Do any BFD (or any digital parametric eq) users have suggestions on how to set the delay times in the 950 to compensate for the millisecond delay? A FAQ on the use of the BFD suggests adding a foot to the distance entered for the sub but we don't have that setting on our 950s. Should I shorten up the mains, center and surrounds?

I set my eq up last week and I have smoothed out the bass response but my subs don't sound as tight as they did before I put this thing in the signal path. Any suggestions would be appreciated. Thanks.

$

With the BFD in the loop and engaged, just set the phase as you normally would between the mains and your sub. That takes care of it.

Thanks, Kevin.

This was passed on to me. (See below)

I agree with it (and always have).

The BFD is a low end piece to begin with. I think that believing the 1ms spec is akin to believing speaker response specs.

I've heard that it could be as much as 20 times that spec as far as the total delay after EQ.
____________________

quote:Tom Ascher wrote:
> It appears that the primary use of equalizers is to get rid
> of room resonances that result from systems being in spaces
> considerably smaller than auditoriums or theaters where music
> is generally heard.

That might well be the most popular use, but it's the wrong use.
It's a waste of an equalizer and it is garaunteed NOT to do the
job.

> But, the broader issue is that any room can be considered an
> extension of the speaker enclosure and will introduce
> resonances at various frequencies throughout the audible range
> and that it really is necessary to have some way of "voicing"
> a system to a given room.

No, any room CANNOT be considered an extension of the speaker
enclosure. This is because of the inherent delays due to room
resonances put these effects well outside of the realm
associated with enclosure issues.

> Of course, speaker selection and placement, mounting, use of
> room treatment materials are all part of adjusting a sound
> system so it sounds good in a given room. But, there remain
> some problems that really cannot be dealt with short of some
> form of equalization.

Let's look at the falacy of this overall approach. Let's assume,
for the purpose of discussion, that the axial frequency response
and the power response of the speaker is perfect: it is dead
flat from 20 to 20 kHz anechoically. Let's further assume that
we want to maintain that response at the listener's ears. (We
can geeralize the assumptions to say that whatever the response
of the speaker is, flat or otherwise, we want to preserve it
by the time it reaches the listners ears.)

Now, let's play the loudspeaker in the room. Sit 3 meters away
from it. The first thing to reach your ears is the direct sound
from the speaker, completely unaltered by the room. Whetever the
response of the speaker is anechoically, THIS is what reaches
your ears first.

Now, let's say your room has a nasty resonance at some
frequency. In order for that resonance to affect what you hear,
the energy from the loudspeaker has to travel to where that
resonance is formed, it has to excite that resonance and that
resonance has the effect of dramatically extending the
reverberation time at that frequency, and the delayed,
reverberated result now has to travel the remaining distance to
your ear in order for you to perceive it.

And it arrives at your ear, AFTER the unaffected direct sound
made it and stimulated your ears.

Now, put an equalizer in and try to correct that resonance.
Let's say that the resonance cause a 15 dB peak in the response
at 300 Hz (not likely, but let's pretend). So you take your
equalizer and dial in a -15 dB hole in the electrical signal
going to the speaker at 315 Hz (because you went out and bought
a fancy 1/3 octave equalizer, none of which have a 300 Hz band,
by the way).

Now, let's play the loudspeaker in the room. Sit 3 meters away
from it. The first thing to reach your ears is the direct sound
from the speaker, with its new 15 dB hole in the response,
COMPLETELY UNALTERED BY THE ROOM. Whatever the response of the
speaker is anechoically, NOW modified by a 15 dB hole in the
response, THIS is what reaches your ears first.

And at the same time, the energy from the speaker travels about
the room, and STILL excites the room resonance (but at a lower
level) and STILL gets delayed and reverberated (sorry, but the
reverberation time at the resonance is not changed by the amount
of energy you put in it).

So, instead of a perfect speaker affected by a room, you now
have a speaker with a big whopper 15 dB hole in the response
that's somewhat less affected by the room.

A room's acoustical problems are fixed by dealing with the
causes of the room's acoustical problems, not by screwing up the
signal fed into it.

The problem is that MOST of the instrumentation available to
most people for "measuring" room response are completely
inadequate to the task. 1/3 octave real-time analyzers, pink
noise generators and sound level meters, warble tones and all of
it simply lose ALL the important time information that describes
WHY the room behaves the way it does. It hides much f the
information altogether.

You have a room with a nasty side-wall reflection? Propose how
an equalizer eliminates that reflection and it's effects? Bad
floor-ceiling slap echo? How and why would an equalizer fix it?
(Hint: it won't). Reverb time of the room about 1/4 second
EXCEPT at 200 Hz, where it's 1 second? How does something which
operates in the electrical frequency domain correct for energy
stirage and delay in the time domain, especially when that agent
is no longer causally connected to the stimulus?

Yes, you're absolutely right:

"It appears that the primary use of equalizers is to get rid
of room resonances that result from systems being in spaces
considerably smaller than auditoriums or theaters where music
is generally heard."

But they are ineffective at doing so.

--
| Dick Pierce
_______________

...Whaddayathink?

From what I've read, the BFD meets or beats its specs. It's a pretty good piece of equipment for equalizing a sub. There's no point in spending a ton of money because you aren't going to hear the difference for low frequencies.

If equalizing is so bad, then how come every professional studio, concert, etc. makes use of it?

tm

I obviously disagree.

I use ETF5 software with my PC and Radio Shack meter to look at freq response in my room. And how can I say this: it does the measurement "fast enough" that there is not enough time for room resonances to occur. (There's a word for what it does, but I can't remember at the moment.)

It's easy to see the benefit. Without eq'ing (and I actually only go after the worst peak), I'm +/- 7.5 dB from 20 to 200 Hz. With the eq, I'm down to +/- 5 dB. I've had up to 5 peaks attenuated in the past, and I can get to +/- 3.5 dB, but in my book, simpler is sometimes better.

Shoot, even SVS recommends eq'ing for sub interactions with a room. And they *build* subs for a living. They probably have a little bit of an idea of how they interact with the room too. And Velodyne incorporates an automatic eq'ing system in their new Digital Drive subs. Not 1/3 octave: parametric digital eq's with precise control over wavelength, bandwidth, and amount of attenuation.

We all know placement and room treatment is the best way to deal with this stuff. But failing that, a parametric eq is a very acceptable way of dealing with sub (low freq) room interactions.

DB is a bass player. I always tend to accept a bassist's opinion of how low freqs sound during playback before most anyone else.

He obviously can hear the time smear. If he can hear it, I seriously doubt that it's a 1ms delay that he hears.

Time smear does NOT show up on a FR graph.

It would be great to think that a phase knob on a commercial subwoofer's plate amp will solve the problem.

Hopefully, DB will let us know.

BTW, I didn't say that I think PEQ is a bad thing. I've used it for many years in my bass guitar rig. I use it to alter the input signal to achieve a particular sound. Never to compensate for a bad room (nor does any studio I've ever been in).

I find the subject interesting, is all.

With the way different frequencies can have different nodes in different places throughout a room, the theoretical problem I have with placing a single microphone at the listening position and letting the parametric EQ do its thing is: things may get better at that one position and worse elsewhere.¹ In my learning and experience, proper EQ cannot be accomplished in a simple plug-n-play, one-shot-does-it approach.

If I use an analyzer that can give me both:

- Pulse or ‘burst’ gated readings (that avoid room reverberation influences for mid-bass through the upper frequencies²) and

- Gather long-duration data (that includes both direct sound from your speakers and reflecting/reverberating sound in the environment),

- and then take readings in about twenty positions throughout the listening area,

- I can compare all the data gathered and determine which frequencies may need boosting or attenuation (because certain frequencies are too weak or too strong just about everywhere),

- and which frequencies average out throughout the listening area and really don’t need attention in the form of EQ at all.

Having both the pulsed and long-duration readings will help evaluate what in the room may need ‘treatment’ vs. what EQ may be needed. I’m one that believes that if the physical environment is creating a problem, I want to address the environment in a reasonable way before trying to electronically compensate.

¹ As a demonstration of “there’s no single, ideal spot to set up my measurement microphone,” try playing a steady sine-wave tone of 40Hz through your system and slowly walk around the room with an RS SPL meter. In most rooms you’ll find that the sound pressure level from area to area can vary by 6, 10, or more decibels. Don’t have a meter? Try covering one ear and walk around. You’ll likely find spots where the bass is really strong and perhaps a spot or two where it seems to disappear almost entirely. Now change the frequency to 60 Hz. Some of the strong nodes and weak nulls are now in different places. How can any one spot tell me the truth about the whole listening area?

² Pop Quiz: Why would pulsed/gated tests be less effective at lower frequencies in normal home environments?

My apologies if I became too insistent - I'm still willing to listen. Together we know more than any one of us individually!

Tuning a room by EQ is something that needs an experienced person to intrepret the data and make informed adjustments beyond what the raw numbers may say. Just sticking a microphone up at the listening position and pushing the "auto EQ" button doesn't cut it. It is way, way better to deal with room nodes by placement and acoustic treatment than by EQ if at all possible - EQ should be the last resort since it does not correct the problem at all, but just covers up the symptoms. The resonances are still there, storing and releasing energy, but are just decreased in amplitude.

I did think about it, and it is an interesting take on it.

Yeah: pulsed or gated is how the software I use measures it. And yeah: averaged across 3 locations. Not just 1.

One stupid question I have, if this is really true, then that dude Russel Hershelsmann (bad spelling, I know) in SGHT is not correct, and even some of the white papers over on Harman International's site about sub placement aren't correct either.

I like the idea of doing short *and* long duration measurements and maybe trying to find the best compromise in terms of what peaks to go after though.

Good info. I need to think about this some more...


> Pop Quiz: Why would pulsed/gated tests be less effective at lower frequencies in normal home environments?

If the time duration of the test isn't long enough, not a good thing? (I.e., low freqs equal longer wavelengths so you'd need more time for low freqs vs high for just 1 wavelength.)


[This message has been edited by Kevin C Brown (edited May 06, 2004).]

Is this thing on? Sorry, I had a glitch with my password.

I had this BFD hanging around and I wanted to see what benefits it could provide after reading how HT enthusiasts are using it as a parametric eq over on hometheaterforum. I put it in the signal path for my sub feed. I have two M&K VX-100s which are 8” down-firing subs, really compact. I hooked my PC up to the 950 with a coax digital cable for tone generation and used TrueRTA for measurements. With three adjustments, I pulled down two peaks. I also tried to fill in a null knowing full well it wouldn't work. I think it was this null that caused problems. It was about a 10 dB null at 55 and again at 110 and boosting the 55 had the expected effect of boosting 110 (and probably 165 and 220 as well). I then made the mistake of trying a DVD-A (Toy Matinee). All my speaker cals were based on the 950 DACs and DSP as were these frequency measurements, so when I switched to the 6 CH Direct, I was already out of phase. My sub placement always minimized the Outlaw 6 CH phase deal that Kevin brought to light some time ago, meaning flipping the phase switch made little difference. However, with the BFD in the chain it was very obvious that something was wrong, at least while playing Toy Matinee. This disc has significant midrange signal in the sub channel.

The 950 employs a low pass filter at 120 Hz. However, the effect I heard was in the lower mids, not the bass. The attack of the bass, bass drum and lower register of the piano was way off. The fundamentals were fine but the overtones that give the instruments their character weren’t happening. When I flipped the phase switch on the subs, it got better but not as good as it was before. I then pulled the 55 Hz boost out and it mitigated the effect some more. My thought is that the boost aggravated the low mid range that the sub was getting and the reported delay in the BFD was inducing a transient or early reflection effect that was audible in these higher frequencies. I didn’t spend a lot of time on it because I simply didn’t like it.

The bass, however, wasn't too bad at all. My subs are not capable of directly producing low bass. I think this is the case for a lot of people. It seems to me that most subs are engineered to rely on room response. In your average sized room, the room response is unavoidable so why not use it. For this reason, I think using eq to make the subs "play the room" is a useful tool to cover the symptoms as Soundhound describes them. Correcting the “problem” is really not an option for me until I decide to build a dedicated room with the correct dimensions and carefully select new equipment.

I measured my room response with pink noise, a quick sweep and a long sweep. The bass was much improved at the listening position with the BFD online, cutting down the peaks and not trying to fill in the nulls. So, until I convince my wife that the two 18” bass bins in the garage would be a nice addition to our family room, I may try to do more eq adjustments. I don’t have much hope that this will work for multichannel music but I think it’ll be fine for movies.

Finally, one problem I had with the dissertation that Bosso posted is that I think the argument starts to fall apart in the bass range. When you are talking about sounds with long wavelengths, all the subject hears is reflections and reverberations if they are standing or sitting too close to the loudspeaker. Also, wouldn’t the “perfect” speaker have to have a very large bass driver in order to produce 20 Hz anechoically? And, I’m pretty sure that what the subject hears or doesn’t hear in the bass range will be based in part on where he or she is with respect to wavelength. I’m sure this is a gross oversimplification of the physics, but for example, if my math is correct, the subject at 9 feet is sitting at a null for roughly 60 Hz and 120 Hz, so unless those frequencies are reinforced somehow by room response, they’ll be lost or significantly reduced.

I agree with Bosso's post for the mid and upper frequencies but not for the sub frequencies. The post describes how you hear the direct sound first then the reflected sound. This holds true except for the sub which may have the speaker not pointed directly at you (like the LFM-1) so all that you get would be reflected sound. I don't currently us any EQ but in the past I have used it for attenuation only which did seem to help.

Whether the the sub fires front or down, there is still a direct radiated wave that hits you before the room-affected wave.

When you attenuate a frequency band that is peaked because of a room mode, you attenuate the direct radiated wave.

This allows the room affected wave to dominate what you hear, as the direct radiated wave is attenuated well below listening level.

So, what you hear at listening level is delayed by the distance disparity of the the sub placement vs. the mains, by the DSP of the PEQ and by the room mode.

You may play a ruler flat line, but what you're hearing is grossly time-smeared (which doesn't show up in the graph) in relation to the rest of the speakers, which changes the kick drum from a 'tock' to a 'thhwock'

If you draw a circle around the LP so that every point on that circle is equidistant to the LP with the mains, and use that line as the possible placement options for the sub (+/-) 1 or 2 feet, you have the start of accurate sub integration.

If you then chart the mains without the sub, so that you can determine the in-room slope of their roll-off, after the HP filter, then shut the mains off and do the same for the sub, you can tweak the slopes to resemble a proper crossover...that's step #2.

Tweaking the phase comes next, to fine tune time arrival to your ears of the sub + mains.

This is followed by level calibration, at the crossover point. If the sub and mains aren't exactly the same volume at the crossover point...there goes your entire 2 octaves of crossover region.

Many times, IMO, a dip and/or a peak are mistaken for room modes, when in fact, they are mismatched crossover slopes and badly out of phase sub/mains in the crossover region (which stretches from 30ish to 200 Hz. Run several graphs with the only difference being a 20 degree change in phase of the sub each graph and see how wide a bandwidth is affected).

I believe this because I'm +/-3dB from 15-30 Hz., and +/- 2dB from 30 to 200 Hz., with NO EQ. And, I assure you that my room isn't some magic size and shape.

Maybe I was lucky, but it sure sounds good, and that's always convincing evidence.

>> When you are talking about sounds with long wavelengths, all the subject hears is reflections and reverberations if they are standing or sitting too close to the loudspeaker. <<

Yes! This is what "corner" loading is all about. But because the walls of the corner are relatively close to the sub, there still isn't much delay between the initial signal, and at least the 1st reflections. Especially considered wrt to the wavelengths involved.

>> You may play a ruler flat line, but what you're hearing is grossly time-smeared... <<

Yes, but this is why I adjust phase *after* setting up the eq. To regain time alignment.

Plus, I'm only attenuating the single peak by about 5 dB. Not a humungous amount, all things considered.

>> If you draw a circle around the LP so that every point on that circle is equidistant to the LP with the mains ... <<

"LP" stands for ... ??

LP = Listening Position I believe.

KCB,

I've always felt that you have a good handle on sub tweaks. 5dB boost is acceptable, IMO, when using a PEQ.

Do you still use Vandersteen sub(s)? It would be cool to hear your setup someday.

LP=listening position..sorry.

This reminds me...DB asked, in the original question, if changing the distances of the satellite speakers would compensate for the delay. The answer is no. The delays are based on the farthest distance being 0 delay. Since there's no subwoofer distance setting, there's nothing you can do to compensate for a subwoofer being farther away from the listening position (physically or virtually), which it almost always is with a corner placement and/or when using outboard DSP equipment.

This is something the next-gen pre/pro should definitely have that the 950 (and many pre/pro/receivers) doesn't.

I need to think on this some more. The nature of bass waves is highly interdependent and I don't believe the direct wave from our little subwoofers dominates what we hear, Soundhound's system notwithstanding. The drivers are simply too small to support waves that are 20 to 30 ft in length without help.

I'm sure I don't have any crossover problems in my room and the position of the subs is fixed. Nonetheless, I'm going to run the tests that Bosso suggests and see what the charts look like.

Thanks for the answer on the delay settings in the 950.

BoB- (I always really learn stuff from you and SH...)

>> If you draw a circle around the LP so that every point on that circle is equidistant to the LP with the mains, and use that line as the possible placement options for the sub (+/-) 1 or 2 feet, you have the start of accurate sub integration. <<

But, and this is why I have become so fanatical about phase *and* time alignment, this does not compensate for the phase delay introduce by the LPF (= low pass filter) of the pre/pro. Plus, this distance might result in nasty room interactions depending one where those equidistant points falls wrt the room boundaries.

Yes, still have the Vandersteen V2W. (Good memory!) I have been looking at the SVS PB2+, but I just can't get past how big and heavy it is. But ... it has flat freq response to 11 Hz. (I've measured mine, and I'm flat to 21 Hz, down 6 dB at 20 Hz.)

One of the reasons I mention stuff like this thread contains is, I remember debating the merits of SH's stereo subs in a thread (or 2, or 3), using logic and accumulated knowledge.

When I actually TRIED the setup, using the appropriate source material, per SH's suggestions, guess what...it worked, as advertised (and thanks to SH for his long-suffering at my expense).

Some of what gets posted is mistaken thought, some is pure gold. Nothing ventured, nothing gained.