Home Theater Forum and Systems banner

House Curve Options

32K views 21 replies 7 participants last post by  Chrisbee 
#1 ·
Wayne A. Pflughaupt said:
Dialing in a house curve is not a "one size fits all" endeavor.
If you're new to house curves, start with Wayne's excellent dissertation here. I’ll refer to those ideas in this thread. I would like to share some of my recent experiences and ideas about house curves. Each of you will have your own preferences, but you might think of something new you'd like to try.

My last sub was a sealed box with one 10” woofer. The house curve that I settled on was probably influenced by the low end limitations of that sub, and not really by my true listening preferences. My recently completed IB subwoofer can dish whatever I ask for, and that’s why I’m tooling around with house curves again. I want to find a curve that sounds great with music and movies, giving both a natural sound and a powerful foundation, and earth shaking impact or rumble when called upon to do so. I call my goal lean and mean. Bass should not sound bloated or overblown, but should have plenty of low end power.

First, I'll look at house curve .txt files. The simplest house curve files use two coordinates to establish a frequency interval and a gain. For example, here’s an 8 dB rise from 80 Hz down to 31.5 Hz.

31.5 8
80 0

This line is not straight on the logarithmic scale. A flatter house “curve” can be written by specifying equal steps every 1/6 octave like this:

31.5 8
36 7
40 6
45 5
50 4
56 3
63 2
71 1
80 0

Here are the previous two house curves superimposed.

Text White Blue Line Pattern


You can see that the target level of the curved line is a couple of dB higher in the middle of the slope. That 2 dB hump doesn’t look like much, right? But a low, very broad hill in your frequency response is more audible than a narrower, higher peak. You might listen to the flat and the rounded house curves above, and you may prefer the rounded, or the flat, or not care at all. (OK, maybe my wife is right and I am obsessive.)

I started developing new house curves for my IB based on my preconceived “rules”.

1. Start the rise at the crossover frequency.
2. Stop the rise at 30 Hz.
3. Use a flat slope on the logarithmic scale.

After I settled on a crossover setting of 80 Hz, I started dialing in and listening to these house curves. They begin ramping up at 80 Hz and level off around 30 Hz. They slope up +6, +8 and +10 dB:

Text Line Pattern Design Parallel


I wasn’t entirely satisfied with the results. I wanted more low end power, but the more aggressive curves were getting too bassy in the musical range.

So I broke my “rule” number 2. I decided to keep the rise constant, and push the shelving frequencies down a little more. These all slope up from 80 Hz at 6 dB per octave and shelve off at 32, 25 and 20 Hz:

Text Line Pattern Design Parallel


Much to my surprise, I listened and discovered that I liked the changes. I prefer not to shelve off at 30 Hz. Gently rising all the way down to 20 Hz does not sound unnatural to me, and the power I was looking for was there. Still, I felt that the sound was a bit bass heavy in the upper half of the subwoofer range.

So I broke my “rule” number 1. I decided to start the rise below the crossover frequency. I moved the beginning of the rise down to 63 Hz and sloped up from there. At 20 Hz it is up +10 dB.

20 10
22 9
25 8
28 7
31.5 6
36 5
40 4
45 3
50 2
56 1
63 0

I have overlayed my new curve over the old to illustrate the difference.

Text Line Pattern Design Parallel


Success! Finally, a house curve that I love. Its lean and mean, natural and powerful. I’d recommend it to anyone with a similar set-up.

As Wayne has long since pointed out, all of this is based on preference, interaction with my room, interaction with my mains, and the behavior of my subwoofer. The “rules” that I broke were self imposed. Wayne's write up has always encouraged experimentation with all of these parameters. I've proven to myself that such experimentation pays off.

To match a target response I:

1. Measure the unfiltered response at four seats and use the average response.
2. Increase the subwoofer gain to get the desired level at 20 Hz.
3. Set a deep, wide filter at 1.00 kHz to adjust the response level around the crossover.
4. Add filters to eliminate the peaks.

Here's the subwoofer target response with an 80 Hz crossover. (This graph is not smoothed.)

Text Green White Blue Line


I have not exhausted all of the possibilities by a long shot. For example, I have only explored straight lines. But I’m homing in on what works best for me.

Please share your ideas and experiences. Play with it. You have lots of options!

Thanks to Wayne, Sonnie, John and Bruce.
 
See less See more
5
#2 ·
So I broke my “rule” number 1. I decided to start the rise below the crossover frequency. I moved the beginning of the rise down to 63 Hz and sloped up from there
Yeah, good subject. You have discovered (I believe), the same thing I was trying to get across in this thread, where I found that by lowering my crossover to 60Hz from 80Hz reduced that mid bass bloated sound.

I think if you have a cross at 80Hz, that it's just too soon to begin a house curve. I solved it by moving my crossover to 60Hz, but perhaps your method is better because it allows you equalization between 60Hz and 80Hz that I may not have..

brucek
 
#6 ·
Yeah, good subject. You have discovered (I believe), the same thing I was trying to get across in this thread, where I found that by lowering my crossover to 60Hz from 80Hz reduced that mid bass bloated sound.
Right. In your case (which is extreme :coocoo: admit it) you lowered the beginning of the slope to around 100 Hz. You could try starting the slope even lower, but I have a feeling you don't want it any leaner.:scratch:
 
#3 ·
Bravo, Naut! :clap:

I’ve often felt that this topic needed more than one man’s perspective, but so far no one has ever got back to me with specifics on their experiences. It would be nice if others would weigh in with details on their experiences, too.

I’ve made a couple of other observations myself since I wrote that article.

For one, I’ve switched from floor-standing mains that had 8” woofers and 10” passive radiators, with good extension down to 30 Hz, to bookshelf speakers with 6-1/2” woofs. I’ve found with these that I like a flatter house curve – there’s only 1-dB differential between 30 and 90 Hz. I can’t help but wonder if it’s because the sub is making up for what the tiny-woofer mains are lacking. Of course, the size of my listening space has increased by 50% (9200 cu. ft now), but I don’t see how that could fully account for a slope change from 7 to 1 dB!

I notice that both you guys have speakers with good bass extension, so that may account for your preference for a curve that starts lower. Maybe someday if you’re bored you could so some experiment using your surround speakers as mains. :)

Second, since I did the subwoofer review, I’ve been theorizing that you get better bass detail when response droops below 25 Hz. Of course, you first have to have some demo material that generates some detail – there are a couple of recommendations in the sub review that I use. Naught, you didn’t mention if your satisfaction with the 20 Hz shelving was as pleasing for music as it is for movies, so I’m still wondering about this one. My subs start dropping naturally at 25 Hz, so I can’t experiment with it myself.

Regards,
Wayne
 
#7 ·
Thanks for your enthusiasm Wayne.
I notice that both you guys have speakers with good bass extension, so that may account for your preference for a curve that starts lower. Maybe someday if you’re bored you could so some experiment using your surround speakers as mains. :)
I thought of the same thing. Some receivers apply a 2nd order roll-off to the mains that is supposed to sum with the natural roll off of the speakers to create a 4th order high pass. At 80 Hz, my crossover is above the natural roll off of the speakers, and so may tend to over-emphasize the mid bass when combined with a house curve that's too high.

I have my reasons to use an 80 Hz crossover instead of 60 Hz.
  • My mains excite a room resonsnce between 90-130 Hz, and this is much tamed by an 80 hz crossover.
  • I don't want to low pass the LFE below 80 Hz.
  • My surrounds poop out around 100 Hz, and I only have one crossover setting for all "small" speakers.
  • I like to unburden the reciever, which is only rated at 115 W/channel.
Pushing the start of the house curve ramp down to the point where the mains naturally rool off seems to result in a much better integration.
Second, since I did the subwoofer review, I’ve been theorizing that you get better bass detail when response droops below 25 Hz. Of course, you first have to have some demo material that generates some detail – there are a couple of recommendations in the sub review that I use. Naught, you didn’t mention if your satisfaction with the 20 Hz shelving was as pleasing for music as it is for movies, so I’m still wondering about this one. My subs start dropping naturally at 25 Hz, so I can’t experiment with it myself.
My sub's last peak is at 17.6 Hz and it's down -6 dB by 16.0 Hz; there really isn't a wide plateau. I think that it sounds better when I don't mess with the natural roll off.

My impression is that the musical information down below 30 Hz is mainly drum pressure. I don't think that what happens below 25 Hz has much impact on musical detail. Can you explain your reasoning?
 
#4 ·
Hey Ayreonaut,

Yeah, nice job. I'm tempted to play with my hourse curve as well, though I don't hate it the way it is. I've recently relegated my last two "very long" RCA cables to basement crawlspace duty, so I don't have a means to get the signal from my RS SPL meter to my PC. But next time I'm off tweaking, I'll give these a try.

I've also gotta give it to you for making it through all those. The averaging stuff must add up as well. It must have taken you quite a while to get through all of it.
 
#8 ·
I've also gotta give it to you for making it through all those. The averaging stuff must add up as well. It must have taken you quite a while to get through all of it.
OK, here's my dirty little secret. :blush:
In my room, using only cuts, and taking the averages, REW is always spot on. John, :yourock:
After a year of learning to nail the target responses with REW, setting the BFD, and measuring the results,
I have come to trust the REW predicted responses completely. And now I try new house curves regularly,
and I don't really need to measure the results every single time. :bigsmile:
 
#9 ·
Hey Naut,

My impression is that the musical information down below 30 Hz is mainly drum pressure.
Actually, drums don’t have any fundamentals that low. That ultra-low impact you get with some recordings is artificial, the result of equalizing.

Aside from that, you’d be surprised at how much information there is in music below 30 Hz. When I did the four-sub comparison a couple of yeas ago, one of them practically brick-walled at 30 Hz, while the others went lower. The sub with the lowest extension sounded the best, while the one that went only to 30 Hz sounded the worst. The lower sub added a noticeable “foundation” or underpinning to the bass that made it sound more substantial.

Most likely it has to do with harmonics. You hear a lot of talk about first and second harmonics and such, which are one and two octaves, respectively, above the fundamental, but harmonics extend below the fundamental as well. A sub with good extension can bring that out.

I don't think that what happens below 25 Hz has much impact on musical detail. Can you explain your reasoning?
You’re probably right about that. That’s why I suggested curbing response below that point. That’s what I meant by “I’ve been theorizing that you get better bass detail when response droops below 25 Hz.”

Regards,
Wayne
 
#19 ·
That ultra-low impact you get with some recordings is artificial, the result of equalizing.
True. I've heard plenty of artificial bass.

When I did the four-sub comparison a couple of yeas ago, one of them practically brick-walled at 30 Hz, while the others went lower. The sub with the lowest extension sounded the best, while the one that went only to 30 Hz sounded the worst.
I haven't done a comparison with a brick wall filter. I'm sure that its important to have the extension. I compared curves that level off, with curves that continue to rise. I noticed no bloating or loss of detail with the curves that continued to rise. (I haven't listented too any pipe organ music yet, I'll have to try that.)
 
#10 ·
that was a lovely bit of work Naut, and you must be a mind reader cause the day you posted it was the day i was gonna check the theory of house curves out!

Wayne, i was very intrigued by the statement that a fundamental had harmonics below it! I remember 'shooting ' someones argument by pointing out by the definition of fundamental it must be the lowest ( cant find an embarressed face! ). Will have to get my head around a new concept

lots of love

terry
 
#11 ·
I remember 'shooting ' someones argument by pointing out by the definition of fundamental it must be the lowest ( cant find an embarressed face! ).
Wow, that guy gave up really easily! If he had checked his Webster’s he could have come back that “fundamental” means “primary, basic or root.” :)

However, for a definition as it applies to music you need a dictionary of musical terminology. This one I found on-line defines a fundamental as “A base pitch from which a series of harmonics is produced.”

Even the musical dictionary and related sites only describe upper harmonics, however, and one site I saw even said that subharmonics do not occur in “natural” tones (whatever that is). I’m not buying it.

Consider the common charts that tell us what frequency ranges certain instruments “live” in, that typically say cymbals generate frequencies between 8-16 kHz. Anyone here think you can get everything a cymbal generates if you play it through a tweeter that’s been brick-wall high passed at 8 kHz?? Not a chance. Anyone who’s ever heard a real one knows they are generating resonances far, far below that, especially the larger “ride” cyumbals. Same thing with hand bells – they are rich in harmonics above and below their fundamental tones.

Regards,
Wayne
 
#13 ·
to be honest that's what I would have thought, but I don't claim to be an expert.

One way to look at it is a single guitar string. the lowest ( this is from memory but we'll see) note it can sound is the longest possible thing 'waving' right? ( wow, precision terms!) That in effect means the whole string vibrating and only anchored at the ends.

As we sound higher and higher notes, the string will 'divide' into two, three etc etc. But of course there is no way for the string to get 'longer', hence 'lower'.

where am I going wrong here??
 
#14 ·
The full spectrum of sound from the instrument includes components from the envelope of the output, i.e. the shaping of the fundamental and its harmonics over time. That lower frequency content might be referred to as sub-harmonics, although the term is technically meaningless.
 
#16 ·
Maybe I’m not getting the terminology right for what I’m talking about – maybe “resonance” is a better word than “harmonics.” In fact, the more I think about it the more I think harmonics probably isn’t the right word, because harmonics are at recognized intervals above the fundamental.

What I’m talking about has to do with the way we are able to clearly hear bass notes, whose fundamentals are down to 40 Hz or even lower, with speakers with 6” or smaller woofers that are only rated for 100 Hz, or even higher.

For instance, I recall reading somewhere that in Motown’s heyday back in the ‘60s, they intentionally mixed and “voiced” bass so that it could be heard on the cheesy transistor radios all the kids listened to in those days. They even had a transistor radio speaker in the studio for reference, they said. Shoot, I used top practice bass guitar lines to music from a portable cassette deck that had maybe a 3-4” speaker!

The point is, there is a lot of information in an instrumental tone that enables us to hear bass, for instance, or cymbals, even when the speakers we’re listening to are unable to generate the fundamentals of the notes.

Whatever you call that, IMO it’s being generated below the fundamental as well as above it, and you can tell it’s there if you have a sub with good extension. Get a 30 Hz sub and a 20 Hz sub side by side and you can definitely tell the difference, even when the fundamentals are 30 Hz or higher.

There - did that help, or make things even more murky? :huh:

Regards,
Wayne
 
#17 ·
aaahhh, the murky world of Mr Pflughaut, aka Waynes Murky World!

I understand your thinking on this one, have pondered it myself at times. Bass wise, it actually is not a 'severely' deficient experience listening to most speakers, even tho they often can only go down to let's say 60 hz for the sake of this argument. At times when I'm mucking about measuring and turning mains and subs on and off, then listen to music etc, forget to turn the sub back on yet still enjoy a relatively 'full' bass experience.

But Wayne, if the FR measurement does not show a signal lower down, then how can we call it a 'lower resonance'? It's not there being picked up by the microphone so we have to conclude it doesn't exist.

Does that mean it is some sort of psychoacoustic thingamebobby?? ie the mind can hear a certain combination of harmonics, and 'back engineer' them to the correct bass fundamental and 'hear' that??

You're right, I've gone and confused myself now ( and everyone else ) .

Bit off topic, so deftlt resurrecting the thread I'll just say that I've worked out how to load a house curve, so may the fun begin

lots of love

terry
 
#18 ·
But Wayne, if the FR measurement does not show a signal lower down, then how can we call it a 'lower resonance'? It's not there being picked up by the microphone so we have to conclude it doesn't exist.
Hmm... I’m not sure you can adequately measure the “frequency response” of a musical note. Maybe John or brucek can comment on that.

Regards,
Wayne
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top