The less is more approach to hi-fi

[lindsayt]

There's a school of thought that when it comes to hi-fi; less is more.

The idea being that the less there is in the signal path, the less there will be to muck up the music.

Here's a few examples of how that might be applied:

Passive crossovers instead of active ones.

First order crossovers (with possibly no crossover on the bass driver) instead of 2nd, 3rd, 4th or higher order crossovers.

Passive pre-amps (or even no pre-amp) instead of active.

No tone controls.

No DSP.

Fewer number of amplification gain stages.

For analogue recordings, listening to master tapes instead of the vinyl versions.

I think there's a lot to be said for this philosophy, even though there is a big fat "it depends" as to how well this approach will work for any particular set of circumstances.

 

[DocG]

I'm convinced that changes in the digital domain cause less problems than changes to an analogue signal (and no, I have no EE background to back that up with evidence; it's an assumption, if you like).

So I want to go: source -- DSP (XO, driver correction, room correction if need be) -- DAC -- wire -- Power amp -- wire -- driver.

Thoughts?

 

[lindsayt]

My thoughts on DSP is that it's all a big fat "it depends".

Here's something for you to ponder. Why do you need DSP in your system? Or why do you think your system will benefit from DSP? Or what shortcomings in your system will be addressed by DSP?

 

[DocG]

My thoughts on DSP is that it's all a big fat "it depends".

Here's something for you to ponder. Why do you need DSP in your system? Or why do you think your system will benefit from DSP? Or what shortcomings in your system will be addressed by DSP?

Of course, it depends!

Now, why do I need DSP?

- I need cross-overs (subwoofer to mid/bass and mid/bass to tweeter)

- No phase issues, that passive, higher than first order XOs inherently have

- No driver break-up nastiness, that's often seen with 1st order XOs (it all depends, I know *wink*) -- I can use a high order XO, without phase issues, to keep the drivers in their 'comfort zone'.

- the DSP corrects for phase, timing and amplitude issues inbetween the drivers.

- I might benefit from some room correction (to be assessed when everything is ready); the living room is rather sparsely furnitured [if that's a word in English]; curtains and an acoustic ceiling will be all the physical damping there is.

And I get a straight wire from DAC to power amp and from amp to driver...

I'm not claiming the absolute truth. I just think that, in my situation, this could make for a performant, WAF-friendly system, that allows me to combine the bits that I like.

 

[lindsayt]

Playing devils advocate here, and taking a less is more view of things:

Of course, it depends!

Now, why do I need DSP?

- I need cross-overs (subwoofer to mid/bass and mid/bass to tweeter)

If the frequency response of your lowest frequency driver dropped off at an appropriate rate, there would seem to be no particular need to have any crossover for that driver.

- No phase issues, that passive, higher than first order XOs inherently have

As mentioned in the first post of this thread, a less is more approach would have first order crossovers and therefore no crossover phase issues.

- No driver break-up nastiness, that's often seen with 1st order XOs (it all depends, I know *wink*) -- I can use a high order XO, without phase issues, to keep the drivers in their 'comfort zone'.

Yep, as you say "it all depends"

- the DSP corrects for phase, timing and amplitude issues inbetween the drivers.

Amplitude issues.

I will take 1 example. Devialet Phantoms. They use DSP to pump a lot more power into the mid-bass drivers at the lowest frequencies than they do at the higher frequencies. The result is that the amplitude has been adjusted and they produce a relatively flat frequency response down to 20 hz.

The trouble is, at generous volumes you can see the drivers flapping in and out. The bass quality is terrible. They make an amorphous bass noise instead of something that bears a good resemblance to the bass instruments being played.

I'd rather listen to a system with an over lean bass than one with bass like that of the Phantoms.

- I might benefit from some room correction (to be assessed when everything is ready); the living room is rather sparsely furnitured [if that's a word in English]; curtains and an acoustic ceiling will be all the physical damping there is.

This would seem to be a problem of having speakers that aren't well matched to your room / a room that isn't well matched to many (if any?) speakers.

The less is more approach would aim to minimise these issues in the first place.

And I get a straight wire from DAC to power amp and from amp to driver...

Where's your volume control?

I'm not claiming the absolute truth. I just think that, in my situation, this could make for a performant, WAF-friendly system, that allows me to combine the bits that I like.

Yes, you may well be right. DSP might be the best solution for you.

 

[jjbomber]

Less technical analysis gives more enjoyment of the music.

 

[DocG]

Quoting gets difficult, using an iPad...

My lowest frequency driver is a single 15" sealed subwoofer. It will easily go up to 200 Hz, I wager. Which would cause directionality issues. Unless I misunderstood your statement. And I obviously need XOs for the other drivers...

I understand that 1st order XOs cause no phase problems, but they can induce break-up noise, and I read somewhere that they're prone to beaming. A DSP XO has neither of these issues.

Amplitude issues. I won't need Phantom wizzardry in the bass, since I'll be using a Rythmik 15" sealed subwoofer. I think it'll go low enough without pumping up anything. I just want an even response in the whole frequency range.

And I might have to correct some bumps induced by the room. Maybe I will, maybe I won't. My wife and I are into the clean, scandinavian style of decoration (less is more and all that *smile*), so no open bookshelves, no rugs, ... The acoustic, suspended ceiling helps a lot though, so we'll see what is needed.

Where's the volume control? Guess where. It's in the DSP! It's a 35 bit chip, so plenty of bits for volume control before it starts degrading my 16bit music signal. And I can adjuste the gain in the DAC, if needed.

And yes, using DSP is a choice here. I was going for Magnepans, modified by Peter Gunn (with a top notch first order passive cross-over), but after long discussions with PG, I finally changed my mind -- he was less than happy back then...

Let me be clear, I'm very much open to other solutions, and believe they can sound excellent! Could you give a concrete example of 'less is more' as you see it?

 

[DocG]

Less technical analysis gives more enjoyment of the music.

How true!

Now the trick is to do the technical analysis beforehand, and forget about it all once the system is ready to start playing music! *yahoo*

 

[lindsayt]

DocG, would your system sound better with a simple analogue active crossover or with a custom made passive crossover? Possibly. Possibly not.

The system that I've used with the most "less is more" approach has been a Denon C630 CD player into Korneff Clone 45 SET power amp into EV Sentry III speakers. If I swapped the Denon for a Beresford type DAC it'd be even more minimalistic.

 

[nopiano]

Hi Lindsay, I think I agree with the premise more than your examples. May I explain?

Simple items of equipment connected with basic wires, installed with a bit of care and appropriate to the room, can give great results. Single, full range drivers would be great, but in the real world two-way is more realistic to make. Bigger woofers are needed for bass that approaches live music, but a credible facsimile can be provided by much smaller units - say 8" or smaller, as has become normal these days. More complex designs rarely work as well in my experience. .

We see in this forum, and on others, folks who get mired in continual upgrades and a type of permanent discontent. That's a shame when there is so much music to listen to, and not just at home, but live too.

Chasing ever better definition, deeper sound stages, more bass slam, more organised treble, better prat, or whatever, is a dream created by journalists, and marketing departments whose job is to sell advertising space. It hasn't got much to do with the latest amp, DAC, or cables.

 

[DocG]

It's all about choosing where to compromise, I guess.

When back in the day I tried the Eclipse TD712z speakers, they sounded really engaging with small jazz combos and chamber music (full range driver, no cross-over), but the lack of bass and higher treble made them something of a one-trick-pony...

 

[plastic penguin]

There's also another school of thought: Just listen to the music.

 

[MajorFubar]

there sees to be a worrying school of thought that DSP will fix just about everything. I'd love to hear what it makes of my living room, where the speakers sound different depending on where you stand or sit. There's no one version of the truth in my living room, because it's just that: a living room, with papered walls (not painted paster at least), probably not enough soft furnishings and too many relective surfaces. All a load of baloney if you ask me; DSP is just the new buzzword. If you're serious about getting your room to sound right, get a pro in and invest in some room treatment instead of expecting to be able to plonk your hifi in a room with livelier acoustics than a cavern and expecting electronics to fix it all for you. If you want to draw a quality painting you start with a good canvas, not some old shite stuff and hope the more paint you throw at it the less the cracks will show. My tuppence worth at least.

 

[DocG]

there sees to be a worrying school of thought that DSP will fix just about everything. I'd love to hear what it makes of my living room, where the speakers sound different depending on where you stand or sit. There's no one version of the truth in my living room, because it's just that: a living room, with papered walls (not painted paster at least), probably not enough soft furnishings and too many relective surfaces. All a load of baloney if you ask me; DSP is just the new buzzword. If you're serious about getting your room to sound right, get a pro in and invest in some room treatment instead of expecting to be able to plonk your hifi in a room with livelier acoustics than a cavern and expecting electronics to fix it all for you. If you want to draw a quality painting you start with a good canvas, not some old shite stuff and hope the more paint you throw at it the less the cracks will show. My tuppence worth at least.

Yes, that's very true. I think with DSP you can make your music sound a lot better in your sweet spot. If the DSP has a lot of work, all the worse it will sound elsewhere. If you sit right against the backwall, your DSP will probably cut the bass (to compensate for the boundary gain), so in the rest of the room, you probably get a rather anemic sound. And there are problems a DSP can't cope with (a long reverberation time to name but one).

DSP for driver correction is a different proposition altogether. If you allign timing and phase, it will sound better full stop.

 

[lindsayt]

Unless the process of aligning the timing and phase of the drivers results in sonic compromises in other areas - eg the addition of a small amount of hash to the signal from passing it through a load of additional circuitry. In which case it would sound better in some ways and worse in others.

 

[ID.]

Unless the process of aligning the timing and phase of the drivers results in sonic compromises in other areas - eg the addition of a small amount of hash to the signal from passing it through a load of additional circuitry. In which case it would sound better in some ways and worse in others.

What is this "hash" you always speak of? I assume it's an audible artifact, but can you link it to some measurement of distortion, etc.? Or is it another one of those that can't be measured but the listener sure as hell knows it when they hear it?

Is it one of those dragons we chase in hifi like blacker blacks, less grit, more something something?

 

[lindsayt]

ID, to a large extent it's empirical. And at the end of the day, any system "less is more" or "more is more" sounds like what it sounds like.

I would urge every person reading this, that's seriously into hi-fi, to listen to a range of less is more systems in direct A/B demos against a range of more is more systems.

The audible compromises from more is more might be measureable, or they might not.

I'll take 1 example. Let's say we add a component to our signal path between the pre and power amplifier that adds 4 op amps to the signal path. And that this component uses Analog Devices AD797 op amps. Data sheet here: http://www.analog.com/media/en/technical-documentation/data-sheets/AD797.pdf

The voltage through these op amps will depend on the volume that we're listening at any particular instant. If we have speakers that are capable of playing at 110 to 120 dbs - which the vast majority of serious high fidelity speakers can - we could have a mind boggling vast variation in the voltage going through or AD797 op amps. Depending on whether we are listening to low level details at low volumes in a quiet room or listening at night-club deafening levels.

Looking at figure 13 (and using extrapolation of the graph beyond the left hand edge) of the AD797 data sheet, these op amps can not cope with such a wide range of voltages without giving us an audible amount of THD+N at some of our listening volumes.

 

[ID.]

Interesting. Thanks for taking the time to provide that explanation, much appreciated.

 

[nopiano]

Interesting. Thanks for taking the time to provide that explanation, much appreciated.

I'm not sure if Lindsay will bear this out, but I believe it is plausible that some background 'white' noise (a bit like the noise between FM stations with the muting turned off) can benefit the sound perception for many listeners. That might also be one of the reasons that LP replay can sound more 'musical'.

 

[bluedroog]

I don't fall in to any one camp, it is just about what is appropiate for me. I was a staunch advocate of active pre-amps but have recently become won over by passives. It entirley depends on the context though.

I have previously taken taken both a DAC and pre-amp out of the chain and gone with a SBT direct to active speakers (very minimalist) and was pleasently surprised but it was a £250 DAC.

I think at times people can assume less boxes is more simply but in terms of what is going on often you'll just be by-passing internal fuctions such as a built in DAC by using a stand alone unit.

 

[ID.]

Interesting. Thanks for taking the time to provide that explanation, much appreciated.

I'm not sure if Lindsay will bear this out, but I believe it is plausible that some background 'white' noise (a bit like the noise between FM stations with the muting turned off) can benefit the sound perception for many listeners. That might also be one of the reasons that LP replay can sound more 'musical'.

I remember reading about such suggestions in the past. When I think back to my first system it was highly detailed, especially with the decay of notes particularly easy to follow, but there was also always a degree of hiss and I was never sure whether that subtle hiss played a part in giving the impression of more detail. That was a purely digital and transistor system, so nothing to do with vinyl playback.

 

[lindsayt]

Having active speakers may be minimalistic in terms of box count. They are most definitely not in the "less is more" philosophy for having the least amount of stuff in the signal path, inparticular ones with high order analogue crossovers.

Here is a circuit diagram for a (stereo pair) analogue 4th order 2 way active crossover:

And here is the circuit diagram for a (mono) first order passive 2 way crossover:

Some speakers, such as Epos ES14's don't even have the L1 inductor. They rely on the natural inductance of the voicecoil combined with the mechanical roll-off of the bass driver.

 

[MajorFubar]

I'm throwing ideas in the ring because this is in no way my specialist, but active systems are essentially quad-amped HiFis with everything in two boxes and where the amps and x-overs have been specifically designed to suit their unique application. Maybe the benefits of that counteract the drawbacks of there being inherrently more electronics. Taken to silly extremes, clearly there comes a point where 'less is more' falls apart as an argument, otherwise some old twin-transister radio with one solitary 'full range' speaker would be the best sounding piece of hifi you could buy. So maybe the truth is it's all about a fine balancing act.

 

[lindsayt]

I totally agree about it being a balancing act.

For a transistor radio the signal comes from the live performance / record player / CD player / DAC, goes through the mixing desk, is passed onto the transmitting station where it goes through the transmitting electronics onto the transmitting antenna, through the atmosphere, into the radio aerial, through the radio tuning circuits, amplifier and speaker. Apart from live performances, that's not really a "less is more" approach.

And the thing with "less is more" is that it's also the quality and suitability of the things in the signal path that contribute towards the total sound quality. For a given manufacturing cost the "less is more" approach does seem to allow the use of more expensive - and therefore possibly better quality - components.

 

[andyjm]

I totally agree about it being a balancing act.

For a transistor radio the signal comes from the live performance / record player / CD player / DAC, goes through the mixing desk, is passed onto the transmitting station where it goes through the transmitting electronics onto the transmitting antenna, through the atmosphere, into the radio aerial, through the radio tuning circuits, amplifier and speaker. Apart from live performances, that's not really a "less is more" approach.

And the thing with "less is more" is that it's also the quality and suitability of the things in the signal path that contribute towards the total sound quality. For a given manufacturing cost the "less is more" approach does seem to allow the use of more expensive - and therefore possibly better quality - components.

It is very difficult (impossible?) to obtain music through a modern signal chain (not analogue tape or vinyl) that hasn't been through DSP at some stage in its travels. Even vinyl of recent vintage will have gone through a digital desk, or digital limiter or some sort of DSP. Worrying about a digital filter in a DSP crossover is locking the stable door after the horse has bolted.