You are here

Log in or register to post comments
Red GTi VR6
Red GTi VR6's picture
Offline
Last seen: Never ago
Joined: Jul 11 2006 - 9:25am
3dB down point questions

In my quest to learn ever more about the technical side of audio, I'm trying to enlighten myself more so on the aspect of a '3dB down point'.

As a way to keep my thoughts coherent and to keep a running list of questions and answers that I've gone over in my head, I've written up a document. Yes, I'm a nerd, I accept that.

This is what I have so far, perhaps someone on here would be able to assist me in filling it out more? Or perhaps correct any misconceptions that I might have encountered?

3db Down Point

The point at which a speakers output is 3 dB down from a normal point (0dB).

From this web site: http://recording.songstuff.com/articles.php?selected=54

It is worth noting that the circuitry used to implement filters is not ideal. This means that the gain does not suddenly change from 0 dB to -18 dB as the frequency crosses the threshold, the gain change is progressive. In fact the quoted cut off frequency is the frequency at which the gain reduction is already 3dB below the input amplitude. This point is also referred to as the 3dB Down Point.

When talking about HP and LP filters, it is talking about the frequency at which point those filters are actively filtering.

Many times you create a notch in a filter, a 3dB notch. It

Jeff Wong
Jeff Wong's picture
Offline
Last seen: 3 years 3 months ago
Joined: Sep 6 2005 - 3:28am
Re: 3dB down point questions

Not sure if this will help:

http://home.ease.lsoft.com/scripts/wa.exe?A2=ind0501&L=vidpro&T=0&P=8588

Welcome to the Stereophile forums, BTW.

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm
Re: 3dB down point questions

I guess I'm not really sure what the question is. Are you asking why aren't the -3dB points marked on a frequency response graph? They are. Graphs are laid out to let you mark off how much variance from 0dB there is at any given frequency. Typically the frequency runs along the horizontal axis and the level along the vertical axis. This surely can't be what you are asking.

If you are asking why -3dB points aren't plotted on a frequency repsonse chart for the overall response of the speaker system, they are. When you see a graph which lays out the response measured from the various drivers in the system, you are seeing the roll-off and roll-in of the drivers' response. As the above article indicates, if both the LP and HP filter are operating properly, the two should combine as -3dB and +3dB to make a flat response which sums at 0dB. Since the overall response of the speaker system is what most people are concerned with, this graph isn't of much interest to the consumer. Where the -3dB points fall and how the filters combine is really only of interest to the designer as the subtleties of the dividing networks is resolved.

What is usually of more interest to the consumer is the -6dB point. This indicates the usable low frequency extension of the woofer. At -6dB you can consider the system down sufficiently in level to say that is where no more usable bass response can occur. At the high frequency end, once again -3dB is considered usable response. After this point the mechanics of the driver tend to suggest the driver has reached a point where distortion will increase dramatically and further frequency response extension will degrade quickly.

Should you be interested in how enclosure type affects the roll-off, you will find virtually all ported enclosures have an acoustic roll-off of -24dB per octave. Mass Loaded Transmission Lines are the exception to ported systems and are generally capable of a roll-off which is more similar to an infinite baffle or acoustic suspension enclosure. These systems all roll-off at approximately -12db per octave. Therefore, an acoustic suspension enclosure will exhibit a -3dB point in the lowest octaves at a slower declination than a ported design, either bass reflex or horn loaded. If both types of speaker enclosure state a system resonance at 50Hz, the AS system will play lower frequencies before reaching its -3dB point than a BR enclosure. Given the same low frequency driver, however, the AS enclosure will probably have a higher system resonance. So making comparisons with the same driver isn't easily accomplished. Since system resonance is lower with a typical BR enclosure, and you can pick up an additional 3dB by utilizing the backwave from the driver to supplement the output of the system, BR enclosures have become rather ubiquitous among consumer designs. Twenty five years ago just the opposite was true as the AS system typically has better transient response and lower distortion at similar frequencies when compared to a BR enclosure. At the high frequencies (shortest wavelengths), cabinet type is not as important as cabinet shape and dimension.

.

Red GTi VR6
Red GTi VR6's picture
Offline
Last seen: Never ago
Joined: Jul 11 2006 - 9:25am
Re: 3dB down point questions


Quote:
I guess I'm not really sure what the question is.

I'm asking for clarification on -3dB down points in general. Specifically I've asked regarding 3dB down points on speakers. Why is it a spec to consider when it seems that it would be highly dependent on the enclosure.


Quote:
Are you asking why aren't the -3dB points marked on a frequency response graph? They are. Graphs are laid out to let you mark off how much variance from 0dB there is at any given frequency. Typically the frequency runs along the horizontal axis and the level along the vertical axis. This surely can't be what you are asking.

Sure, they are marked by deduction, but it seems to me that if they were of such importance, they would be market with a pointer/tick/etc to specifically locate them. I do know how to read a graph.


Quote:
Should you be interested in how enclosure type affects the roll-off, you will find virtually all ported enclosures have an acoustic roll-off of -24dB per octave.


This is slightly confusing to me. Reading this the way I have would suggest that all ported enclosures are 4th order...which isn't true?

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm
Re: 3dB down point questions


Quote:
would suggest that all ported enclosures are 4th order...which isn't true?

Correct. A Mass Loaded Quarter Wave (somewhat similar to a transmission line, though essentially a hybrid of TL and BR) will generally roll off at a slower rate. Done properly such an enclosure can actually extend the system response beneath driver resonance much the same as a back loaded horn enclosure. However, the horn and the quarter wave do not share the same roll-off. The MLQW more closely assimilates an infinite baffle with approximately a -12dB per octave roll-out. Otherwise a ported or vented enclosure goes away at a rate of -24dB per octave.

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm
Re: 3dB down point questions

I would say the -3dB point is not anything particular to consider regarding speakers unless you are manipulating the crossover. So there is no reason to mark it on a consumer oriented FR graph. Within a room the position of the speaker relative to reinforcing surfaces will slightly affect the -3dB point so it is somewhat floating according to speaker location. Room resonances and nulls can change the bass response dramatically over just a few feet of distance within the room.

I don't know why you feel it is "highly dependent" on the enclosure type. For the most part, the roll-out is either second or fourth order acoustic depending on whether the out of phase rear wave is allowed to exit the enclosure.

What makes you think -3dB is of "great importance" in the first place? Did someone suggest to you this was true?

Red GTi VR6
Red GTi VR6's picture
Offline
Last seen: Never ago
Joined: Jul 11 2006 - 9:25am
Re: 3dB down point questions


Quote:

I would say the -3dB point is not anything particular to consider regarding speakers unless you are manipulating the crossover.

This is what I needed to know. I was having trouble understanding the significance of something that was so dependent on other things.

I was thinking so my self:

"How can something be a 'standard' of sorts, when the methods used to come to the 'standard' are not solid and unchanging?" (i.e. the 3 dB down point for a speaker will change based on enclosure, etc).


Quote:
What makes you think -3dB is of "great importance" in the first place? Did someone suggest to you this was true?

Based on some readings I've done on the net. But because I was failing to see the significance, I kept researching, and that led me to here, where I've since learned that it's not the case!

  • X
    Enter your Stereophile.com username.
    Enter the password that accompanies your username.
    Loading