Wayne/Brucek - House curve SVS 16-46PC sub - need a starting point suggestion . . (1 Viewer)

[gomez_a]

Wayne, these Eq threads just never end do they huhh?
. . . . as you know, I have an SVS 16-46PC sub, since this subwoofer goes strong down to 16Hz, what type of house curve would you recommend for me to start with?
I know this varies from room to room, personal pref's, actaul sub in use, human perception of loudness at various freq's, etc, (I've been reading all these BFD/EQ/housecurve threads) - but I need a suggestion just to get me started.
I have a 12wx19lx8h room (1824 cu ft) with my fronts, center, TV, and front-corner-loaded sub all located along the 12w dimension - this listening room opens up to a kitchen,which I'm guessing, is 12x12x9h. The opening between my listening room and kitchen is about 8'wide and 7' tall.
In one post for a forum member you suggested he start with a "9dB house curve from 100 to 32Hz" , which is fine. But given the 16-46's low end extension, would you suggest I start with a:
1. 9dB curve from 100 to 20Hz
2. 9dB curve from 100 to 32Hz and plateau/flat from 32 to 20?
I just need something to get me started . . as always, thanks!

 

[brucek]

gomez_a,
If indeed you have wisely taken Wayne's advice regarding the Polk, and if the response of your SVS 16-46PC is as shown on your web site here:
http://ncstatesman.tripod.com/angelswebsite/id1.html
I would then say you've already got a fairly nice natural curve happening except for the peak at around 37Hz. My advice would be to start there and reduce that simple peak and see how it sounds. If you like it, carry on and smooth out the rest.. If it's too heavy in the bottom, tilt the curve down a bit.
Wayne may disagree..
brucek

 

[Wayne A. Pflughaupt]

Anyone who disagrees with brucek usually ends up extremely embarassed. Ken, it’s great to see you’re finally getting active on this forum.
Angel,
Brucek is right, you already have the makings of a good house curve, due in part to cabin gain from your relatively small room. While your room does open to the kitchen, you still have most of four walls in place, which will give you more cabin gain than if the area was entirely open. (By the way, I hope the sub is not in a corner near that opening...)
My 9dB slope is in a room more than three times larger than yours; the house curve in your room you will need a steeper slope. I can’t make an exact recommendation, since I’ve never set-up a room that small, but I imagine something like 12-15dB would be a good starting place.
Here’s my recommendation for a seat-of-the-pants method to determine your curve: Ignore the numbers and go by ear. Test tones at say, 32Hz, 56Hz and 100Hz should sound like they are at the same volume. The lower frequencies will measure louder, but they should sound about the same. Once you get in the ball park, then you can start paying attention to the readings.
But first you need to smooth out your response. In addition to the situation at 37Hz that brucek mentioned, I would reduce the area at 89-98Hz by 5-6dB (to the 60dB level on your chart). This should set you up nicely for a 15dB curve. Of course, see what you can about the low spot at 70Hz and the high spot at 62Hz as well.

Given the 16-46's low end extension, would you suggest I start with a 9dB curve from 100 to 20Hz?

I found I didn’t like the way things sounded when I continued the curve upward beyond 32Hz. It gave me a lot of annoying and unnatural-sounding ultra-low energy. For instance, when a car door slams you shouldn’t hear it all the way down to 20Hz. That’s the kind if stuff I was getting.
But again, your room is much smaller than mine - that might make a difference. All I can say is experiment. However, running the same slope from 100 to 32Hz out to 16Hz will get you another 6-9dB or so of gain - that’s pretty substantial.
Another thing relating to the differences in our rooms is that my starting point of 100Hz might not work for you. You might need to move the starting point up or down, say to 80Hz or 125Hz, for instance. Again, listen as well as measure.
Hey, I never said dialing in a house curve was easy.
By the way, when you take your next readings, Angel, you need to raise the volume. When you get below about 65dB you run the risk of ambient room noise polluting your readings.
Good Luck,
Wayne A. Pflughaupt

 

[gomez_a]

Wayne/Brucek, Thank you for the response and pointing me in the right direction.

The sub is in the front left corner opposite to the opening to the kitchen, ie, it's fully corner-loaded for max effect. I should probablty take down the readings in my website - those reading were taken with the SVS sub sitting on top of the polk sub. I'm going to retake the readings over the next few days, but this time the SVS will sit on the ground (w/the Polk removed from the area - I'm trying to sell it right now). I'll alos retake the reading with the sub situated in the front right corner of the room (again, the corner oppo to the opening to the kitchen) . Once I have these 2 new curves, I'll have a much better idea on how to apply your suggestion in EQ'ing the sub with the BFD. I'll try EQing a flat response and a housecurve according to your suggestions.

Brucek, since I'm getting rid of the Polk sub, I have an extra Monster bass RCA cable that is about 18' long, so in effect, I'll have : RCVR===18'===>BFD===18'===>SVS sub, will this length cause any signal problems?

Wayne, I tooke these readings relatively low because i was conserned about frying up the coil in the driver. I'm using an Autosound #102 CD w/sine waves from 10Hz to 98Hz. This CD does not warn about playing the sinewaves high as I think they were U]not[/u] recorded at 0dB Full Scale. Using the test tones in the rcvr (which I know are not very accurate) - I calibrated both subs to 75dB (includes the 3db comp factor) and set my Onkyo 898 rcvr vol to -20. Can you please recommend a higher vol setting?

Again, Wayne, Brucek, thanks for the guidance!

Angel

 

[brucek]

Angel,

concerned about frying up the coil in the driver

It's a good idea to be taking your readings (as Wayne says) of a sufficient level as to not be influenced by room noise. Certainly average somewhere around 85dB SPL would be fine. That way the low end might go as high as 90dB and the top end at the cross point might be around 80dB.

You are correct about frying your driver though. When I made my CD for tones, they are recorded as 3 second tones each. To me, this is long enough - others may disagree and say I haven't fully excited the room mode enough. Well, that's fine, but I won't be overheating my driver either. Playing a single low frequency tone is very hard on the driver - there's a big difference between a single tone and music.

On that same note (pardon the pun), I've read many times about overcompensating for room modes a bit too much with equalizers. The theory being that normal music will not excite a mode as highly as a single tone will. Warble tones at these low frequencies will not excite the mode enough, so a single tone is used instead, but maybe there's a bit of a compromise needed.

I've certainly noticed at frequencies where I know the room has a large influence, that when the tone first starts, my SPL meter will jump immediately to a value and then as the tone plays for the next few seconds, the SPL slowly increases at least by a few dB. What value of SPL do you use in these instances. Maybe Wayne might like to give his thoughts on this subject.

brucek

 

[gomez_a]

Hi Bruce, when taking the readings i was careful to eliminate all ambient noise (I even had to turn of my heating system). To verify, throughout my readings I would check for ambient noise by setting my SPL to it's lowest setting (60) and checking for noise - none registered on the meter.

Also, I adjusted my subwoofer gain to yield a 75dB (72dB on the meter) reading while playing the rcvr's test tones (I know they are not accurate, but all things being equal . . .) Also, all spkrs were off/disconnected, slow responce/C weighting used.

Thanks,

Angel

 

[Wayne A. Pflughaupt]

brucek wrote:

To me, this is long enough - others may disagree and say I haven't fully excited the room mode enough. I've read many times about overcompensating for room modes a bit too much with equalizers. The theory being that normal music will not excite a mode as highly as a single tone will.

I’ll be the first to admit that I never “got it” regarding room modes, how to fully “excite” them, etc. Seems to me if sound travels at the speed of – well, sound – then what’s with this “they take time” business?

Nevertheless, brucek, your observance of having the meter climb after initial readings at certain frequencies is a good place to explore this. I take 5-6 second readings with sine waves, so maybe I was giving enough time for the adjustment without realizing it. With filtered pink noise I would set the meter for a 20-second time lapsed reading, and use the average.

I originally set my EQ with filtered 1/3-octave pink noise, which supposedly doesn’t excite room modes. Recently I took new 1/6-octave sine wave readings (5-6 seconds each, as described above). Other than a couple of situations the 1/3-octave readings couldn’t reveal, there was nothing in the readings that gave an indication that I needed to change anything.

Bottom line, (in ignorance) I didn’t consider room modes when taking my readings, but I successfully EQd my system anyway. This may be because I used my ears as the final decision-maker, not the readings alone.

Regards,

Wayne A. Pflughaupt

 

[Richard Greene]

YOU WROTE:

I’ll be the first to admit that I never “got it” regarding room modes, how to fully “excite” them, etc. Seems to me if sound travels at the speed of – well, sound – then what’s with this “they take time” business?

RG REPLIES:

-A room mode builds up in the time it takes sound to

reflect between at least two surfaces and return to your ears. That means less than one second to build up ...

and then after the test signal is stopped, a typical plasterboard wall room will "ring" for up to 200-300 milliseconds before the sound attenuates by -30dB.

- Only steady-state sine waves and very slow sine wave sweeps will fully excite room modes.

- Sine wave warble tones (often five warbles per second) don't stay at a specific room mode frequency long enough to fully excite the standing wave. Pink noise is too random to excite room modes very much.

- Track 5 on the Stryke Basszone test CD ($12 from Stryle.com) is a very slow sine wave frequency sweep from 1Hz. to 100Hz. (why 1Hz. rather than 10Hz. as the starting point I'll never figure out).

- Use that track to measure the total bass frequency response deviations (including fully excited standing waves and fully excited wall rattles, floor rattles, ceiling rattles and resonances of every other room object!

-Use your parametric equalizer to reduce the frequency response peaks measured with steady state sine waves

(1/6 octave spacing is okay, but 1/10 octave spacing between sine waves is more accurate) to reduce frequency response deviations.

-I think a good target after equalization is +/- 5dB from

20 Hz. to the upper limit of the subwoofer (+/- 5dB may seem like nothing special ... but remember that all the room resonances are included and fully excited by a very slow sweep tone). Before equalization, frequency response deviations measured at the listening position using a slow sweep tone are usually worse than +/-10dB. (sometimes a lot worse).

YOU WROTE:

Nevertheless, brucek, your observance of having the meter climb after initial readings at certain frequencies is a good place to explore this. I take 5-6 second readings with sine waves, so maybe I was giving enough time for the adjustment without realizing it. With filtered pink noise I would set the meter for a 20-second time lapsed reading, and use the average.

RG REPLIES:

-The standing wave will probably reach their maximum amplitude faster than you can read the sound meter and write the data on a piece of paper.

-I've been equalizing subwoofers for about 20 years using my Rat Shack meter set on C-weighting and Fast Response. The meter reads approximately the same after one second as it does after five seconds or ten seconds or 100 seconds.

-Pink noise is great for equalization of frequencies above 100Hz. but sine waves and slow sine wave sweeps work better below 100Hz. because they allow you to accurately measure narrow standing-wave-related frequency response peaks rather than smoothing the peaks as pink noise does.

- A very simple way to start experimenting with a bass house curve is to use the Rat Shack meter on C-weighting without any corrections. C-weighting already has a bass house curve built in compared with unweighted dB measurements (U-curve).

 

[brucek]

Thanks to Wayne and Richard for interesting replies.
I have to admit whenever I have used my 1/6 octave tones to record and graph my low frequency response, that once I sit down to enter a filters center frequency I will play "every" one hertz tone around the peak to find it's highest point. This is the center frequency I use for the filter - seems to work.
Richard, I was aware that warble tones were used for higher frequencies and steady state tones for low frequency response testing. In that regard it makes me wonder if using a single frequency is representative of the subs end use. Music's complex spectrum hardly allows for many sustained tones, so I was wondering if we perhaps over compensate while using single tones as our test source for low frequencies?
My feeling on the increase in room SPL level at some frequencies that I have observed (since you indicate it isn't increasing room gain) may well be attributed to the noise of items in the room rattling at those frequencies and then adding to the overall SPL - who knows.
brucek

 

[Wayne A. Pflughaupt]

Great information, Richard!

My feeling on the increase in room SPL level at some frequencies that I have observed (since you indicate it isn't increasing room gain) may well be attributed to the noise of items in the room rattling at those frequencies and then adding to the overall SPL - who knows.

Believe it or not, this isn’t really a problem as long as you’re taking your readings high enough. I was taking some informal readings one day while my daughter was rattling dishes and running water in the kitchen (behind the living room and wide open to it). When I kicked in the sine waves, I was more than 10dB above all that!

It’s the Fletcher-Munson thing: Bass frequencies that are perceived to be as loud as higher frequencies will register higher SPL than the latter.

Regards,

Wayne A. Pflughaupt

 

[Richard Greene]

RG:

- I'm not sure what you meant about not EQing for room modes as the only reason I've used equalizers with all my DIY subwoofers for the past 20 years was to reduce frequency response peaks caused by room modes.

-If placed outdoors where there are no standing waves, my subwoofers measure relatively flat (most subwoofers measure relatively flat when located outdoors or in an anechoic chamber) and would only need equalization to boost output under 50Hz. (no room gain outside either).

- I have measured many rooms using slow sine wave frequency sweeps over the past 30 years and have only found one room whose bass frequency response was within +/-6dB without equalization. So that was the ONLY room where bass equalization would not have made a major improvement.

.

.

.

.

.

Track 5 on the Stryke Basszone test CD is a very slow sine wave frequency sweep from 1Hz. to 100Hz. Use that track to measure the total bass frequency response deviations

--------------------------------------------------------------------------------

Hmm. The only thing I don’t like about this method is that you have no way to precisely pinpoint offending frequencies.

&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

RG:

The sine wave sweep is used to measure frequency response deviations before and after equalization. It works pretty well just using your ears too. Since the primary goal of subwoofer equalization is to reduce frequency peaks heard at your listening seat, the result of equalization should be considerably smaller frequency response deviations -- perhaps +/-15dB before equalization versus +/-5dB after equalization from 30 to 80Hz.

You still need steady-state sine wave tones to equalize subwoofers & rooms properly ...

... and you still need pink noise or warble tones to equalize satellite speakers properly.

.

.

.

.

.

.

--------------------------------------------------------------------------------

- A very simple way to start experimenting with a bass house curve is to use the Rat Shack meter on C-weighting without any corrections. C-weighting already has a bass house curve built in compared with unweighted dB measurements (U-curve).

--------------------------------------------------------------------------------

I found my house curve needed to start at 100Hz, so that wouldn’t have worked for me – the RS meter’s serious response deviations don’t kick in until about 40Hz.

&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

RG:

My own house curve needed to start below 50Hz. (I use a 50Hz. 24dB/octave low-pass filter for my subwoofer) and ended up similar to the standard C-weighting curve

(that's what C-weighting was intended for -- it's an "average" house curve developed from "average" hearing measurements" at an "average" SPL.)

I understand your concern about applying a house curve below 30Hz (as I do)because sometimes that is counterproductive:

- Sometimes there are unintentional very low frequency sounds in recordings caused by HVAC systems in the recording studio? or subway trains running under the recording studio? that would be less annoying if my subwoofer rolled off more below 30Hz. These problems are rare but do seem more common in older recordings where the ultra low bass artifacts were obviously not heard through the (ported?) speakers used by the mixing engineer.

 

[AndyDC]

Richard Greene,

How low is "very slow" for a sine wave sweep?

As I've posted elsewhere, I get similar results with a ~1 minute 16-100hz sweep as with 1/6 octave tones, for the frequencies that are common to both measurements (the sweep gives much more detail, which I found to be important). That implies to me that 1 minute may be slow enough, but I wonder whether that's what you had in mind? (It may be that if I did 1/12 octave tones, for example, I'd find that the sweep was not matching so well).

Thanks,

Andy

 

[Steve Morgan]

BruceK, Wayne,

You mentioned that the sinewaves are very hard on voice coils. I downloaded the tones from snapbug and I believe they are 8 second tones fading in and then out. I would hate to see my new PB2 voice coils burn out. Are these tones reasonably safe? I want to thank you guys for sharing your thoughts and info on EQ'ing a room. Wayne has helped me each time I purchased a new sub and so has BruceK when it came to wrangling with the BFD. Not to highjack this thread but , Say Wayne don't those new numbers look strange?

Once again thanks for sharing.

Later,

Steve