Compensating for mains and sub delay

[ErLan]

Hello ASR members,

I would like to get the forum perspective (and hopefully tips) on the following.

I am using a fairly streightforward signal path:

PC (Intel NUC, source: Tidal, Spotify, EqAPO) >> DAC (Topping E50) >> Sub (Whearfdale WHD10, active, including crossover 100Hz) >> Amp (Behringer A800) >> L+R Mains (Kef LS50).

Unexpectedly, REW measurements reflects a ~20msec delay between main and sub (mains prior sub).

In the past weeks I familiarized myself with solutions like Multichannel dacs/ miniDSP / DBX / Muto m4 /.... however each is quite expensive in my area (400-600€).

Prior deciding to pull the trigger on (yet )another HiFi investments. I wonder is there any other alternative that may help to resolve this imperfection in an elegant and cheaper way (preferably in the <100€ range).

Many thanks!

 

[dasdoing]

well, why not just delay in EQ-APO which you allready use?

 

[pseudoid]

Simply stupid question: Are you 100% certain that all of the 20mS delay is due SOLELY to the WharfEdale subs?

 

[ErLan]

Simply stupid question: Are you 100% certain that all of the 20mS delay is due SOLELY to the WharfEdale subs?

Most likely, as got similar numbers also with other set up of speakers; also, prior current set up I used an AVR where I had control on each speaker delay/distance. Compensating sub by this number worked like a charm.
Anyway, what other 'sources' do you jave in mind? (Path is fairly minimalistic)

 

[ErLan]

well, why not just delay in EQ-APO which you allready use?

The output from the PC is L+R, while crossover is done on the sub (downroad). Adding a Delay via EQAPO at the beginning of the chain will lead to a delay in the signal's input to the sub as well hence will keep the relative delay between mains and sub as is.

 

[Chrispy]

Why is it surprising you have delays inherent in the subs and don't have gear to compensate for such?

 

[pseudoid]

Have you considered replacing the WHD10 with one of the KEF recommended (hopefully complementary to LS50) subs?

 

[radix]

You could try a Y cable and make sure it's the sub that's adding delay. 20 msec seems large.

Do you have 0* or 180* phase set? Phase could add 6 - 7 msec.

You could likely find a used minidsp DDRC-24 (if you want dirac).

As you play from your PC, a Motu M4 (4-channel dac) plus a convolver on the PC likely makes most sense. You need some software and a measurement mic to create the filters, then use a convolver to get dynamic room correct and equalize phase and delay.

 

[DWPress]

Since you are PC based you need a DAC that can control 2+ individual channels. A simple Minidsp 2x4 will do what you need and pretty close to budget.

 

[dasdoing]

The output from the PC is L+R, while crossover is done on the sub (downroad). Adding a Delay via EQAPO at the beginning of the chain will lead to a delay in the signal's input to the sub as well hence will keep the relative delay between mains and sub as is.

ooh, the subwoofer has high-passed main outs? surprising that it hasn't a delay compensation than. how do you meassure the delay in REW? a good way is to look at excess group delay. in this example the sub is delayed about 25ms:

if your sub response is near the 0 line, you don't have delay

 

[dasdoing]

whatever, this is still solvable in EQ-APO, since you can create a partial group delay inversion (phase correction) in REW.

 

[ErLan]

whatever, this is still solvable in EQ-APO, since you can create a partial group delay inversion (phase correction) in REW.

Thanks, Actually I looked and measured it at the spectrogram/wavelet view which AFAIK aligned with the Group Delay.
Can you further elaborate on (or link me to) the group delay inversion via. REW approach?
Is it the same as described in other threads using RePhase & FIRs?
Assuming FIRs, is not such correction on the low end imposes noteable pre-ringing?

 

[dualazmak]

Hello OP @ErLan,
OK, now you started this new thread on the topic. Let me copy-paste my message from here as follows.

I am looking forward to your new thread on this topic. Please kindly announce here when you would start the new thread.

While waiting for start of your new thread, let me share with you my methods for time alignment measurements and tunings (applying group delay in DSP software EKIO in 0.1 msec precision) which I summarized in my post here on my project thread. As for the time alignment and XO configuration between subwoofer and woofer, my posts here and here would be also of your reference, I assume.

 

[Keith_W]

ooh, the subwoofer has high-passed main outs? surprising that it hasn't a delay compensation than. how do you meassure the delay in REW? a good way is to look at excess group delay. in this example the sub is delayed about 25ms:

if your sub response is near the 0 line, you don't have delay

What is the cause of the peak at 50Hz? I was thinking the crossover looks to be about 80Hz.

 

[OCA]

What is the cause of the peak at 50Hz? I was thinking the crossover looks to be about 80Hz.

It's probably European AC interference.

 

[Keith_W]

It's probably European AC interference.

Does AC interference cause group delay? How? If it was a ground loop it would manifest as a 50Hz hum but I don't know how it would cause group delay? I mean, that's a pretty substantial peak on that excess GD plot, and if there was a hum it would surely be audible.

 

[dualazmak]

For group delay adjustment especially between subwoofer and woofer, I would like to suggest that at first you should not over-trust sophisticated software such as REW and RePhase.

You should first measure and try time alignment by simple and straightforward "room air sound recording and analysis at your listening position" using independent second PC like I did here (time-shift tone burst method) and here (energy peak matching method).

If you would be seriously interested, I will be happy to share all the test tone signals I prepared for these studies summarized in my post here.

Just like @zerxia kindly shared here as well as here, and also I pointed here, advanced DSP software (in his case Equalizer APO) does not always give/measure accurate delays (time alignments) at your listening position.

Of course, you (we) should not mix-up (misunderstand) the "absolute total delay (total latency)" and "relative delay among SP drivers", as I pointed in the beginning of my post here.

As far as I can, I would like to perform the measurement(s) and fine tuning with no or minimum (and understandable) incorporation of "black box" type software manipulation or signal handling, as I have been sharing throughout in my project thread. I believe it is also important that we always need to establish objective "validation" procedures plus evidences for any of the sound measurement methods we would like to apply in our audio system tuning.

By "validation", I mean that if you intentionally delay your subwoofer sound in 25.0 msec, your "measurement and tuning" method should be able to accurately find the 25.0 msec delay in 0.1 msec precision in recorded room air sound at listening position.

Regarding pre-ringing and/or post-ringing, just for your info and reference, I use DSP(XO/EQ/Delay) software EKIO;
- EKIO uses IIR filters; cascade 2nd order direct form II biquad in 64 bit floating point: #138-#142
- EKIO (using IIR filters) gives no audible post-ringing nor pre-ringing at all, at least in my XO/EQ configurations: #143, #181on my project thread.

 

[OCA]

Since you don't have a way of separating subwoofer output in your setup, your only free option is to apply a filter like below in Equalizer APO which will send frequencies below 100Hz around 20ms earlier. Since the filter itself causes a processing delay of 170ms, it MAY not cause pre-ringing. If it does, then you can try delaying above 100Hz by 20ms instead. Let me know if you need any help with that.

Here's the group delay graph after the filter is applied to a Dirac pulse:

 

[dasdoing]

Thanks, Actually I looked and measured it at the spectrogram/wavelet view which AFAIK aligned with the Group Delay.
Can you further elaborate on (or link me to) the group delay inversion via. REW approach?
Is it the same as described in other threads using RePhase & FIRs?
Assuming FIRs, is not such correction on the low end imposes noteable pre-ringing?

It is complicated and requires a lot of experimentation. I plan to do a topic about this, soon. probably this week
you have to restric the window and the frequencies, mostly because of

What is the cause of the peak at 50Hz? I was thinking the crossover looks to be about 80Hz.

this, for example. which is a no minimum phase problem in the room response. in the FR this is a dip.
we can't phase correct these

 

[Ron Texas]

Hello ASR members,

I would like to get the forum perspective (and hopefully tips) on the following.

I am using a fairly streightforward signal path:

PC (Intel NUC, source: Tidal, Spotify, EqAPO) >> DAC (Topping E50) >> Sub (Whearfdale WHD10, active, including crossover 100Hz) >> Amp (Behringer A800) >> L+R Mains (Kef LS50).

Unexpectedly, REW measurements reflects a ~20msec delay between main and sub (mains prior sub).

In the past weeks I familiarized myself with solutions like Multichannel dacs/ miniDSP / DBX / Muto m4 /.... however each is quite expensive in my area (400-600€).

Prior deciding to pull the trigger on (yet )another HiFi investments. I wonder is there any other alternative that may help to resolve this imperfection in an elegant and cheaper way (preferably in the <100€ range).

Many thanks!

Could you show us the measurement?