Can Harman preference curve be applied to home Theater applications, center, surrounds, etc?

[Vacceo]

I have a 7.1 all KEF system. It works perfectly for me. The only thing I did was adding a bit extra gain on the subwoofers, but that´s it.

My speakers are all part of the old IQ line and honestly, I´m not missing anything.

 

[ernestcarl]

Can you give an example with some specific loudspeaker, measurements and recordings?

These were convolved measurements from before (though he uses different speakers now):

What is your favorite house curve

Extremely low frequency. MORE, MORE!

www.audiosciencereview.com

I imagine the flatness of the envelope (after processing) is similar. But I do not know whether @dasdoing thereafter applies a more downwards sloping target or keeps it as it is.

 

[dasdoing]

there are two types of neutrality

trying to explain better as this probably sounds stupid, since neutrality is only one state: same FR balance on playback as original.
now this is in an anechoic chamber. we are now in a room, which has it's own sound, no matter how much you treat it. let's say you pop a baloon in a church and record it with a perfect mic...everything is perfect. the soundfile you got is perfect. now stereo will never reproduce that ambience perfectly. it will always sound like if you are standing outside the church with your head in an open window. with the flat direct sound you bring the baloon closer to the window. that's obviously not realy correct, since the direct to reflected ratio doesn't change...but that's the psychoacoustic impression I have. with the flat envelope you move the midrange backwards where the bass will always sit since the room will dominate there.
so you have two "neutralities" here even the midrange meassures diferently. they both seam to replicate the real event in terms of frequency response. what changes is the image.
on a sidenote: a none smoothed flat envelope in my room, when looked at it with var smoothing looks close to the Dr. Floyd Tool curve (not Harman) mentioned here: https://www.audiosciencereview.com/...esponses-in-csv-txt-format.16401/#post-531654

 

[DVDdoug]

There are some screwy things about movie audio... something about mis-interpreted measurements and mis-applied results ending-up with movies mixed on inaccurate systems. See this post by Amir: Validity of X-curve For Cinema Sound.

If I understood and remember correctly... The movie-mix generally ends-up with rolled-off highs. Or, maybe I've got it wrong and it's the commercial theaters that are EQ'd wrong.

 

[dasdoing]

I imagine the flatness of the envelope (after processing) is similar. But I do not know whether @dasdoing thereafter applies a more downwards sloping target or keeps it as it is.

I keep the top (envelope) as flat as I can get it.

 

[dasdoing]

btw this whole envelope method builds on the fact that we percieve peaks, not the dips (*). the fact that it seams to produce (smoothed) house curves close to those created by trail and error (sans bass boost) is strong evidence towards it imo. one can still aply a harman bass boost to it, just adust the envelope curve acordingly.

(*) there are papers out there studying envelope of instruments, and how the envelope is important to percieve the instrument as what it is.

 

[ROOSKIE]

Care to supply a link to the other one?

The typical Harman speaker "curve" is a downward slope from about 200hrz to 20k and rise from 200hrz to 20hrz. This is in listening room only. The slope varies and is not fixed or a goal.
Generally .4-1db per octave during the fall to 20krz and maybe 5-10db boost from 200hrz-20hrz.

The info is all over the site and including all speaker reviews and is easy to find on the net as massive research went into it.
Flat (on axis)speakers in anechoic chambers with even, smooth fall of in off axis develops a downward slope in room and people need extra bass.
Read Floyd Toole's book.

 

[kongwee]

It is time for your home theater to be audiophile grade.

 

[ernestcarl]

btw this whole envelope method builds on the fact that we percieve peaks, not the dips (*). the fact that it seams to produce (smoothed) house curves close to those created by trail and error (sans bass boost) is strong evidence towards it imo. one can still aply a harman bass boost to it, just adust the envelope curve acordingly.

(*) there are papers out there studying envelope of instruments, and how the envelope is important to percieve the instrument as what it is.

@mitchco's way of showing it in an overlay view together with the psychoacoustically "smoothed" traces seem more natural to me here:

Audiolense Digital Loudspeaker and Room Correction Software Walkthrough

In this article, I walk through the steps of using Audiolense 5.0 (AL5) to biamp my loudspeakers using a linear phase digital crossover. Followed by frequency and time domain correction, so the frequency response is tonally neutral and the direct sound is arriving at my ears all at the same time....

audiophilestyle.com

individual Responses Before EQ

Still, I find that my ears tolerate peaks and dips as seen in the measured traces perfectly fine as long as the overall full (and direct) response remains "flattish" enough -- i.e. I personally wouldn't necessarily try to clip off all the peaks just to keep the measured upper envelope "flat" based on a single-point measurement.

 

[freemansteve]

I've just read up on this Harman curve stuff..
I conclude that I cannot prefer the same cat food that 9/10 cats "prefer".
Too many systems are just wrong because of this consumer vox-pop nonsense. Just look at what the dweebs listen to in their 8 minutes of paying attention to music in any given week.

 

[Haint]

From Dr. Toole's "Measurement and Calibration of Sound Reproducing Systems" paper. Looks like a "good average" room curve to most listeners would be a linear ~+6dB bass boost from 20 - 60hz, ~-5db roll off to 200hz, then maybe a ~-2dB roll off to 10Khz. The gotcha here is that even 1dB adjustments here are hugely audible as the regions are so broad and poorly designed speakers probably won't equalize well.