Sub 20Hz EQ

[Malice]

With the 1124SP specification at 20Hz -3db, I presume sub 20Hz EQ is a

I've been thinking about an upgrade to a PB12Plus/2 and tuning it for 16Hz because I appear to have significant room gain below 30Hz so a drop in overall SPL for a 16Hz tune would not have a significant effect IMO.

However, if my room gain < 20Hz is significant (> 10db) will I run into trouble with EQ on a sub that is producing a response below the BFD's lower operating spec?

[brucek]

The response of a BFD is quite good actually.

Here's the response of my unit using REW Vers 4.




brucek

[Malice]

Marvellous! Looks fine to me!

[DrWho]

So what's the lowest frequency filter you can employ on the BFD?

[brucek]

So what's the lowest frequency filter you can employ on the BFD?

20Hz center frequency..............

[Malice]

So not much good for taming a 16Hz peak which is greater than a 20Hz peak.

So, I think my the opening question is still valid! We're stuffed below 20Hz?

[brucek]

We're stuffed below 20Hz?

Well, why not and see what the response is.... then you can decide.

brucek

[Hakka]

You could always use a wide filter centered at 20hz to cut the 16hz peak, and then add a narrower filter at maybe 25hz to pull the 20hz are back to where it was.


Hakka.

[Sonnie]

If you really have a problem, you could always look for a used Rane PE17. It will reach down to 10Hz.

[cyberbri]

I don't see how a big rise below 20Hz is a problem, as long as it's room gain and not artificial boosting. The lower you go, the more dB it takes for the signal to be noticeable. So low frequency boosts from room gain should be welcome. I have my sub eq'd flat to 30Hz, where most bass found in music lies, with some small cuts below 30Hz to tame the rise centered around 22Hz. But the peak is still pretty good, which sounds great with movies and video games, and that occasional music track that extends that deep.

[Malice]

cyberbri wrote:

The lower you go, the more dB it takes for the signal to be noticeable. So low frequency boosts from room gain should be welcome.

I thought the point in C-weighting a response was to offset this phenomenon so you could get an "even" and comparable reading across all frequencies. By flattening a response to a C-weighted set of readings, IMO to the ear the low frequencies would sound "as loud" as the higher ones. Dunno, have I missed something?

[brucek]

I thought the point in C-weighting a response

We don't weight the response to C-weight.

The RS meter outputs a C-Weight reading and so we apply the inverse reading in REW to negate it, making it flat.....

The meter doesn't track its C-weight design well enough, so we substitute our own in the form of a calibration file, but the end result is the same - a flat response, not c-weight.

brucek

[Malice]

Must ask Sonnie to find a "The penny's dropped" emoticon!!

[Malice]

And bounced again......

Sorry Ken, I'm being extremely dense here and need your clear mind to point out where my logic is falling down.

Let me explain from a theoretical standpoint.

Let's say one is able to produce a flat 85db signal (absolute SPL) across all the key frequencies of the sub's output, i.e. the perfect sub (absolutely speaking) as measured with no weighting whatsoever.

Now, with the Ratshack meter we have limited options in what we use to measure the SPL. So although there may be limitations in the unit's ability to faithfully recreate a weighted response (for whatever reason), it nonetheless applies a weighted correction factor (C-weighted) to the absolute 85db SPL. So what the meter produces is in fact a set of lower values for any given frequency and dB level.

For instance, at 18 Hz and an absolute SPL of 85db, the C-weighted meter reading will read an adjusted reading of 77.79db or 7.3 db lower, see here for an on-line C-weighted calculator. At 30 Hz the absolute SPL of 85db will produce a c-weighted reading on the meter of 81.71 db, or 3.3 db lower., anyway I think you'll get my drift.

So when these c-weighted reading are read in the REW and c-weighted compensation is applied, lo and behold in the REW you get a flat 85db line. Now in theory this is the ideal sub response (ignore the 80Hz roll off for crossover): flat and true across all frequencies. Nirvana. There is no "EQ" required on the sub because the non-weighted response line is as flat as a ruler.

The point I was trying to make in my other reply was that if in relation to the imperfect ear, if one is trying to produce an experience where the loudness of a 30Hz signal is the same as that for a 100 Hz signal, then because the ear in effect thinks that an absolute SPL of 85db at 30 Hz is perceived to be 3.3db lower than in real life, then it follows that if the ear is to think it is experiencing 85db at 30Hz, the real world absolute level has to be 3.3db higher. If one EQ's a sub to a flat non-weighted target then aren't we understating the what the sub needs to produce in real life for the ear to experience "equal" loudness at all frequencies?

Perhaps this is why people have house curves! Perhaps a study of an ideal "house" curve is actually an inverted c-weighted curve!! Or the fact that people's hearing is different, notably so as one gets older, then an individual's house curve is perhaps a reflection on their own "c-weighting" and the inverse applied to a "flat" EQ of a non-weighted line.

Do I make sense, or I should I call a shrink or a medic?

As with all "arguments" there can be just a single assumption which is taken as fact and when pointed out by the counter argument, all the "theories" go out of the window!

So, why do we try to EQ a sub to a flat non-weighted response when our ears would then perceive the actual output SPL's to be lower in lower frequencies? Or is this a preference thing? Or do the sound producers increase their source levels by an inverse c-weighting so we can EQ to a flat line?

My brain hurts!

[brucek]

Do I make sense, or I should I call a shrink or a medic?

I don't know about that, but you're disregarding the most important element.

Loudness.

The human ear has different frequency response characteristics for different loudness. I know you've seen and read about the Fletcher Munson Equal Loudness curves. As the loudness increases our hearing response flattens and vise-versa.

What would you have as an ideal response then? The only sensible starting point would be a flat response. From there, depending on the room, the loudness, your age, it's all a shoot and personal preference.

brucek

[Malice]

Mmm, I guess C-weighting doesn't take that into account! EDIT : Other than representing a 100 Phon curve However, if one has to have a sensible starting point, and everyone with a sub EQ's to an arbitrary (mostly) 85db (or 75db if you use the Denon test tones!), then IMHO a flat line non-weighted target response is perhaps not the best option. Maybe using A-weighted and ignore compensation into REW may be a better option. I must have a go at this soon!

But as you say it is down to personal preference at the end of the day. Some like it hot!

[Wayne A. Pflughaupt]

Malice wrote:

For instance, at 18 Hz and an absolute SPL of 85db, the C-weighted meter reading will read an adjusted reading of 77.79db or 7.3 db lower, see here for an on-line C-weighted calculator. At 30 Hz the absolute SPL of 85db will produce a c-weighted reading on the meter of 81.71 db, or 3.3 db lower., anyway I think you'll get my drift.

That doesn’t sound right. My understanding is that when the meter is set for A weighting, it “ignores” response below about 500 Hz – in other words, meter response rolls out below that point. Set to C weighting, the meter “sees” response down to about 31.5 Hz.

That said, a C-weighted meter would register a lower reading than correct at 18 Hz, because the meter rolls out at a rate of about 6 dB/octave below 31.5 Hz.

Check the Rane Pro Audio Reference and Hyperphysics sites.

Regards,
Wayne

[DrWho]

Malice wrote:

Or do the sound producers increase their source levels by an inverse c-weighting so we can EQ to a flat line?

Loudness is taken care of in the studio - so a flat line should be the goal of our playback systems.

However, it used to be common practice to filter out the very bottom frequencies - especially during the vinyl days where 30Hz material can get the needle to jump outta the gap. I'm sure some of that has carried over, but it's slowly moving lower in frequency. A house curve would be a way of counteracting the roll off employed in the studio.

Nevertheless, I would think of house curves as tone controls for tweaking the tonal balance of recordings to taste....not as any form of equal loudness compensation.

[Malice]

Thanks for that Mike.

Wayne, I suspect for sub EQ we need a sub-weighting system!

Here's a picture and as short explanation of A & C weighting

Quote

Generally speaking the overall level found from A weighted spectrum correlates well with subjective assessment of loudness. The C weighting curve gives equal emphasis over the normal hearing range from 31.5Hz to 8kHz.

As far as meter roll off/accuracy is concerned, I thought the meter calibration file took care of that in the REW?

Oh well, good debate guys! And thanks. Not sure I am still 100% clear, but why change the habits of a lifetime?

Bob