Login / Register
 
Flatten Monitors Response - Recommended Software?! Filter Impulse Response?
New Reply
Subscribe
#31
29th March 2012
Old 29th March 2012
  #31
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Link

#32
29th March 2012
Old 29th March 2012
  #32
Gear maniac
 
Joined: May 2009
Posts: 160

JohnPM is offline
Quote:
Originally Posted by Glenn Kuras View Post
I am sure it would shorten (by a little) but it still is going to be there. That really is my point.
It could be eliminated completely for that single listening position. It is important to recognise that EQ filters do not modify "the room", they modify the Transfer Function from the point where the speaker is to the point where the mic is. Between those points the room acts like a filter, is it really such a leap to realise that one filter can alter the effect of another? The main difficulty for EQ is that the room's "filter" is different for different pairs of speaker and listener positions, so any EQ that hopes to address more than a single position is of necessity a compromise based on the various room filters it is addressing for the listening positions at which it must work. There are many room filters, one for every listening position, but we can only apply one EQ solution.
__________________
REW Author
#33
29th March 2012
Old 29th March 2012
  #33
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Source

Hi John, thanks for joining us. If you would indulge me....
Does Eq shorten modes?
Very roughly, what sort of reduction could one expect?
Does this shortening apply to listener position only or does the reduction in mode level and (relative) length pertain throughout the space?

DD
#34
29th March 2012
Old 29th March 2012
  #34
Gear maniac
 
Joined: May 2009
Posts: 160

JohnPM is offline
Quote:
Originally Posted by DanDan View Post
Does Eq shorten modes?
It can, but only if the filter settings (more technically, the zero positions) match the resonance it is countering.

Quote:
Very roughly, what sort of reduction could one expect?
The decay of the resonance is replaced by the decay of the filter, which depends on how much cut the filter has. The greater the cut, the faster the decay, but the filter decay will always be faster than the original mode. Some images may make that more clear.

Here is a measurement of a room that has a strong resonance at 50Hz.
Name:  original.png
Views: 907
Size:  56.7 KB

Here it is with a filter that is matched to the resonance, but has only 1dB of cut.
Name:  minus1db.png
Views: 913
Size:  56.3 KB

Now with 3dB cut.
Name:  minus3db.png
Views: 871
Size:  54.4 KB

6dB.
Name:  minus6db.png
Views: 879
Size:  52.1 KB

And finally 15dB.
Name:  minus15db.png
Views: 883
Size:  45.5 KB

Hopefully the way the rate of decay increases as the filter cut is increased is easy to see.

Quote:
Does this shortening apply to listener position only or does the reduction in mode level and (relative) length pertain throughout the space?
For a given mode, the rate of decay of that mode is the same everywhere in the space but its amplitude varies with position. Unless the mode is fairly isolated from its neighbouring resonances it can be difficult to see that in measurements, however. Applying a filter will change the decay of the specific mode it is matched to everywhere, but the level at the modal frequency at any particular location will depend on what it was before the filter was applied - all locations will be shifted by the same amount, so if the peak at the listening position was 10dB and a -10dB filter matched to the mode was applied, there would be no peak at the listening position, but at a location where the mode did not add anything the level would also be 10dB lower than it was.
bwo
#35
30th March 2012
Old 30th March 2012
  #35
bwo
Lives for gear
 
Joined: Aug 2009
Location: Norway
Posts: 568

bwo is offline
Quote:
Originally Posted by DanDan View Post
G
Thanks bwo. Does DRC do something more than Eq, some time domain damping thing?

DD
I believe Audiolense uses both IR and FIR filters. You might find more here:
Juice HiFi

I was asked in a pm if I had the raw measurements files, but I don't. They are not mine. Mitch has posted them in his review:
Hear music the way it was intended to be reproduced - part 1 | Computer Audiophile

Hear music the way it was intended to be reproduced - part 2 | Computer Audiophile

Hear music the way it was intended to be reproduced - part 3 | Computer Audiophile

Hear music the way it was intended to be reproduced - part 4 | Computer Audiophile

Hear music the way it was intended to be reproduced – part 5 | Computer Audiophile

Hear music the way it was intended to be reproduced - conclusion | Computer Audiophile
#36
30th March 2012
Old 30th March 2012
  #36
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

What's missing from all of these graphs is what happens six inches away from where the microphone was when the EQ was calibrated.

--Ethan
__________________
Ethan's Audio Expert book
Quote
1
#37
30th March 2012
Old 30th March 2012
  #37
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
While the statement is true in itself, on Nyal's page he says he measured at every seat. Toole's quotation seems logical, and JohnPM's answer says the same thing. i.e. Modal decay rate doesn't change with position.

Do you have some test or theory to suggest otherwise?

And as we know some Eq fixes, e.g. Soffit Filter, cover the whole room.

I must say I am quite amazed at the slew of remarkable successes being shown. Even if they were only at the mix position. Which I do not believe. Until I see tests or theory otherwise I am going to accept Nyal, Floyd, and John's considered opinions. I do believe things have moved on and there is now considerable benefit to be had using this extra tool.



DD
#38
30th March 2012
Old 30th March 2012
  #38
Gear maniac
 
Joined: May 2009
Posts: 160

JohnPM is offline
Quote:
Originally Posted by DanDan View Post
JohnPM's answer says the same thing. i.e. Modal decay rate doesn't change with position.
The decay rate doesn't change, but the level can vary dramatically.
Quote
1
#39
30th March 2012
Old 30th March 2012
  #39
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Location Location

Unless you are referring to something else John, the variation of mode levels within tiny distances has been well observed. The first example I saw was a classic Ethan test.
Indeed this is why I encourage two measuring positions, binaural mics etc.
We typically measure what the nose might hear!

However, let's say we are sitting mid width. Second axial mode has a level peak around the room centre. Let's say this modal peak qualifies for an Eq cut, tuned exactly at the nose for present purposes.
By your' and other descriptions the level and rate of decay of that mode will be diminished, even fully corrected.

The big question for me (same as Ethan I reckon) would be how big is the area 'corrected',
for both Level and Decay rate?
If it collapses over a few inches, a head width, it would not be particularly useful IMO. However I would be delighted to hear 2-3 feet.
Furthermore is it possible that the Eq and Decay rate correction could make things worse in terms of the differences observed at the sweet spot and relatively nearby?
Is does this difference remain relatively the same?



DD
bwo
#40
31st March 2012
Old 31st March 2012
  #40
bwo
Lives for gear
 
Joined: Aug 2009
Location: Norway
Posts: 568

bwo is offline
#41
31st March 2012
Old 31st March 2012
  #41
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Mac

Thanks bwo. Glad to see someone is thinking of Macintosh, i.e. Dirac.
Is that the same company that makes the Dirac sold by Bruel and Kjaer for KiloEuros?
They have some weird ideas as to what a Demo is. I just tried the Trial which failed. Digi Avid Hardware doesn't like to be used as a SoundCard!
Working on finding a way to try REW through an Audio Unit Eq. Wish me luck.
DD
#42
31st March 2012
Old 31st March 2012
  #42
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

Quote:
Originally Posted by DanDan View Post
The big question for me (same as Ethan I reckon) would be how big is the area 'corrected', for both Level and Decay rate?
A friend of mine is a DSP expert, and here's how he explained it to me: He said that the more correction that's applied, the smaller the corrected area becomes. And we know for sure from my previous tests that improvements in one location can make things even worse elsewhere.

--Ethan
Quote
1
#43
1st April 2012
Old 1st April 2012
  #43
Lives for gear
 
jhbrandt's Avatar
 
Joined: Feb 2009
Location: Jakarta, Indonesia
Posts: 3,989

Send a message via Yahoo to jhbrandt Send a message via Skype™ to jhbrandt
jhbrandt is online now
Quote:
Originally Posted by Ethan Winer View Post
A friend of mine is a DSP expert, and here's how he explained it to me: He said that the more correction that's applied, the smaller the corrected area becomes. And we know for sure from my previous tests that improvements in one location can make things even worse elsewhere.

--Ethan
++1

"Jim, ye canno' change the laws 'o physics!" - Scotty

Cheers,
John
Quote
1
#44
1st April 2012
Old 1st April 2012
  #44
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Test

Well that's well believable Ethan, no free lunch.
However, the Audessey software shows no significant reduction of the modes to my eye. All of the other apps here show very decent reduction of the modes.
That night and day difference, plus the newer ones all state that they involve time, suggest to me that no comparison is possible. Nor are any conclusions drawn from the Audessey test valid for the others.

There is extensive reading in the linkage on the laws of physics and math.
Much of it says the same as JohnPM. i.e. some anomalies are amenable to successful Eq. If Mathematically equipped one might debate the reasoning John and Toole use, but ultimately I don't think the answer I/we want lies there. i.e. How wide is the area usefully corrected?
Time for a new test IMO, as time passes an old wisdom may become a new myth. And let's test in a non comparative way, i.e. not Eq vs Treatment.
And let's test one that is highly regarded by those that use these things.

Aside- I wonder is there any practicing Acoustician who does not use Eq?
Present company included!
We all agree that Speaker corrective Eq is OK, but I believe the Room/Speaker are very interactive. Looking at the Waterfalls of speakers is also enlightening.

Hopefully I will crack the issue of feeding audio from one app to another in the Mac OSX shortly. If so I will do the test.


DD
#45
1st April 2012
Old 1st April 2012
  #45
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

Quote:
Originally Posted by DanDan View Post
All of the other apps here show very decent reduction of the modes.
EQ is useful to reduce modal peaks. That's not the point. EQ doesn't reduce the rate of decay, which is at least as damaging as the peaks themselves because bass notes still run into one another. Further, resonances can be excited by nearby bass notes (sympathetic vibration), which tends to make many different notes all the sound the same. EQ can't fix that either because it merely lowers the volume at those frequencies, and doesn't solve the underlying problem. Nor can EQ improve nulls, which are the larger problem in most home-sized rooms. For at least the 40th time, and as explained clearly in my Audyssey Report, I use one band of EQ cut to reduce a 40 Hz mode in my living by a few dB. So it's not that EQ is never useful. But it's not the solution many claim. It's a band-aid.

Why are we still arguing about this?

--Ethan
#46
1st April 2012
Old 1st April 2012
  #46
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Differing

Debating I hope. But there are very substantive differences of opinion and even fact.

I regard Audessey as useless. It shows negligible effect on modes, all the others are spectacularly successful in contrast. I regard any information derived from tests on it as per se unreliable.

Quote:
EQ doesn't reduce the rate of decay
I don't concur. I am convinced by the theories, using the laws of Maths and Physics, rehearsed by John and many, actually pretty much all the others, that Eq can and does alter the rate of Modal Decay. When the readings show it possible, i.e. the Excess Group Delay thing.
Looking at the slopes of the mode in John's graphs shows this.
It gets steeper, an increased rate of decay. It's not just level.

I have extensively treated a nearly impossible room here at SoundSound.
It is not viable in several respects to treat it more or differently.
I can work it reliably but to be frank the measurements and actual response are pretty embarrassing.
Many share this position. Test will ultimately show if this situation can be improved or not. I will even stoop to a Behringer if I have to .



DD
bwo
#47
1st April 2012
Old 1st April 2012
  #47
bwo
Lives for gear
 
Joined: Aug 2009
Location: Norway
Posts: 568

bwo is offline
#48
1st April 2012
Old 1st April 2012
  #48
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

Quote:
Originally Posted by DanDan View Post
I regard Audessey as useless. It shows negligible effect on modes, all the others are spectacularly successful in contrast.
I'll believe it when I measure it for myself.

Quote:
Originally Posted by bwo View Post
I also don't see the point that it will not work 6 inches away. Most of us listen at one spot.
I said six inches because that's the space between your ears. So if you optimize the EQ for one ear, the response at the other ear will not be improved and may even be made worse. This graph shows the response measured at two locations only four inches apart in a typical small room:



The graph above and the explanation below are from my article A common-sense explanation of audiophile beliefs:

Quote:
Originally Posted by Ethan's article
Conventional wisdom holds that the bass response in a room cannot change much over small distances because the wavelengths are very long. (A 40 Hz sound wave is more than 28 feet long.) Yet you can see in Figure 1 above that the peak at 42 Hz varies by 3 dB for these two nearby locations, and there's still a 1 dB difference even as low as 27 Hz. The reason the frequency response changes so much even at low frequencies is because many reflections, each having different time and phase delays, combine in different amounts at each point in the room. In small rooms the reflections are strong because the reflecting boundaries are all nearby, so that further increases the contribution from each reflection. Also, nulls tend to occupy a relatively narrow physical space, which is why the nulls on either side of the 92 Hz marker have very different depths. Indeed, the null at 71 Hz in one location becomes a peak at the other.
So how could EQ fix the peak at 71 Hz without making the null even worse four inches away? Note that the graph above is unrelated to EQ, it's just how small room acoustics works.

--Ethan
Quote
1
#49
1st April 2012
Old 1st April 2012
  #49
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

Quote:
Originally Posted by bwo View Post
Can you explain why?
The nulls have improved here:
Yes, you can force a null to be improved by applying EQ, but again that works for only one location. Nulls are typically 20 dB deep or even deeper, which means you'll need a ten-fold increase in power.

--Ethan
#50
1st April 2012
Old 1st April 2012
  #50
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Subjective

BIR Nulls are the result of cancellation due to reflections arriving out of phase.
There are sibling peaks but the peaks max at +6dB while the nulls as Ethan said can be very deep. A perfect reflection should null infintely, but 20-30dB is common.
Clearly diminishing such reflections does far more good to the nulls than the peaks.
I suggest that diminishing modal peaks, really diminishing them, by changing the rate of decay, should have the inverse but similarly more useful effect of shallowing/widening the nulls.
I reckon that is what we are looking at in bwo's example graph.

Ethan asked Nyal if his 'proof' of Eq shortening modes was tested in more than one location. He replied that is was, at every seat in the home theatre in question and said he would show the results. Did this happen?
It would save me bother of testing!

DD
#51
2nd April 2012
Old 2nd April 2012
  #51
Gear Guru
 
Glenn Kuras's Avatar
 
Joined: Jul 2005
Location: Atlanta, GA
Posts: 16,617

Glenn Kuras is offline
Quote:
I said six inches because that's the space between your ears. So if you optimize the EQ for one ear, the response at the other ear will not be improved and may even be made worse. This graph shows the response measured at two locations only four inches apart in a typical small room:
Also to add most people can not put there head exactly in the same spot every time, so to be off 4" (front to back and or side to side) is common.
__________________
Glenn Kuras
GIK Acoustics USA
GIK Acoustics Europe
http://www.gikacoustics.de (German Translation)
404 492 8364 (USA)
+44 (0) 20 7558 8976 (Europe)

Built in Slat design (Scattering/Diffusion) on all Bass Traps click here
#52
2nd April 2012
Old 2nd April 2012
  #52
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Off

The variation in frequency response in relatively short distances has been illustrated very well by an Ethan test.
This is why I suggest placing the mic at each ear position and taking two measurements with single speaker drive. When the two responses are averaged we should have something close to what we actually hear.
One could go even further to find out what is really happening at LF. Drive two speakers, place the mic at each ear position with a blocking disc to simulate the effect of the head.
But I digress...

Dirac are claiming to be doing something quite different.
Quote:
Technically, Dirac Live® improves on other digital room correction systems - as customer benchmarks show - by not just making minimum-phase or linear-phase adjustments to the frequency response. Dirac Live® is unique in offering true mixed-phase impulse response correction. This is achieved not just in a small listening region, but in a volume of arbitrary size. Several measurement positions and separate time and frequency-domain criteria are used to make sure all of the common pitfalls (over-compensation, under-compensation, pre-ringings, decorrelation, etc.) in advanced room correction are avoided.
Unfortunately their demo is a buy and try, 100% refund if not satisfied.

Elusive.

DD
#53
2nd April 2012
Old 2nd April 2012
  #53
Gear interested
 
Joined: Apr 2012
Posts: 2

Mitchco is offline
Quote:
Originally Posted by Ethan Winer View Post
What's missing from all of these graphs is what happens six inches away from where the microphone was when the EQ was calibrated.

--Ethan
Hi, in Audiolense, you can make multi-seat room correction across the listening area. For example:



Cheers,

Mitch
#54
3rd April 2012
Old 3rd April 2012
  #54
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

I'd rather see actual measurements rather than a graph showing simulated responses. And also without any smoothing applied.

--Ethan
#55
3rd April 2012
Old 3rd April 2012
  #55
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Your wish is my command Ethan. Dirac have put me on their Witness Protection Program...;-)

Beta testing. I will give it a good spin shortly.

DD
#56
3rd April 2012
Old 3rd April 2012
  #56
Gear interested
 
Joined: Apr 2012
Posts: 2

Mitchco is offline
Quote:
Originally Posted by Ethan Winer View Post
I'd rather see actual measurements rather than a graph showing simulated responses. And also without any smoothing applied.

--Ethan
The trial can be downloaded here: Juice HiFi if you want to give it a go yourself.

Btw, thanks for the acoustic links to your site. I used them in this article:

Speaker to Room Calibration Walkthrough | Computer Audiophile

Cheers!

Mitch
#57
4th April 2012
Old 4th April 2012
  #57
Gear Guru
 
Ethan Winer's Avatar
 
Joined: Oct 2002
Location: New Milford, CT, USA
Posts: 13,663

Ethan Winer is offline
Lightbulb

#58
20th April 2012
Old 20th April 2012
  #58
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
First Impression of DIRAC

I am Beta testing DIRAC on Mac. The Digidesign Hardware HD192 won't play nice so I am confined to the Mac Line Out.
This does sound inferior to the HD192 to my ear.

So, I did a quick first test. The Beta is limited to analysing just the centre of the sweet spot. Called C below.
The full version does much more than that of course, with user chosen locations and so on.
So the filters here were generated using the C position only. The Nose. A worst case scenario I guess.
I took a shot with no Eq, called RAW below.
The other shots were with DRC Left Right and Back of the Nose spot, all by about 0.5M.
Sonically I like the difference. More Bass, Less Honk, More Clarity, Better Stereo and Depth.
Walking about the room, the sound stage holds up quite well and is much wider than the speakers positions would suggest.
The pictures can do the rest of the talking.
Name:  Sweet Circle.jpg
Views: 746
Size:  60.4 KB
Name:  DRC.jpg
Views: 824
Size:  57.3 KB

Enjoy, DD
#59
20th April 2012
Old 20th April 2012
  #59
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 10,958

DanDan is offline
Can't Wait

Off to see some live music, but thought I had better show these before I go.
Name:  RAW.jpg
Views: 724
Size:  117.0 KB
Name:  DRC.jpg
Views: 788
Size:  107.8 KB
DD
#60
26th March 2013
Old 26th March 2013
  #60
Gear nut
 
Joined: Dec 2006
Location: Frankfurt
Posts: 79

ak23 is offline
Hello DanDan,

do you still testing/using DIRAC ?
I am running the free demo right now.
Quite impressive. Try to use it only up to 400-500 Hz.
I flattend my freq. response a lot. OF course this limites the volume
but much better then a totally uneven freq. response.
But i have the feeling it does something above the frequency window.
what are your opinios after a longer period?

greetings

Ak
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
BLueROom / So much gear, so little time!
3
Kyle S / Electronic Music Instruments & Electronic Music Production
4
A27Hull / Geekslutz forum
9
esaias / Remote Possibilities in Acoustic Music & Location Recording
0
Harley-OIART / Music Computers
3

Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.