Kegworth Show Information

The information for this year's show is now live at www.thehifishow.com or by clicking on the show banner below.

NON EXHIBITOR ROOMS NOW OPEN - Special Offer....Stay Saturday, join in all the build up excitement on Sat night and then gain FREE access to the show on Sunday

ONLY 4 EXHIBITOR ROOMS LEFT - BIG SYSTEM SPACES SOLD OUT - ONLY 2 NON EXHIBITOR ROOMS LEFT

insider9

Chasing system accuracy... Is there a point?

Recommended Posts

Moderator
1 hour ago, bencat said:

Difficult question to answer really as for most if not all of us the final link in the chain is different that being our rooms . One thing I did learn from the work I did making active speakers is that measurements do have there value and place they will tell lots of things about how items perform and what changes to those paramters then do to the output sound .

As always though they do not tell you everything and my belief is that they never have . I have heard a speaker and measured and recorded a speaker with a flat response from 10 - 2000 hz twice . Once at a feloow wammers house who was helping me and had some incredibly expensive but effective DSP / Crossover units that could be sett to ensure any speaker fed by them had a perfect flat response . Second was at home after measuring and creating a flat response , this one was only from 20 - 20000 hz as the bass speaker was not stable below this so signal was stopped .

In both cases the sound was flat unimpressive and lacking either dynamics and the main sound cues I like to hear in music . They made Earth Wind and Fire sound boring andthey stipped Otis Redding of his soul feeling . Now the measurements may state that this is a flat reasponse but I have discovered I do not want a perfect flat response . Instead I like many others add what is know as a house curve which I adjust till it sounds right to me , with my equipment and my music in my room . In general I found that my curve was similar to many used by other people and even makers . We tend to push the Bass up slope back to flat in the Mid and then slightly drop off the treble . I have seen lots of variations on the sizes of the Bass increase and the treble drop off and in some cases the mid is not quite flat but the general shape of the curves is the same .

So for me at least I do not and would not ever decry the usefulness and use of measurement in order to get any product to be working in the right area. Once oyu get there though then it needs and requires a human ear to listen to music and finesse the sound to be enjoyable and acceptable to the majority . Not a win for either side . if I was pushed then I would have to say I woud pursue musical enjoyment rahter than perfect measurement as to date I have not heard any perfectly measured system that sounded enjoyable .

I agree - flat is not a desirable end result. One of the articles I came across (a paper by some guys at Harman) on the miniDSP site addresses this issue:

https://www.dropbox.com/s/3ocp8oxi93yaqml/Sean Olive on Target Responses.pdf?dl=0

Conveniently, the miniDSP DDRC-24 software includes a Harman target curve as one of the options for the Dirac correction - I have been experimenting with it and the results are pleasing to my ear at least.

  • Upvote 1

Share this post


Link to post
Share on other sites
33 minutes ago, PuritéAudio said:

Considering reflected sound is one element of sound quality, a loudspeaker with wide even dispersion ( constant dirctivity) will sound less coloured than a speaker with a ragged off-axis response.

EQ can be used to reduce a bass ‘boom’ that is going to hugely improve sound quality, a fine measuring loudspeaker is always preferred in listening tests.

The new actives aim and succeed in making a domestic space behave  more like a fully treated studio.

Keith 

First off, can we use the term Controlled Directivity instead if Constant Directivity because the latter is an ideal that doesn't exist since at some point, even in the best speakers, the LF becomes omnidirectional and the top of the HF still beams. Despite waveguides, cardioids and such. These technologies control directivity to some limit.

Cardioid response to better control directivity and make it more even, as with Kiis and 8C speakers is avoiding early reflection by essentially beaming sound forward in the cardiod pattern. This is not wide dispersion like Toole meant. At Harman, the aim was to ensure that the additive power of reflection and direct sound summed correctly at the listning position. Therefore, a good speaker would result in an in room SPL at the listening point that correlated to the Bruel Kjaer curve, since higher frequencies would still be absorbed to an extent by the room.

The two are very different concepts.

Your comment on wide dispersuon suggests you should be selling Revels, not Kiis or 8C speakers, which have widened HF due to waveguide, but narrowed bass/mid due to cardioid to essentially match at a dispersion narrow enough to avoid early wall reflections.

You can still have a cardioid whereby early reflections are absent, but total reflections and direct sound summed at SPL are correct.

Wide dispersion a la Toole is a design that uses early reflections, while cardioids do not.

So, your last claim that cardioid speakers make a domestic space more like a studio only goes so far as if there are noeaearly reflections, then absorption materials don't need to be used. The rest if the domestic room response is nothing like a studio.

Share this post


Link to post
Share on other sites
52 minutes ago, Tony_J said:

I agree - flat is not a desirable end result. One of the articles I came across (a paper by some guys at Harman) on the miniDSP site addresses this issue:

https://www.dropbox.com/s/3ocp8oxi93yaqml/Sean Olive on Target Responses.pdf?dl=0

Conveniently, the miniDSP DDRC-24 software includes a Harman target curve as one of the options for the Dirac correction - I have been experimenting with it and the results are pleasing to my ear at least.

I am not so sure about flat is not desirable - after all the only way the music can be replayed accurately is with something that does not add or take away.  So if the equipment has any deviance from flat then the recording is being altered.   Now I agree to some that alteration is desirable but to others it is not ... it all comes down to the individual and what their ears tells them they like  :)  

Share this post


Link to post
Share on other sites
Super Dealer
2 minutes ago, Metatron said:

First off, can we use the term Controlled Directivity instead if Constant Directivity because the latter is an ideal that doesn't exist since at some point, even in the best speakers, the LF becomes omnidirectional and the top of the HF still beams. Despite waveguides, cardioids and such. These technologies control directivity to some limit.

Cardioid response to better control directivity and make it more even, as with Kiis and 8C speakers is avoiding early reflection by essentially beaming sound forward in the cardiod pattern. This is not wide dispersion like Toole meant. At Harman, the aim was to ensure that the additive power of reflection and direct sound summed correctly at the listning position. Therefore, a good speaker would result in an in room SPL at the listening point that correlated to the Bruel Kjaer curve, since higher frequencies would still be absorbed to an extent by the room.

The two are very different concepts.

Your comment on wide dispersuon suggests you should be selling Revels, not Kiis or 8C speakers, which have widened HF due to waveguide, but narrowed bass/mid due to cardioid to essentially match at a dispersion narrow enough to avoid early wall reflections.

You can still have a cardioid whereby early reflections are absent, but total reflections and direct sound summed at SPL are correct.

Wide dispersion a la Toole is a design that uses early reflections, while cardioids do not.

So, your last claim that cardioid speakers make a domestic space more like a studio only goes so far as if there are noeaearly reflections, then absorption materials don't need to be used. The rest if the domestic room response is nothing like a studio.

Well no, if you have a waveguide 8C/Kii  then you get excellent dispersion even off axis, likewise if you push the speakers close to a wall there is no bass ‘ reflection’ because a single wave front is presented.

Cardioid is reducing the rear propagation  , thus reducing reflected sound, Harman’sdesigns haven’t included Cardioid yet, although a few do, Geithain and Gradient for example and it really works.

I believe the boundary coupling of the 8Cs produces a  hemisphoidal  propagation which matches the propagation of the cardioid midrange .

The benefit if you compared them directly to your loudspeakers would be immediate and obvious , tighter better controlled bass ( full-range) and greater resolution through the mid range.

Keith

Share this post


Link to post
Share on other sites
Super Wammer
1 hour ago, PuritéAudio said:

Considering reflected sound is one element of sound quality, a loudspeaker with wide even dispersion ( constant dirctivity) will sound less coloured than a speaker with a ragged off-axis response.

EQ can be used to reduce a bass ‘boom’ that is going to hugely improve sound quality, a fine measuring loudspeaker is always preferred in listening tests.

The new actives aim and succeed in making a domestic space behave  more like a fully treated studio.

Keith 

Doesn't wide dispersion increase the amount of reflected sound?

What is meant by a fine measuring loudspeaker? Who's definition of that term do you choose? And for those who choose a different definition of fine measurements come up with a different set of preferences to those that align with yours? The sweeping assertions are less than helpful.

  • Upvote 1

Share this post


Link to post
Share on other sites
1 hour ago, PuritéAudio said:

Considering reflected sound is one element of sound quality, a loudspeaker with wide even dispersion ( constant dirctivity) will sound less coloured than a speaker with a ragged off-axis response.

EQ can be used to reduce a bass ‘boom’ that is going to hugely improve sound quality, a fine measuring loudspeaker is always preferred in listening tests.

The new actives aim and succeed in making a domestic space behave  more like a fully treated studio.

Keith 

First off, can we use the term Controlled Directivity instead of Constant Directivity because the latter is an ideal that doesn't exist since at some point, even in the best speakers, the LF becomes omnidirectional and the top of the HF still beams. Despite waveguides, cardioids and such. These technologies control directivity to some limit.

Cardioid response to better control directivity and make it more even, as with Kiis and 8C speakers is avoiding early reflection by essentially beaming sound forward in the cardiod pattern. This is not wide dispersion like Toole meant. At Harman, the aim was to ensure that the additive power of reflection and direct sound summed correctly at the listning position. Therefore, a good speaker would result in an in room SPL at the listening point that correlated to the Bruel Kjaer curve, since higher frequencies would still be absorbed to an extent by the room.

The two are very different concepts.

Your comment on wide dispersuon suggests you should be selling Revels, not Kiis or 8C speakers, which have widened HF due to waveguide, but narrowed bass/mid due to cardioid to essentially match at a dispersion narrow enough to avoid early wall reflections.

You can still have a cardioid whereby early reflections are absent, but total reflections and direct sound summed at SPL are correct.

Wide dispersion a la Toole is a design that uses early reflections, while cardioids do not.

So, your last claim that cardioid speakers make a domestic space more like a studio, only goes so far as if there are no early reflections, then absorption materials don't need to be used near the speakers. The rest of the domestic room response is nothing like a studio.

  • Like 1

Share this post


Link to post
Share on other sites
Super Dealer
13 minutes ago, sunbeamgls said:

Doesn't wide dispersion increase the amount of reflected sound?

What is meant by a fine measuring loudspeaker? Who's definition of that term do you choose? And for those who choose a different definition of fine measurements come up with a different set of preferences to those that align with yours? The sweeping assertions are less than helpful.

It means that the speakers on and off axis response are the same, of course the off-axis is greatly diminished , colouration is due to the off-axis response notmatching the ‘on’ axis.

I believe what constitutes a good measuring loudspeaker is pretty much understood, read any of John Atkinson’s loudspeaker measurements for Stereophile if you are unclear.

Keith

Share this post


Link to post
Share on other sites
40 minutes ago, PuritéAudio said:

Well no, if you have a waveguide 8C/Kii  then you get excellent dispersion even off axis, likewise if you push the speakers close to a wall there is no bass ‘ reflection’ because a single wave front is presented.

Cardioid is reducing the rear propagation  , thus reducing reflected sound, Harman’sdesigns haven’t included Cardioid yet, although a few do, Geithain and Gradient for example and it really works.

I believe the boundary coupling of the 8Cs produces a  hemisphoidal  propagation which matches the propagation of the cardioid midrange .

The benefit if you compared them directly to your loudspeakers would be immediate and obvious , tighter better controlled bass ( full-range) and greater resolution through the mid range.

Keith

Sorry Keith, your original post and last one still conflates wide dispersion with KIIS and 8Cs. Although you deferred to saying "excellent dispersion" as opposed to wide dispersion. I know Harman speakers don't use cardioid - you've ignored where and how I pointed out they are different.

"I believe the boundary coupling..." And go on to only mention the rear wall behind the speakers. The reason both Kii and 8C can work in a corner is because side wall influence is ameliorated.

If the speaker at the tweeter is 40cm from side wall, but you don't get side wall early reflections you must be dispersing sound narrow enough not to.

This doesn't diminish the fact the waveguide widens HF dispersion to meet that of the bass and mids, but the result is not a wide dispersion speaker in the manner Toole spoke of.

Edited by Metatron

Share this post


Link to post
Share on other sites

For me, it's just about listening enjoyment as there is no way I have ever been able to create a true live feel in a domestic living space. In fact even live music can be a challenging experience in the wrong space or if the sound engineer believes volume equals enjoyment. 

Some of the most musical enjoyment I've had recently has been whilst listening to the radio on a old Denon shoe box thingy whilst working in the kitchen, no deep base and limited everything else but really enjoyable. In fact I often find myself dancing around the kitchen as I work.  

I beginning to believe this search for the perfect sound system is a mild form of madness but unfortunately in my case not one that seems curable.   

  • Like 5

Share this post


Link to post
Share on other sites
Moderator
48 minutes ago, wHIZZY said:

I beginning to believe this search for the perfect sound system is a mild form of madness but unfortunately in my case not one that seems curable.   

For me it isn't a search for the perfect esound, which I agree is a form of madness. Pretty much from the start I have been into DIY, particularly speakers but occasionally amps as well. The combination of building something that not only works (and looks half decent) but also sounds good as well is quite addictive, and quite soon becomes another form of madness...:22:

Unfortunately, the advent of affordable tools for building active crossovers with DSP makes it too easy to build new stuff and provides infinite possibilities for tweaking.

  • Like 3

Share this post


Link to post
Share on other sites
33 minutes ago, wHIZZY said:

For me, it's just about listening enjoyment as there is no way I have ever been able to create a true live feel in a domestic living space. In fact even live music can be a challenging experience in the wrong space or if the sound engineer believes volume equals enjoyment. 

That is true most concerts I have been to the sound balance is not right, many singers end up shouting rather than singing just to be heard. Even at a jazz concert the balance was off, the drums were much louder than the double bass and piano was almost lost. Classical is probably ok but depends where you are sitting.  

Share this post


Link to post
Share on other sites
Super Dealer
1 hour ago, Metatron said:

Sorry Keith, your original post and last one still conflates wide dispersion with KIIS and 8Cs. Although you deferred to saying "excellent dispersion" as opposed to wide dispersion. I know Harman speakers don't use cardioid - you've ignored where and how I pointed out they are different.

"I believe the boundary coupling..." And go on to only mention the rear wall behind the speakers. The reason both Kii and 8C can work in a corner is because side wall influence is ameliorated.

If the speaker at the tweeter is 40cm from side wall, but you don't get side wall early reflections you must be dispersing sound narrow enough not to.

This doesn't diminish the fact the waveguide widens HF dispersion to meet that of the bass and mids, but the result is not a wide dispersion speaker in the manner Toole spoke of.

The 8Cs have wide even dispersion,  Toole states that a speaker should have flat on axis response and an off axis that is smooth even and mirrors the on axis, the 8Cs do exactly that.

They also have a cardioid response through the midrange ,and use boundary coupling to prevent cancellations, boundary coupling is well known and recommended by Genelec amongst others, all totally in line with Toole/Olive.

Why Harman don’t produce and active loudspeaker with cardioid /boundary filtering is another question of course.

Just to add, Martijn discusses and explains the 8Cs on the D&D thread on Gearslutz, if you just search his user name ‘mensink’.

Keith

Edited by PuritéAudio

Share this post


Link to post
Share on other sites

@PuritéAudio

Dutch 8C Horizontal Directivity

(and more importantly, how to read it)

image3.png.e6f75d5826828ce0647b3a40b0ca4

On first appearance, we can say that this looks like a wide-dispersion as the main colour change is along the +/- 90 degree arc in front of the speakers. However, it is not so.

Only the dark red is at 0dB, so this means there is a circa 30 degree arc either side of the tweeter in which the sound has the same SPL as on-axis. This itself is extraordinarily good.

However, we have a drop off by about 45 degrees from axis of -2dB, and -4db by 60 degrees. -6dB (orange) is the point at which half the perceived loudness would be encountered. While this is circa -70-80 degrees in the low frequency, we still see narrowing to the high frequencies. If we draw a line across the map at the high end of -6dB...

image.thumb.png.0522efb7ebe92ab0216d0ec20b7194d7.png

We see that half perceived volume above 1kHz is roughly 45-50 degrees off axis.

So if we then toe in the speakers 30 degrees or enough to cross in front of the listener, this edge where frequency fall off occurs is going to not meet the side wall for quite some time, especially the more space there is between the speaker and the side wall. The mid to high frequencies are used by our brains to localize direction of sounds (which we use to build our perception of 3D soundstage with stereo), so it's really the higher frequencies in this arc that we will perceive, which with 30 degree or more toe-in, we will experience in the entire area in front of the speakers from left side of the room (less the mount the speaker is from the side wall) and equally the right. The result is a sound that is remarkably consistent wherever you happen to sit on your couch.

And this is my point. The aim is for the band down the middle of the above to cover a wide arc to give a very good sweet spot, but then for SPL to fall off quickly after that, so that toe-in combined with control of directivity at each frequency results in very little side wall interaction.

An earlier attempt at better controlling directivity via the Geddes Abbey speaker (which I own) manages keeping the high frequency directivity from 1kHz up between 30 and 50 degrees much like the 8C speaker. However, the bass and frequencies up to circa 800Hz have no controlled directivity and the SPL drops off far less than in a cardioid speaker like the 8C. Geddes utilized this to an advantage by having a sealed box for more accurate bass but sacrificing low frequency depth, and then, saying that the placement should be in the corner of the room  (with 45 degree toe-in) such that there is immediate reflection of the rearward bass energy into the room. As such, you can get close to the 8C sound of plenty of spaciousness and a wide sweet spot, without the frequencies the brain uses for localization of sound (imaging/soundstaging) being marred by wall reflections at or above half-volume. The black line on the graph below is the half-perceived volume (-6db) point.

image.thumb.png.1271b63ca9c3a531dc57a4f276c67281.png

Now if we examine speakers based on wide dispersion, the aim was not to narrow the bass dispersion and widen HF to align with the an ideal directivity somewhere in the middle, but rather to widen the other frequencies through the baffle design and waveguides. This was mainly to avoid a narrow sweet spot that is gone the moment you move your head or sit slightly off centre. Something both the above more recent designs already accomplish.

Here's a wide-dispersion design...

image.png.0457d4685ed8657188c46e7b3a648481.png

Here the design used an ultra-wide dispersion tweeter achieving +/- 40 degree dispersion at 20kHz!! -6db fall-off at 1kHz  is off-axis by 80 degrees, and still 60 degrees by 10kHz!  While the directivity smoothness is fairly controlled, it is wide to 10kHz. This will mean that even with a speaker toe-in of 30 degrees, we will have sound leaving at 80 degrees (or 50 degrees relative to the front-to-back axis of the listening room). In other words, we will have significant frequencies reflecting off the side walls in all eventualities that are used by the brain to localize sounds. This will create a diffuse (some might say 'spacious', including Toole) soundstage that is less accurate as the listener is now hearing a lot of reverb in everything that is played from early reflections.

Floyd Toole was a proponent of early reflections in the work he did. It's speakers like the one above that conform to Toole's esposued research.

I've heard speakers in both camps. What sounds right to me? In short, lose the side wall reflections that blur the soundstage and make 3D imaging unrealistically large and full of reverb. The Kii's and 8Cs do it right by using cardioid dispersion to lessen the bass dispersion SPL off-axis, and a waveguide to widen the HF dispersion. The ideal dispersion pattern is quite simply this: for the amount of toe-in you wish to apply, you want little to no early side-reflections (inside the first 10-12ms, so about similar amount of feet in front of the speaker on the side wall) and within the arc in front of the listener you want an even SPL at all audible frequencies.

So what about studio reference sound? How do they setup the room? Commonly, studios have oodles of absorption at the front of the room where they place the speakers. The mixing engineers can ill afford to have reflections blurring the time locality and stereo position of an image. Hence it is common to use LEDE absorption setup in a studio. LEDE stands for Live-End, Dead-End. It means one end of the room is almost free of damping materials, while one end is so damped as to not get reflections from frequencies that affect the brains perception of a 3D stereo image.

In the studio environment, wide-dispersion can be great since listening is near field, and walls are damped to avoid reflections, so a wide sweet spot is obtained and the mixing engineer can move around the length of the mixing desk without issue. However, if the studio is not LEDE in design, such wide-dispersion will cause phantom images. And in the home environment where there is likely little to no damping, the result is likely to be terrible. Don't get me wrong, it won't necessarily sound bad since if the wide-dispersion still has smooth changes in directivity rather than ragged ones, the perception will be of a big diffuse ans spcaious sound. However, pinpointing sound sources, or listening for micro-details will be marred by different sounds overlapping each other due to the different time it takes for a direct sound to arrive at the listeners ear versus a reflected one.

2 minutes ago, PuritéAudio said:

The 8Cs have wide even dispersion,  Toole states that a speaker should have flat on axis response and an off axis that is smooth even and mirrors the on axis, the 8Cs do exactly that.

No, the graphs above make the differences clear. You are conflating two different aspects.

A cardioid speaker nullifies sound to the rear and sides by wave cancellation or reflection through a porous membrane that alters phase and will cancel the original wave partially or fully at the point the cancelling wave emanates. That's what's happening in the 8C midrange through the side slots.

A wide dispersion speaker is like the last directivity plot above where even response is aimed for over the entire audio range.

The cardioid give a narrower response every time and does not meet Toole's characteristic wide-dispersion. Toole was an advocate of using early reflections to create a subjectively spacious sound. The merits of which are hotly debated and hinge on views about early reflections.

https://www.gearslutz.com/board/studio-building-acoustics/473846-early-reflections-not.html

  • Upvote 1

Share this post


Link to post
Share on other sites
18 hours ago, PuritéAudio said:

Well no, if we take the example of the 8Cs, we have wide even dispersion, due to the tweeters waveguide and a cardioid response, they are also designed to be placed close to a wall, so effectively there is no rear bass reflection, totally phase coherent, and room modes can be ameliorated with the built in EQ.

Keith

How did this thread suddenly develop into a discussion about a pair of expensive speakers that you may just happen to sell?

  • Haha 1
  • Upvote 1

Share this post


Link to post
Share on other sites
20 hours ago, insider9 said:

I went to see a band I've not known on Thursday.  Only put on a few tracks prior to the gig and though I'd like them. Band is called Birds of Chicago. Must say they were outstanding! Trio playing acoustic guitars, banjo, etc. and singing. Small venue packed max 100 people. They sounded excellent! As part of encore they came down and sang unamplified so you had a point of reference to how vocals sound. I spoke to them after the gig as well. 

Been listening to all their recordings since and must say that the vocals sounded so much better live. It's not just the atmosphere and I've only had two pints (0.5%). But some of their recordings sound too clean compared to live performance. Don't get me wrong SQ is good but there was marked difference. 

Considering room acoustics (both) concert venue and studio, different mics used, guitar strings or even picks... Is there a point sweating over playback accuracy? When there's so much variation during recording process.......

Should we give up on reproducing the recording with the highest possible fidelity just because reproduced sound, even of a recording of a live event, doesn't sound like the real thing? That doesn't make sense to me.

In my experience adding distortion, whatever the kind, will detract from the sonic realism even though in some cases it may make the sound more 'pleasant' or 'enhance' the listening experience for some people; but ultimately it's your call

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.