noob question about house curve

[Marcus Aseth]

just a beginner quick question about an house curve guys, I was following a tutorial about equalizing using REW and EQ APO and the person explaining said that they've "modeled (the curve in blue in the image below) after Harman’s preferred listening curve":

but watching it (with the eyes of a neophyte) I was thinking "isn't that "valley" in the 20 to 200Hz area too extreme?" - looks like a 6dB difference that's too steep to me
Can you guys give me some opinions about it and if I should use it?

 

[Rednaxela]

First thing I would do is try and double check if the modeling was done correctly.

 

[Rednaxela]

I mean otherwise you may not know if your question is about the modeling or the Harman curve itself.

 

[NTK]

OP may want to read Dr. Toole's post (the Harman curve's "creator") on AVS Forum regarding his "thoughts". To summarize, the Harman curve is not a "target" intended for EQ'ing your speaker response to, at least not for the frequencies above a few hundred Hz. It is the "estimated in-room response" of a well designed front firing speaker in a typical domestic listening room. This estimated in-room response is calculated from a weighted average of the loudspeaker's anechoic response curves at multiple different radiating directions (12% listening window, 44% early reflections, 44% sound power). Reference: ANSI/CTA-2034-A section 13.

https://www.avsforum.com/threads/official-jbl-synthesis-pro-revel-home-theater-thread.2515137/page-168#post-57820394

 

[Marcus Aseth]

then what should it be a good target to EQ towards for awesome music listening experience?
And if you (or someone else here) happen to have it, could you share with me the .txt of said house curve I can import into REW?
Would really appreciate ^_^

 

[NTK]

The best known method is, if you have access to the on-axis or listening window anechoic frequency response, then EQ to give an anechoic flat response at frequencies above ~300-500 Hz. For below ~200 Hz, measure the in-room response at the listening position and EQ to flat. However, refrain from trying to boost deep nulls/troughs as they are likely cause by destructive interferences, and adding power at those frequencies will not solve the problem but will put a lot of stress on the amplifier and speaker (every 3 dB increase will double the power). Also, try not to fight too much the inherent low and high frequency roll-off of your speaker. Bass management, i.e. incorporating subwoofer(s) into your system, is the most effective method to get good bass into your room.

If you are lucky, @pierre's site may have the measurements and the EQ parameters for your speaker. After that, you can adjust to taste.

A collection of loudspeakers measurements

 

[Sancus]

For below ~200 Hz, measure the in-room response at the listening position and EQ to flat.

This is a matter of taste and what recordings you're using though.

 

[NTK]

This is a matter of taste and what recordings you're using though.

True. Those are the starting points. Then adjust to taste.

 

[Marcus Aseth]

The best known method is, if you have access to the on-axis or listening window anechoic frequency response, then EQ to give an anechoic flat response at frequencies above ~300-500 Hz. For below ~200 Hz, measure the in-room response at the listening position and EQ to flat. However, refrain from trying to boost deep nulls/troughs as they are likely cause by destructive interferences, and adding power at those frequencies will not solve the problem but will put a lot of stress on the amplifier and speaker (every 3 dB increase will double the power). Also, try not to fight too much the inherent low and high frequency roll-off of your speaker. Bass management, i.e. incorporating subwoofer(s) into your system, is the most effective method to get good bass into your room.

If you are lucky, @pierre's site may have the measurements and the EQ parameters for your speaker. After that, you can adjust to taste.

A collection of loudspeakers measurements

It does have them!

A collection of loudspeakers measurements

pierreaubert.github.io

Sorry to test your patience, but I'm new to this, can you explain me what am I supposed to do with the data on that link? (in terms I can more easily understand) ^_^

Meanwhile, I'll be EQing the subwoofer to flat, thanks for all the help so far, really appreciated ^_^

 

[NTK]

The EQ section should have the parametric EQ parameters for your speaker.

 

[Marcus Aseth]

I mean, probably I'm failing to understand you, because I've put that file into the EQ APO and the change in measuring was minimal, and the graph is still overall terrible... I'm looking for a house curve to import into REW EQ

Here's a screen

So I don't quite get it, what house curve should I load into REW? Where can I download it?

 

[Rednaxela]

So I don't quite get it, what house curve should I load into REW?

No house curve.

Fix your bass (not just your subwoofer), and apply the anechoic PEQ filters you found.

Then give it a listen. You may find you have achieved your goal.

If not, try seasoning to taste with a little high and/or low shelving. Careful not to overdo this.

Finally you may want to have a look at what could be placement-related cancellation around the 700hz mark. Depending on the nature of the issue, this may not be fixable with EQ or only to a certain extent.

HTH.

 

[Eetu]

I mean, probably I'm failing to understand you, because I've put that file into the EQ APO and the change in measuring was minimal, and the graph is still overall terrible... I'm looking for a house curve to import into REW EQ

Here's a screen
View attachment 201715


So I don't quite get it, what house curve should I load into REW? Where can I download it?

You can download the Harman house curve here. Remember to disable REW's 'Add room curve' which is not to be confused with the house curve you add via txt file.

Then let REW create the filters for you below ~300-400Hz (bass). Don't run full range correction based on the house curve, use it only to as a guide in applying bass correction.

Then add the 'speaker correction' EQ (mids & highs) from Pierre's Spinorama site separately.

Fine tune the 6dB bass shelving +- to taste.

 

[Marcus Aseth]

Well, I had already started (and finished) EQing following the Harman's house curve before you guys answered here, so I've ended up with this:

That is with Var smoothing. I don't know what's going on on 10k Hz, I swear in the prediction it was fine... could it be that those very high frequencies do crazy stuff like that if I mistakenly place the mic 1cm in the wrong position?
Anyway how terrible is it? x_x
Psychoacoustic smoothing is patting me in the back telling me "all is alright", but with the other smoothing there is a lot of this up and downs and I don't know what to think, need your opinion guys x_x

I should also not that's 3 files loaded/overlapped in EQ APO, the first has 17 filters, the second 11, the third 7, so is a total of 35 filters... how bad is that? xD

 

[HarmonicTHD]

Well, I had already started (and finished) EQing following the Harman's house curve before you guys answered here, so I've ended up with this:
View attachment 201768
That is with Var smoothing. I don't know what's going on on 10k Hz, I swear in the prediction it was fine... could it be that those very high frequencies do crazy stuff like that if I mistakenly place the mic 1cm in the wrong position?
Anyway how terrible is it? x_x
Psychoacoustic smoothing is patting me in the back telling me "all is alright", but with the other smoothing there is a lot of this up and downs and I don't know what to think, need your opinion guys x_x

I should also not that's 3 files loaded/overlapped in EQ APO, the first has 17 filters, the second 11, the third 7, so is a total of 35 filters... how bad is that? xD

As NTK quoted the inventor (Toole) of the Harman curve. It is not meant to be EQed to, but it is the natural response of a speaker which measures flat in an anechoic chamber. So don’t use it. Start simple - just EQ you bass (ca below 250Hz) flat. Then listen and and maybe add ca 3dB in that bass region depending on your taste. Also you have forced your speaker to play in the 20Hz region, where it is physically incapable of doing (see the speaker measurements from Pierre) this will at least add over proportional distortion if not worse. Don’t EQ above 250Hz for now, as you have completely distorted your speaker in that region (see the 10kHz bump) with what you have done. Plus it may kill your tweeter if you turn up the volume.

 

[ernestcarl]

I should also not that's 3 files loaded/overlapped in EQ APO, the first has 17 filters, the second 11, the third 7, so is a total of 35 filters... how bad is that? xD

That’s quite a lot… Though, I’ve been there before — going nuts with EQ…

If you don’t mind me asking, what speakers are these?

 

[Marcus Aseth]

I think I need some specific tips on how to properly EQ, because even if I'm aiming for flat, I get a seesaw result in the end...
First thing first, in the EQ tool from REW, should I have my flat line in the average point of the bass region or at the lowest trough, or somewhere inbetween?
Second, how much positive Gain am I allowed to have in my filters?
Anyway I've wasted at least 8 hours on this today, so I'm spent, I'll try again tomorrow x_x

Would really appreciate if you could share any information for doing this that you think a beginner might benefit from, so that tomorrow I can do actual progress rather than going insane down the wrong path xD

 

[HarmonicTHD]

That’s quite a lot… Though, I’ve been there before — going nuts with EQ…

If you don’t mind me asking, what speakers are these?

Kali LP6 provided the link the OP gave, is correct.

 

[Marcus Aseth]

Kali LP6 provided the link the OP gave us correct.

yes, Sub SVS SB1000PRO and speakers Kali LP6 v2

 

[Marcus Aseth]

Plan for tomorrow:
・measure the whole system together
・from REW EQ tool, set a flat curve at the lowest trough in the bass region
・let REW generate filters for the range 20 to 500Hz (since the measurement shown in post #11 has a big peak at 500Hz, so might as well go up there)
・import the filter in EQ APO
・repeat the process to generate a second filter which hopefully will flatten the response even more
・finally apply the correction mentioned earlier to EQ APO

Does this sounds correct? Following this plan, there will be parts in which REW will need to apply -15dB corrections, is that ok?