Does DSP belong in State of the Art Systems?

[Mike Lavigne]

To make this less controversial lets put to one side analogue sources.

why? it's what is most important to optimize. it's where 'suspension of disbelief' occurs frequently.

So lets say your digital source is something like JRiver running on a dedicated PC and you are using JRiver’s own convolution engine to apply your combined crossover, time domain and room correction filters derived from AudioLense or Acourate. Absent the DSP filters you would still have had DA conversion, so I can only see upsides from using DSP, and those upsides are far from trivial.

let's say you have a Wadax Reference dac and server and don't want to 'dumb them down'? in that case digital is just like analog. messing with it takes you the wrong direction. the dsp engines are not even close to as capable as the Wadax.

OTOH i agree that there is a case where the level of digital playback is complimented by dsp in most cases. but where is that level? i can't really say where it is. or if this direction is relevant to higher level digital playback?

A two channel system that doesn’t need fixing? Let me know when you come across one of those.

no comment; other than if you come and listen with me then you tell me what i need to fix. from what you hear.

then allow me to hear a system that you consider is fixed, for me to understand what your reference is. or maybe it's just an idea, and not a reality.

of course; system performance is much more than frequency response and have almost infinite reasons for their performance. and really almost impossible to compare rooms and systems.

 

[schlager]

Mike, I live in Denmark, you are always welcome, haha.

 

[Mike Lavigne]

Mike, I live in Denmark, you are always welcome, haha.

as are you here, east of Seattle.

 

[Atmasphere]

I've used DSP a lot when mastering recordings. I've found that the less you use, the better the recording will turn out unless it was a mess to begin with! So for digital releases I use only normalization, which isn't a thing in home audio. It seemed to have no downside as opposed to DSP compression and EQ, which always seemed to degrade the result.

When mastering LPs, I would ask the producer if they had a non-DSP version of the original master file; in that way it was possible to master an LP that sounded better than the digital release: more natural and greater dynamic range (since most digital releases are compressed for better results in a car).

On this account I suspect that DSP will have a sonic fingerprint in a home system as well. However, the place where its least likely to do harm is in the bass. I don't see this worth pursuing unless you already have a Distributed Bass Array since DSP can do nothing about cancellation due to standing waves except eat up amplifier power without results.

But if you have a DBA, then you could apply DSP to it and really get the bass right in the room. So it could be inserted only between the subwoofer amplifiers (whether built-in or not) and not the main amps driving the main speakers. Food for thought...

 

[schlager]

If you just grab the high SN DAC from the ASR measurements then you will get something technically good measuring but listeners would disagree that it sounds very good.
Well, then something else is terrible wrong in the system. 99 % it will be speaker/room related.

 

[Hear Here]

Hera here, using Stereophile as a guide won´t necessarily yield good results. I only use them for the measurements.

Ah - I think that's where we differ when choosing speakers. I have little interest in measurements as long as the speakers offer the life-like excitement factor I like to hear. I wouldn't take measuring equipment with me when I listen to live music and I accept that no concert hall will offer perfect sound at every seat, so I'm not really concerned if my speakers' responses left to right are not quite 100% identical.

The Stereohile Avantgarde Uno review was what prompted me to seriously consider this bizarre-looking system that I'd always previously passed by at HiFi shows. Robert Deutsch's excellent description (especially if reading between the lines) convinced me that this was where I wanted to go after a real disappointment with my short-lived ATC Active 50s. A quick listen at a most unsuitable London showroom convinced me to place an order.

Absolutely no regrets and since upgraded to Duos and now Duo XDs, by way of unsuccessful excursions into electrostatics. The Unos were great at my London flat (350 sq ft 5 sided room) and later in my present 945 sq ft open-plan almost semi-circular room. Engaging Dirac definitely spoils the excitement factor, even though my version adjusts only sub 500 Hz, so I don’t engage it. Horns seem to be far more tolerant of peculiar-shaped rooms than most other types. I will do REW measurements soon that will allow me to adjust the bass response using the XD processor built into the bass amplifier. I'm sure that will improve the sound, but it leaves the precious mid and top unprocessed.

 

[schlager]

I like the life-like excitement factor as well and still play on horns, just not conventional ones. Keeping horns like Avantgarde and the listening position well away from side walls and back wall, can give a nice result. Or you could damp most surfaces with at least 4" (10cm) absorbers (insulation) and toe in the speakers, to keep 1. side reflection at bay. That will give you a high ratio of direct sound compared to reflections. Sorry for the OT.

 

[schlager]

I'm not really concerned if my speakers' responses left to right are not quite 100% identical.
You should care, because that is an important part of the combined sound in the SS and the sense of center imaging.

 

[ecwl]

Absolutely no regrets and since upgraded to Duos and now Duo XDs, by way of unsuccessful excursions into electrostatics. The Unos were great at my London flat (350 sq ft 5 sided room) and later in my present 945 sq ft open-plan almost semi-circular room. Engaging Dirac definitely spoils the excitement factor, even though my version adjusts only sub 500 Hz, so I don’t engage it. Horns seem to be far more tolerant of peculiar-shaped rooms than most other types. I will do REW measurements soon that will allow me to adjust the bass response using the XD processor built into the bass amplifier. I'm sure that will improve the sound, but it leaves the precious mid and top unprocessed.

So at my stereo dealer’s store, I’ve heard multiple systems using Dirac. I’ve heard some amazing sound and some where as described, Dirac robs the “excitement factor”.
Having played around with time-domain correction myself, I learnt that it’s very easy to over correct or mis-correct for time domain issues and cause loss of excitement because all corrections are a bit of a compromise because you don’t have active crossover so you’ll never be able to align your tweeter with your midrange and woofer and subwoofers perfectly. Hence I think if people are unhappy with Dirac, it’s likely that the DSP is suboptimal, rather than DSP not being suitable for a state of the art system. The challenge is that if your time-domain Dirac filter is suboptimal, there is no easy way to manipulate it further other than to change the frequency response target curve.

 

[schlager]

Dirac filter is suboptimal
Never put tools in the wrong hands, that is the road to hell, but please don´t blame the tool. You should probably "upgrade" your software and skills.

 

[ecwl]

Dirac filter is suboptimal

Never put tools in the wrong hands, that is the road to hell, but please don´t blame the tool. You should probably "upgrade" your software and skills.

Fair enough. It’s been a while since I’ve used Dirac and I definitely haven’t tried the latest version. In the past, Acourate or Audiolense has many more options and features to adjust the time domain correction compared to Dirac. It is possible this has changed.
But it’s also true that sometimes my dealer gets lazy and didn’t do the Dirac measurements properly and got a bad result. After all, garbage in garbage out. But I just don’t want to be misquoted. I said sometimes the Dirac filter can be suboptimal. Whether that’s fixable would depend on the circumstances.

 

[Hear Here]

Dirac filter is suboptimal
Never put tools in the wrong hands, that is the road to hell, but please don´t blame the tool. You should probably "upgrade" your software and skills.

I wonder how many orchestral musicians take their instruments to a lab to get them measured?

How many piano tuners use electronics to do their job?

Why should the instruments we use to provide music need to be tested in the way that the instruments that created the original music don't need to be so tested?

Bit of a leg pull, but it does need thinking about. In the end, it's only the sound that matter if it gives pleasure to our ears.

 

[schlager]

Fair enough, have no intention to misquote you. Dirac is probably to constrained in terms of tweaking. Different time windows (FDW) in both frequency and time domain, will have quite an impact on perceived sound quality.

 

[Atmasphere]

I wonder how many orchestral musicians take their instruments to a lab to get them measured?

How many piano tuners use electronics to do their job?

Why should the instruments we use to provide music need to be tested in the way that the instruments that created the original music don't need to be so tested?

Bit of a leg pull, but it does need thinking about. In the end, it's only the sound that matter if it gives pleasure to our ears.

This analogy doesn't hold up! FWIW though electronic tuning has been a thing with musicians since the tube era. Conn used to make a tube tuner and Korg has made ones popular with guitarists for decades. Most piano tuners use them as well.

But the real point where measurements help out is when the engineer sets up the microphones... There's an enormous difference between what musicians do with instruments as opposed to how they are recorded!

 

[schlager]

In the end, it's only the sound that matter if it gives pleasure to our ears.
Subjective bable, that leads nowhere, at least not in any progressive way.
Instruments are instruments, not a reproducing device. Speakers are reproducing devices, not instruments, see the difference? And by the way, I tune my guitar for a clean A 440 hz, not by ear, but using a digital tuner.

 

[schlager]

Microphones captures everything we need to know, ears not so much.

 

[Hear Here]

Microphones captures everything we need to know, ears not so much.

When I go to concerts there's normally no sign of a microphone unless the concert is being recorded and no sign of any loudspeaker. Violins built in the 17th century are the absolute best and, as far as I'm aware, that was long before any mic or speaker had hit the light of day. Same goes for most acoustic / orchestral instruments. All created with only one aim in life - to offer ultimate pleaseure to the ear - not to the microphone.

PS - Still a bit of a leg pull, but I'm simply making the point that measurements are secondary to the music, though of course accurate transducers are essential.

 

[godofwealth]

Microphones captures everything we need to know, ears not so much.

Ugh, I prefer my ears any day of the week compared to the absolute dreck of sound that comes from most microphone feeds. I’ve been attending live concerts for 35+ years and collecting CDs etc. for the same length of time. There’s just no comparison between the beauty of live sound and the pale shadow of live music that is in the bits of a CD or a flac stream.

 

[Kal Rubinson]

Bit of a leg pull,

Yes.

Instruments are instruments, not a reproducing device. Speakers are reproducing devices, not instruments, see the difference?

There we go.

 

[SeagoatLeo]

So it´s the ADA that is troublesome? That problem was solved many years ago. Transparent DAC´s are easy to get hands on, also cheap ones.
Most speakers also expensive ones, are far from linear, compared even to the cheapest audio gear.
Let´s take a look at your Avantgarde Uno, I believe.

View attachment 100917
Not that bad, but could still need some top end eq. This will sound a bit bright. You don´t fix that with cable swapping or another amp. You tackle the problem by the root, fixing the FR with EQ.
Dirac Live have full FR EQ, probably not FIR above 500 hz. As I said only a handful of software does this correctly, Audiolense, Acourate, DRC Designer, Trinnov, comes to mind. They can be tweaked endlessly, to suit your speakers, room, listening distance and what not (taste of the day).

I worry about the 130 to 300 Hz 5db dip. That's really concerning.