[flyte3333]

Yes let's see what Rob thinks (questions were to him anyway).

But I'm not sure how talking about nano seconds of timing differences in the digital domain, and some 300dB down (now 400dB down is being mentioned)

When in typical room with passive speaks, bass arrives milli seconds after tweeter...

Headphones don't have this issue

And time coincident DSP speakers also shouldn't have this issue

 

[Rob Watts]

Hi Rob, this passage below was from Dave thread but I don't want to de-rail Dave DAC discussion , so asking here in your generic non-product specific thread.



If phase is this important and audible, are you doing phase correction in your B&W 802 system?

And are you doing driver time alignment?

As you can see from your speakers step response measurements, each driver has sound arriving at different times, with milli seconds delays...

If you're not doing this, why not , if timing is so critical ?

Or have you by-passed the passive crossover and doing your own phase correct with driver time alignment?

Obviously these questions/ discussion relates only to speaker systems, not headphones.

Cheers!

https://www.stereophile.com/content/bw-802d-loudspeaker-measurements

Actually, it is not phase shift itself that is important - and it's not that that I measure - but phase shift with amplitude. So I run an FFT of my digital modules, with an input at 0dBFS, and record the absolute value of amplitude and phase at 6kHz. Then I take my 6kHz signal and attenuate it by 301.0299957 dB, to create my -301.030dBFS signal. Then I re-run the FFT, and for an ideal module the OP should be identical phase (i.e. no phase shift with amplitude) to the 0dBFS figure. For the module to pass, it needs to be accurate to +/-0.001dB and +/-0.001 deg of phase.

Listening tests ran before indicates that if the phase changes with amplitude, it will degrade the perception of depth. I have no evidence that a phase shift that is constant (that is does not vary with amplitude or time) degrades depth perception; indeed, going into differing acoustic environments (with huge phase and frequency changes) does not damage ones ability to determine depth perception. The brain is adept at recognising the acoustic footprint of the space one is in, then subtracting that footprint from depth calculations.

So it's not the absolute value of timing that is important, but that the timing does not vary with amplitude (giving small signal errors) or with sampling (giving transient timing errors). The issue here is that vanishingly small errors that change with amplitude or time is vitally important subjectively.

 

[The Jester]

Interesting,
What I discovered when I first got my M Scaler/Qutest and decided to go through the setup routine again that a lot use for setting up Sonus Faber speakers, including rake angle via adjustable spikes, with my previous DAC the L/R position, toe in etc was relatively easy but spent a lot of time listening and adjusting rake angle, when that is “just right” the speakers open up and disappear acoustically, with the Chord front end I had it “dialled in” within half an hour … adjust, listen, repeat and it wasn’t hmm, does this sound better, more like someone threw a switch and there it was ..

 

[ecwl]

I have to admit I do have phase correction in my convolution filter for my speakers and I like the sound. But I don’t know how much of the correction improvement is due to the change in phase.

but I can also say that the convolution filter itself also causes a loss of depth perception in the playback. Whether this is because the convolution algorithm causes a loss in bit accuracy, I don’t know for sure.

 

[The Jester]

Maybe why some decide against phase correction in the crossover which will add more components in the signal path, and go with front rake angle or multiple staggered boxes for each driver and risk aesthetic (and WAF) concerns …

 

[flyte3333]

I have to admit I do have phase correction in my convolution filter for my speakers and I like the sound. But I don’t know how much of the correction improvement is due to the change in phase.

but I can also say that the convolution filter itself also causes a loss of depth perception in the playback. Whether this is because the convolution algorithm causes a loss in bit accuracy, I don’t know for sure.

Are you attempting phase correction with passive crossover speakers?

 

[ecwl]

Are you attempting phase correction with passive crossover speakers?

No. You’re correct. Aggressive phase correction at the crossover with passive speakers would lead to other sonic artifacts that would be unpleasant (I believe audible pre-ringing being one of them). I see your point.
That said, I still far prefer Chord DAC driven system than active DSP speakers that I’ve heard, e.g. Linn, Meridian, Kii, etc.

 

[flyte3333]

That said, I still far prefer Chord DAC driven system than active DSP speakers that I’ve heard, e.g. Linn, Meridian, Kii, etc.

You've jumped from one extreme to the other.

What about something in the middle?

What about a DSP crossover with a few Chord DACs feeding each speaker driver individually?

Something like this keeping them in sync?

If you can get a few 2nd hand Mojo's, could be fun experiment - until Rob's multichannel interface comes out

Heck I might try it !

 

[flyte3333]

Actually, it is not phase shift itself that is important - and it's not that that I measure - but phase shift with amplitude. So I run an FFT of my digital modules, with an input at 0dBFS, and record the absolute value of amplitude and phase at 6kHz. Then I take my 6kHz signal and attenuate it by 301.0299957 dB, to create my -301.030dBFS signal. Then I re-run the FFT, and for an ideal module the OP should be identical phase (i.e. no phase shift with amplitude) to the 0dBFS figure. For the module to pass, it needs to be accurate to +/-0.001dB and +/-0.001 deg of phase

Thanks Rob. All of these impressive numbers are purely in digital domain (you also mention digital modules).

My questions were about in the analogue domain... like with your B&W 802D with passive crossover network...

Wouldn't nano seconds discussions in the digital domain get easily dwarfed by milli seconds 'stuff' in the analogue domain?

Not with headphones or full-range driver speakers. I'm moreso talking about with multi-way passive crossover speakers.

 

[The Jester]

The answer was in Rob’s last sentence, the brain can accept slight errors that are constant and having even slight errors constantly changing is the issue, so with a passive speaker and the listener sitting still how can the slight timing error between drivers change ?

 

[flyte3333]

Deleted

 

[All Day Breakfast]

Why dont you set the display to automatically turn off? Display option 4. It will turn back on whenever you alter volume etc.

Just came across this post. Thanks for this info. The manual is totally useless for the DAVE. @Rob Watts Why isn’t this explained in the manual?

 

[Rob Watts]

Just came across this post. Thanks for this info. The manual is totally useless for the DAVE. @Rob Watts Why isn’t this explained in the manual?

It does - see page 34 on manual on Chord's website. I don't write the manuals, and its not an easy job either!

Happy listening.

 

[All Day Breakfast]

It does - see page 34 on manual on Chord's website. I don't write the manuals, and its not an easy job either!

Happy listening.

Thanks, I stand corrected. I love the DAVE and how my music sounds now. Thanks for that!

 

[feddar]

Electronics are more reliable if left on. If you are worried about the 12W or so power consumption, then use the standby - this keeps the clocks running. I keep my TT2 and Dave on all the time.

Can I ask what you mean by ‘more reliable?’

I only plug in my HMS and TT2 when I use it. I worry about power surges in the house.