Is any of the EQ Simulators in REW comparable to Inguz?

[hybris]

Hi. I'm using Inguz EQ as a plugin to my squeezebox to EQ my system, mainly the bass area (20-500hz).

So far I've been doing this manually by measuring individual sinustones with radioshack and manually applying EQ to inguz until I got a decent house curve.

I've now discovered REW which so far seems to be a great program. Easy to use and excellent help! It takes the pain out of measuring the system. However, it is no practical way of plugging inguz into the loop so that I can directly measure the result of inguz.

So I was wondering if I could adjust the EQ in REW until I got the frequency response I want, and then apply these EQ settings manually to Inguz. For this to work properly I assume I would need to use an EQ simulator with a EQ type somewhat comparable to Inguz.

As I understand Inguz works a bit different from many Equalizers, and you cannot set the Q of each band directly. As stated by the author : The InguzEQ system doesn't have any separate control of bandwidth (or "Q"), unlike fully parametric EQs. Instead it just uses cosine-shelves between the points specified; so bandwidth is implied from the points and their spacing. In this sense it's much more like a N-band "graphic equalizer"

Can I simulate this with the EQ in REW in any way? If not exactly, at least in a way that should give me roughly the same effect when applying say 20 different EQ points between 20-500hz?

Thank you for any help!

[hybris]

After some fiddling I adjusted my subwoofer placement, volume and crossover until I got the following result, and then manually configured inguz to reduce more or less the same peaks as rew suggested. Added two peaks that REW ignored, and also added some 0db marks to limit the bandwidth of the EQ settings of inguz. My final EQ is as follows, and it sounds good. good low end extension and overall the bass response sounds balanced I have to do some more listening to see if additional adjustment is needed.

Comments and suggestions welcome!

The response is above the 75db reference down to about 24.5hz The subwoofer crossover is now 80hz. If anyone wonders the subwoofer volume is intentionally turned up to extend the low end is much as possible, then EQing down the overall volume.

Subwoofer is an XTZ 10" Class D active subwoofer, the speakers is B&W 804s. The system is mainly used for music, not movies.

INGUZ CONFIGURATION:

<EQ Bands="16">
<Band Freq="30">0</Band>
<Band Freq="36">-7.5</Band>
<Band Freq="58">-9.5</Band>
<Band Freq="78">-10.2</Band>
<Band Freq="90">3</Band>
<Band Freq="98">-4.4</Band>
<Band Freq="107">3</Band>
<Band Freq="113">-9.4</Band>
<Band Freq="122">-5.3</Band>
<Band Freq="132">1</Band>
<Band Freq="140">-9.1</Band>
<Band Freq="149">0</Band>
<Band Freq="156">-3.0</Band>
<Band Freq="165">0</Band>
<Band Freq="5000">0</Band>
<Band Freq="15360">-3</Band>
</EQ>
Based on this measurement of a fullrange (woofer+speaker) signal, attempting to EQ a hard knee house curve:

[hybris]

I guess the waterfall indicates that I have some serious issues around 60hz.. I guess this might explain why my previous configuration had the subwoofer crossover at 40hz when I had measured manually. It's really bass heavy on some tracks with the current EQ filters and crossover, even though the frequency response measurement doesn't indicate that that should be the case. But I guess the issue at 60hz might be the problem.. ?

Probably not so easy to EQ my way out of that one?

(note that the graph go all the way out to 1000ms...)

[brucek]

I guess the waterfall indicates that I have some serious issues around 60hz

When you see a waterfall with a correct decay of signal over time (as you have at 60Hz) in the first 300msec, and then a narrow signal jutting out of nowhere, it is usually an indication of noise, and not a room problem. You would likely see this noise if you examined the spectrum without any stimulus from a sweep. 60Hz is especially curious since it is associated with hum.

Take a non-sweep spectrum reading of your system and see what you get.

Read this post to give you idea of how to do that.

brucek

[hybris]

brucek wrote:

When you see a waterfall with a correct decay of signal over time (as you have at 60Hz) in the first 300msec, and then a narrow signal jutting out of nowhere, it is usually an indication of noise, and not a room problem. You would likely see this noise if you examined the spectrum without any stimulus from a sweep. 60Hz is especially curious since it is associated with hum.

Take a non-sweep spectrum reading of your system and see what you get.

Read this post to give you idea of how to do that.

brucek

EDIT: forgot to turn on the radioshack..(purple graph) hehe.

Hi, thanks for the tip. I attempted to do as described in the post. Not 100% sure if I did that correctly, but if so - here is the result. Does that tell us anything? (black graph)

[hybris]

here is spectrum with no signal running (black graph) (was that was I was supposed to do?)

[brucek]

here is spectrum with no signal running (black graph) (was that was I was supposed to do?)

Yep, but I see no noise out of the ordinary, so I can't really account for the ringing you had at 60Hz in your waterfall, other than it may have been something turned on at the time that caused it (refrigerator, etc).

Try another waterfall and it may be gone. be sure to do your testing at 75dBSPL at the listening position, so you can then see the level of noise ringing (if there is any). Anything below about 55dB you can ignore.

Your response seems fine though. It doesn't extend too low (around 35Hz), but it seems fine.

brucek

[hybris]

brucek wrote:

Yep, but I see no noise out of the ordinary, so I can't really account for the ringing you had at 60Hz in your waterfall, other than it may have been something turned on at the time that caused it (refrigerator, etc).

Try another waterfall and it may be gone. be sure to do your testing at 75dBSPL at the listening position, so you can then see the level of noise ringing (if there is any). Anything below about 55dB you can ignore.

Your response seems fine though. It doesn't extend too low (around 35Hz), but it seems fine.

brucek

Well, there seem to be something going on around 60-70hz each time I measure. Below is a waterfall of the speakers alone (the sub turned off). The issue seem to have shifted up to about 70hz here. I guess that would indicate that this is a room / phase effect of some sort?

[hybris]

Hm, perhaps you're right after all. I made some new measurements now of all three listening positions in my sofa, and now the ringing effect suddenly appears to be gone, and then it appeared again a little while later.

Here is the center position waterfall while it was gone:




I also plotted graphs for all three listening positions. The one with the heaviest dips is unfortunately the center position... Could that be due to cancellation between my two main speakers or something? .

[brucek]

now the ringing effect suddenly appears to be gone, and then it appeared again a little while later.

Yeah, could be something turning on and off.

The one with the heaviest dips is unfortunately the center position... Could that be due to cancellation between my two main speakers or something?

Those are fairly sharp dips. Narrow sharp dips are hard to detect by the human ear.

Hopefully, you're not using a soundfield in your receiver to get the center channel measured. The receiver must be set to stereo only. Any soundfield will operate its steering with phase and so the results are to be ignored since REW is a mono signal.
To get the Center to play, you must disconnect a main speaker and plug in the center in its place. You can have the other main playing if you want to check for interaction.

brucek

[hybris]

brucek wrote:

Yeah, could be something turning on and off.


Those are fairly sharp dips. Narrow sharp dips are hard to detect by the human ear.

Hopefully, you're not using a soundfield in your receiver to get the center channel measured. The receiver must be set to stereo only. Any soundfield will operate its steering with phase and so the results are to be ignored since REW is a mono signal.
To get the Center to play, you must disconnect a main speaker and plug in the center in its place. You can have the other main playing if you want to check for interaction.

Hi, no I'm playing stereo only. The three graphs represent measurements from the three different seating positions(left, center, right) in my sofa measuring my main speakers only.

Nice to know that the sharp dip at 64hz probably should be ignored. That brought me back to a simpler version of the same EQ settings I had before I started with REW, only with fewer bands as I've ignored some of the sharpest dips. Thank you for all your replies

I'm now back at 40hz crossover frequency and I've set the following EQ based on the graph from REW

<EQ Bands="9">
<Band Freq="30">0</Band>
<Band Freq="36">-4</Band>
<Band Freq="50">-10.4</Band>
<Band Freq="60">0</Band>
<Band Freq="115">0</Band>
<Band Freq="119">-5.4</Band>
<Band Freq="132">0</Band>
<Band Freq="5000">0</Band>
<Band Freq="15360">-3</Band>
</EQ>

REW measurements (dotted line is an approximate simulation of the correction). This sounds way better than 80hz crossover, which was to bassy and muddy. What I really wanted was a response similar to my speakers with extended low end, which is what I got now

[brucek]

the three different seating positions(left, center, right

Ahh, OK, my bad. I thought you were referring to cancellation from mains to center channel.

Anyway, it's too bad you can't identify how they define bandwidth and Q. Then you could associate it with a particular equalizer that uses the same math in REW.
What you're doing though seems to be working.

brucek

[hybris]

brucek wrote:

Ahh, OK, my bad. I thought you were referring to cancellation from mains to center channel.

Anyway, it's too bad you can't identify how they define bandwidth and Q. Then you could associate it with a particular equalizer that uses the same math in REW.
What you're doing though seems to be working.

brucek

As far as I have gathered, the inguz eq creates a smooth sine slope to the next defined band. So you can define/limit a band by adding 0db bands in between as I've done below. I've done manual measurements earlier that confirms this. So my settings will probably follow the house curve slightly better than the dotted line from the REW simulation.

<EQ Bands="9">
<Band Freq="30">0</Band> - To make sure I'm not reducing below 30hz
<Band Freq="36">-4</Band>
<Band Freq="50">-10.4</Band>
<Band Freq="60">0</Band> - Limits the width of the 50hz band above this line
<Band Freq="115">0</Band> - makes sure the 119hz band is narrow
<Band Freq="119">-5.4</Band>
<Band Freq="132">0</Band> - also makes sure the 119hz band is narrow
<Band Freq="5000">0</Band>
<Band Freq="15360">-3</Band>
</EQ>

I'd have to manually measure to know exactly how this ends up, but it sounds right.

Here is an explanation about how the inguz works: http://inguzaudio.com/blog/2007/09/2...nd-parametric/