024-tap FIR on just the mains, to line them up better with the subs (works better than a flat delay). went from +880deg to +560deg at 20Hz, +250deg to +120deg at 100Hz.

however, after a ton of unscientific A/B testing, I think I can only hear it on one very odd techno track, the change was basically inaudible on everything else despite the measurements looking a lot better.

The only real way I know to achieve better phase flattening for the sub-main combo, is to tie them together with a complementary linear-phase crossover. Any other type attempt, like applying "global" phase correction across their summed acoustic response, is just throwing darts hoping to get lucky ime.
The sub and KH-120 II will also need to be co-located for it to work with the complementary linear-phase crossover. If you try this, I'd suggest a steep lin phase LRxover, maybe around 100Hz.

Phase linearization of the sub's low end roll-off is not a good idea ime (same goes for the KH-120's roll-off on its own)

 

[dan_remaincalm]

> Can you describe what you mean by feeling the pre-delay? Audible or tactile, or both? And if audible, how the pre-delay sounded?

Absolutely tactile - a real chest thump vs uncorrected, despite measured frequency response being ~identical. Again, might have just been the sub content showing up before the attack on the kick, or it could be having the impulse lined up right just gives a higher instantaneous SPL.

> The only real way I know to achieve better phase flattening for the sub-main combo, ...

Sub/main integration is pretty good now (should be visible in the graphs). Mains are on stands on desk, tilted up about 5deg, tweeters aimed slightly above reference. Subs are on floor, sortof co-located with mains but offset a bit from each other in all three dimensions.

A few other things to resolve still - from digging around on ASR I found a bunch of data on SVS subs and they seem to have a ton of GD and ringing, which is what's murking up the step response a lot. Also got some desk bounce to sort out.

 

[levimax]

i've been playing around with this stuff too, to room-correct and integrate a set of neumann kh-120 ii with a pair of svs sb-1000 subs with a minidsp flex, REW and rephase

I wish I knew what I know now about group delay and pre-ringing caused by DSP for these "too small" subs that try to go to 20 Hz. If I had I never would have bought them. The idea that subs and mains, especially the modern ones with "small enclosures and lots of DSP" are easy to integrate" is repeated by sub and software vendors but it is not true. For me the only thing worse than struggling manually with it is when I let DIRAC Bass Control try it. The results were radically inconsistent but always nonsense with the timing being set at least 125 ms off between sub and main. The FR was fine, the timing was not. This much of a timing error is quiet audible. Now looking at my setup with REW the mains basically have no GD and the 2 subs with 100+ ms of GD @ 20 Hz plus 6 ms of "DSP" delay and strong pre-ringing it is no wonder DIRAC couldn't figure it out.

 

[dan_remaincalm]

I wish I knew what I know now about group delay and pre-ringing caused by DSP for these "too small" subs that try to go to 20 Hz.

similar here, although subjectively it sounds pretty good now. i know it could be a lot better though.

over the years, i've experimented with sonarworks, dirac, a bunch of other stuff, but i've settled on hand tuned minidsp IIR+FIR + delay informed by REW data, plus whatever is running native in the neumann, and equalizerAPO for some zero-latency headphone stuff that doesn't get routed to the minidsp. no processing above 1kHz apart from a very gentle treble tilt.

(i re-ran dirac on my tuned system and it didn't really add anything useful)

when i got these SVS subs they were genuinely a huge step up from what i was previously running. but now i have all this other hand-tuned DSP i kinda wished i had some pure analog ones so i could do all the integration by hand and get it perfect. like - these are sealed, so the low end behaviour should be pretty clean, but all the DSP makes the phase relationship go wonkier than a ported analog sub. would cost me another couple of grand to get anything meaningfully better, though.

 

[levimax]

similar here, although subjectively it sounds pretty good now. i know it could be a lot better though.

over the years, i've experimented with sonarworks, dirac, a bunch of other stuff, but i've settled on hand tuned minidsp IIR+FIR + delay informed by REW data, plus whatever is running native in the neumann, and equalizerAPO for some zero-latency headphone stuff that doesn't get routed to the minidsp. no processing above 1kHz apart from a very gentle treble tilt.

(i re-ran dirac on my tuned system and it didn't really add anything useful)

when i got these SVS subs they were genuinely a huge step up from what i was previously running. but now i have all this other hand-tuned DSP i kinda wished i had some pure analog ones so i could do all the integration by hand and get it perfect. like - these are sealed, so the low end behaviour should be pretty clean, but all the DSP makes the phase relationship go wonkier than a ported analog sub. would cost me another couple of grand to get anything meaningfully better, though.

Out of frustration and curiosity I just ordered one of these https://shop.gsgad.com/products/roundover-full-marty-single-unit. Cheaper than my SVS SB-3000 but of course I have to put it together and finish it . Even though it is ported it has only about 20 ms of GD at 20 Hz and has considerably higher SPL and plays lower. It is also huge....

 

[Digital_Thor]

If you remove the high-Q issues with sub-placement, you do not need that many taps. The taps are more about resolution, and you can go by with not that many of them for group delay because it's not that critical anyway. Ballpark is good enough.

I think this is a VERY important point!
It should always be the main goal to correct as much as possible with acoustics, system setup and choice of subwoofers/loudspeaker - before you add any type of digital room correction - even though some can be heard more or less than others. I personally often experience that excessive digital room corrections sound "flat" and "wrong" when the basic design of the speaker in question is more or less ignored, and an overly fixated focus on a given house-curve in the MLP has been chosen over everything else.
Read somewhere that a linear distortion, like an uneven response of a driver, can be corrected at line level with any type of EQ (IIR, FIR-DSP) - because this is a 2-dimensional problem - I think reading about Geddes and Toole, inspired me to this approach - if I recall correctly.
But as soon as the speaker design, room and listening position is taken into the calculations, then we enter the 3-dimensional playing field - and no type of DSP can correct this, optimally. Only going back to the start, and looking at the speaker design, room acoustics and basic system setup, will give you a better result.
So, we can jump fences all we want, but let's be honest about the compromises we chose to initially have with the overall sound setup, budget, space, preference, looks, etc. , before we try and move forward with a solution - IMO

 

[Digital_Thor]

Out of frustration and curiosity I just ordered one of these https://shop.gsgad.com/products/roundover-full-marty-single-unit. Cheaper than my SVS SB-3000 but of course I have to put it together and finish it . Even though it is ported it has only about 20 ms of GD at 20 Hz and has considerably higher SPL and plays lower. It is also huge....

This is something I learned the hard way. I like closed boxes, with all the compromises included. But I had to acknowledge my abilities and preference, and building a proper reflex box can be quite difficult - and they are always bigger - difficult to hide