Full range EQ à la Française

[fouiz]

Hi there,
here i intend to post my experiences about full range eq corrections on my system using REW.
I' discovered this great tool a few weeks ago, after trying to get the best from a DEQ24ç6 i recently bought. After some testing and auto eq, i realized that i needed to get more informations about the measures before applying any corrections, informations that the DEQ could not give.
After some surfing on the appropriate forums, i discovered REW, and here i am !
I'm gonna try to explain all the process i'll follow, just to share the adventure.
One more thing : i'm french, and my english is a little bit rusty since school, so my apologies for potentiel inappropriate wording ...!

OK, now for the context :

Dedicated room :
5 meters long, 3,75 meters wide. Ceiling is at 1,9 meter high.
Walls in bricks.
Front 2/3 covered with tense fabric on flannelette. 2 bass traps in the corners (filled with glasswool ? fiberwool ?)
Back 1/3 covered with wood
Tense Fabric on the ceiling
Carpet on the ground.
One couch on the left side
Front wall has a big screen for projection too.
>> I'll try to make a plan and be more precise about this later.

Audio equipment involved :
ECM8000 > M-audio mobilepre > laptop PC > DEQ2496 >optical> Sherbourn PT7000 (preamp) > Outlaw ICBM (Bass manager) > Parasound Halo A23 (amp) > BC Acoustique NIGER + REQ Q200E

Settings :
Using newecm.cal, mic is pointing to the ceiling at my listening position.
M-audio calibrated on channel 1 (left) using MIC input (balanced, phantom off) and line output (unbalanced) - see response curve below - not bad in my opinion.
Checking levels at 75db (with the ECM8000 as the input and a Radioshack analog SPL meter at hand just near the mike) gives a result of -11db for sweep for -18db signal on input

To be continued ...

[Wayne A. Pflughaupt]

That's some impressive "rusty" English there, Cedric. Looking forward to the next post...

Regards,
Wayne

[fouiz]

OK, first graphs from yesterday night.
First one is Full range stereo + subwoofer phase 0.
The ICBM applies the filtering at 60Hz between front and sub

Seconde one is stereo without sub.

i did some other measurements, but look really wrong, i think i have a levels problem, got to check it before posting other graphs.

[brucek]

Post your graphs using the standard used here:

Vertical axis = 45dB-105dB.

Use a 75dB target.

brucek

[fouiz]

OK, graphs updated.
I set a 75db target but it does not appear un 'measured' window. Does it matter ?

By the way, here are the RT60 results for the Stereo+sub and stereo only measured. Looks quite good to me but is it realistic ?


RT60 STEREO + SUB



RT60 STEREO ONLY


Next one is the Waterfall for Stereo+sub. I'm not sure about the interpretation of this ...
WATERFALL STEREO+SUB

[brucek]

The waterfall is most useful (and use to be limited to) the subwoofer range.

Use a time axis of about 300ms and set the vertical (45dB-105dB) and horizontal (15Hz-200Hz).

Now you'll be looking at successive slices in time. Move the slider to one and you'll have a match to your frequency response graph. Start moving the slider out to 30 slices to watch how the level drops in the room with each slice.

Above the subwoofer range, the RT60 is useful.

brucek

[fouiz]

Hello,
took me some time to get back to measuring and fix my levels problem.
So here are my observations :
First comparing L+R without sub (blue) , then L+R+Sub phase 180 as it was set (red) and L+R+S phase inverted (green).

We can see that Mains and sub meet together around 60Hz.
I noticed a big dip with phase 0, just before 90Hz.


Same one zoomed below 200Hz


Waterfall for the sub


My main speakers are supposed to go down 40Hz (that we can see on the first graph) but there was a matching problem with the sub.
After a little searching, i found out that changing the crossover filter on the bass manager from 60 to 120Hz would do the trick !

Sub phase 0 initial (dark green) and with modified crossover (light green)


Waterfall for the two measures


So, no eq yet, just trying to optimise with settings changes.
Next i'll see what i can do to work on higher frequencies by changing speakers position.

Regarding EQ, it seems obvious that i need a first filter around 40Hz for the sub.

What do you think ?

[brucek]

i found out that changing the crossover filter on the bass manager from 60 to 120Hz would do the trick

Anything above 80Hz would cause the sub to be localizable.

it seems obvious that i need a first filter around 40Hz for the sub

Yeah...

brucek

[fouiz]

brucek wrote:

Anything above 80Hz would cause the sub to be localizable.

brucek

Well, in fact no.
Maybe because his position is right behind the left main speaker, firing perpendiculary (from left to right along the wall). And it doesn't feel like hearing more on the left or the right side ...

[brucek]

A mono source (sub) is far more localizable than stereo (mains), irregardless of its position ......................

[fouiz]

brucek wrote:

A mono source (sub) is far more localizable than stereo (mains), irregardless of its position ......................

Of course, But if the signal above 80Hz coming out of the sub is properly merged with the signal coming out of the mains, then localisation is not that obvious ...

[brucek]

My understanding is that the presence of a soundfield is the over-riding influence with regard to localization (at frequencies over ~80Hz). There is no soundfield present in a mono signal, so no matter where it is located or merged or whatever, it can be easily identified by the human ear as to its point of origin. But I've been wrong before....

brucek

[fouiz]

Well, i have to do more testing on this point (not possible for the moment, as i have someone sleeping in my dedicated room - which can be a friends bedroom too), i'll check in may when i'll be back from vacation.

Anyway, about the other measures, any comments ?
For example, do you think the RT60 is too low ?

[spreston]

Cedric-

Have you continued your testing?

: P

[fouiz]

Hello,
well, absolutly no time for the moment.
I hope next week i'll be able to start again ...