Is modeling sealed sub waste of time in car?

1393 Views 23 Replies 9 Participants Last post by sqshoestring, Apr 21, 2009

monkeybutt

Discussion Starter · Apr 20, 2009

I was inquiring on a classified ad that when I modeled this particular sub using T/S parameters I got a much larger box volume of like 2.5 cu ft. for a Qtc of about .82. instead of the sellers recommended .88 cu ft. His reply follows.

Like i posted and PMed, you guys can NOT rely on modelling programs for sealed box woofers in a car! They do NOT account for the car's cabin gain, which changes everything! All those programs do is show you the response charachteristics in an anechoic open-air environment.
Trust me....0.88 cube sealed per driver is the KEY!

This has not been my experiance as the relative response of higher vs lower Qtc alignments have been predictable. With the bump in response from typical cabin gain adding the bump from a higher Qtc alignment seems to muddy up the upper bass and makes the transition to the midbass more difficult and requiring lower LP. A big friggin' 'if' is if the T/S parameters are accurate, but a sealed alignment is pretty forgiving for some production variation compared say to ported.
By the way the sub in question is an RF 12 Audiophile w/ Fs=21 hz, Qts=.38, and VAS=9.182 cu ft.
What do y'all think?

Click to expand...

Save

1 - 20 of 24 Posts

89grand

· #2 · Apr 20, 2009

How much eq power do you have?

If you have a decent eq you can tailor the response of a sealed sub to either dial in a little more low end or dial out some upper bass bloat. It's true that the modeling programs don't tell you what the in car response will be because the software doesn't know unless you measure and input it.

If you don't have much in the way of eq, then you'll want to tailor the box for the desired frequency response.

Save

Lanson

· #3 · Apr 20, 2009

Sure there is cabin gain, but the sub itself needs to be aligned with the "air spring" created by the box. I'm a believer in .707 as a golden ratio, so I use it.

I have found better sound quality by using these size boxes compared with the manufacturer's ratings. I've tried this and proved it to myself in three different situations:

Eclipse SW8000, the infamous LMT sub. A 10" sub that is recommended in a .8 to 1.0 cubic foot box according to Eclipse, but it sounds REALLY boomy and muddy in that box. Changing to a 1.8 cubic foot box smooths it out to .707ish and it sounds flawless.

JL Audio 10W3, according to JL Audio it should be in a .75 cubic foot box, or as low as .60 cubic feet (if I recall..this is the v1 version so its old!) Anyway, I build this first and sure enough it was too boomy and very muddy. I went to double that size, and the response smoothed out and sounded much more like music.

Fi Audio SSD 10's, again put them in a small box, again got the muddiness. Switched to a bigger box (about .7 cubes per sub), and my response was near perfect.

Truth is, cabin gain is tricky but you can generalize it as a boost like a hump centered around ~80hz. You can always EQ it out if you need to, but most like a little extra in that range.

The deeper truth is that .707 has a strong meaning to me, since it is the base number to virtually everything that occurs in nature, from the spiral of the galaxy to the harmonic scale of music. I don't feel as if it is a coincidence that .707 works as a box Q, given the way it prevails through everything.

See less See more

Save

dubbreak

· #4 · Apr 20, 2009

Like i posted and PMed, you guys can NOT rely on modelling programs for sealed box woofers in a car! They do NOT account for the car's cabin gain, which changes everything! All those programs do is show you the response charachteristics in an anechoic open-air environment.
Trust me....0.88 cube sealed per driver is the KEY!

By that logic it doesn't make sense to model any alignment (let alone sealed).

Yes you have to take cabin gain into consideration for overall frequency response, but that doesn't make modeling useless.

Save

monkeybutt

Discussion Starter · #5 · Apr 20, 2009

89grand said:
How much eq power do you have?

If you have a decent eq you can tailor the response of a sealed sub to either dial in a little more low end or dial out some upper bass bloat. It's true that the modeling programs don't tell you what the in car response will be because the software doesn't know unless you measure and input it.

If you don't have much in the way of eq, then you'll want to tailor the box for the desired frequency response.

So if I had adequate EQ ability I can pretty much chuck a sub into any size box that would accomodate my space requirements? Makes me wonder why I have spent so much time lately comparing several of the lauded subs here on the forum when I just could have EQ'd my way to good performance. Still seems to make more sense to choose a sub carefully and optimize w/ install first before tuning w/ EQ.
However, maybe my less is more attitude is holding me back as I do have some EQs in the garage, just never used them. I've been even loathe to use my tone controls, until recently.

Save

monkeybutt

Discussion Starter · #6 · Apr 20, 2009

fourthmeal said:
Sure there is cabin gain, but the sub itself needs to be aligned with the "air spring" created by the box. I'm a believer in .707 as a golden ratio, so I use it...
The deeper truth is that .707 has a strong meaning to me, since it is the base number to virtually everything that occurs in nature, from the spiral of the galaxy to the harmonic scale of music. I don't feel as if it is a coincidence that .707 works as a box Q, given the way it prevails through everything.

Your first statement is what I have experianced w/ regard to good sealed sub performance in car. Qtc of .707 or a somewhat lower or a bit above has sounded optimal for me for any given sub.
However, I have been playing around w/ the DIYMA12 for the past four days and in alignments from .62 to .7 it still sounds a bit bloated and not having good impact when crossed over higher at say 100hz. A lot of folk say they have good performance at a .55 Qtc in a 1 cuft net box. So I dunno.
Your last statement kinda 'mystifies' me re. nature and so forth. Fill us in as it sounds interesting.

Save

89grand

· #7 · Apr 20, 2009

monkeybutt said:
So if I had adequate EQ ability I can pretty much chuck a sub into any size box that would accomodate my space requirements? Makes me wonder why I have spent so much time lately comparing several of the lauded subs here on the forum when I just could have EQ'd my way to good performance. Still seems to make more sense to choose a sub carefully and optimize w/ install first before tuning w/ EQ.
However, maybe my less is more attitude is holding me back as I do have some EQs in the garage, just never used them. I've been even loathe to use my tone controls, until recently.

I don't think I went as far as to say you could chuck a sub into any size box. I meant taking whatever sub you are interested in, you can tailor the response with an eq for a sealed box that is either a little too small or a little too big. Sure eq's have their limits, but exact box size in a sealed sub is not too important as long as you can eq the response.

It's not like there are huge response differences between a sub in say 1 ft3 vs 1.5 ft3. If you have the room, by all means make the enclosure as large as you want, but when space is an issue, you can tailor a smaller box with an eq.

Save

monkeybutt

Discussion Starter · #8 · Apr 20, 2009

89grand said:
I don't think I went as far as to say you could chuck a sub into any size box. I meant taking whatever sub you are interested in, you can tailor the response with an eq for a sealed box that is either a little too small or a little too big. Sure eq's have their limits, but exact box size in a sealed sub is not too important as long as you can eq the response.

It's not like there are huge response differences between a sub in say 1 ft3 vs 1.5 ft3. If you have the room, by all means make the enclosure as large as you want, but when space is an issue, you can tailor a smaller box with an eq.

Okay. However, when tuning a sub several weeks ago a relatively small change from 1.0 to 1.2 cuft made for notably better transient response and tighter sounding bass that played out in two different car systems.
The seller in this case was promoting a box size of .88 cuft giving a Qtc=1.27 :surprised: and I would be amazed if such an alignment of any sub regardless of EQing didn't sound muddy/bloated. However, having been very happy w/ the 10 inch version of this sub in a .6 cuft box, I am suspicious that the whopping RF published VAS=9.182 cu ft might just be wrong. Perhaps that .88 cuft might be the proper size after all.

Save

Lanson

· #9 · Apr 20, 2009

monkeybutt said:
Your first statement is what I have experianced w/ regard to good sealed sub performance in car. Qtc of .707 or a somewhat lower or a bit above has sounded optimal for me for any given sub.
However, I have been playing around w/ the DIYMA12 for the past four days and in alignments from .62 to .7 it still sounds a bit bloated and not having good impact when crossed over higher at say 100hz. A lot of folk say they have good performance at a .55 Qtc in a 1 cuft net box. So I dunno.
Your last statement kinda 'mystifies' me re. nature and so forth. Fill us in as it sounds interesting.

I don't want to take this thread off topic so I'll just post this for now and we can cover it separately if you want to in another thread or in PM.

In nature (and by that I mean the natural laws of the universe as we know them), there are recurring patterns and numbers. Some feel that they are coincidence, but I feel that the force driving the patterns is due to "intelligent energy", or torsion waves as they are known in the Eastern Bloc countries.

At any rate, a Russian scientist took notice of the way plants, trees, shells, etc. all grew, and noticed they follow a distinct pattern. In fact, one of the most obvious of these to us humans is in our own structure. Take note of the distances you see between points in your outstretched arm. Your fingertips, knuckles, palm, wrist, elbow, and shoulder all follow the same pattern mathematically. The pattern you see there is exactly the same (numerically) as the phi spiral the galaxy has, or the spiral of a nautilus shell. This can go on and on, for instance the musical Octave scale uses the same principles exactly, and even the cycles of the stock market follow the same mathematical form. The core to all these is the golden ratio (.707) and the direct mathematical derivatives of that number.

If you want more info on this, I can point you in a few directions (beware that you are stepping into "fringe knowledge" at this point and it can get weird as hell and there is no turning back once you know how it ALL really works), so I'll leave that to you if that's what you want. In the meantime, to satiate your thirst for this info, you can google golden ratio or Fibonacci and see what its all about for a on-the-surface basic view.

See less See more

Save

releasedtruth

· #10 · Apr 20, 2009

I did some calculations based on boxes I've had over the last 10 years and it appears that whether I made them or not, most were in the high .7s to mid .8s so there has to be something to it. All sounded good to me, although I'm no pro.

Save

npdang

· #11 · Apr 20, 2009

It may not offer the absolute results you are looking for, but it does offer an interesting point of comparison does it not? Especially given that vehicle cabin gain is a constant.

The one caveat, and one that I often encounter... is that many drivers have a wide variance in small signal parameters from the stated manuf. spec. You also have to consider how well small signal modeling correlates with large signal performance.

Save

Brian Steele

· #12 · Apr 21, 2009

monkeybutt said:
I was inquiring on a classified ad that when I modeled this particular sub using T/S parameters I got a much larger box volume of like 2.5 cu ft. for a Qtc of about .82. instead of the sellers recommended .88 cu ft. His reply follows.

Like i posted and PMed, you guys can NOT rely on modelling programs for sealed box woofers in a car! They do NOT account for the car's cabin gain, which changes everything! All those programs do is show you the response charachteristics in an anechoic open-air environment.
Trust me....0.88 cube sealed per driver is the KEY!

Click to expand...

This has not been my experiance as the relative response of higher vs lower Qtc alignments have been predictable. With the bump in response from typical cabin gain adding the bump from a higher Qtc alignment seems to muddy up the upper bass and makes the transition to the midbass more difficult and requiring lower LP. A big friggin' 'if' is if the T/S parameters are accurate, but a sealed alignment is pretty forgiving for some production variation compared say to ported.
By the way the sub in question is an RF 12 Audiophile w/ Fs=21 hz, Qts=.38, and VAS=9.182 cu ft.
What do y'all think?

I'm going to guess that something's wrong with the quoted t/s parameters. What's the rated efficiency or sensitivity of that driver? The t/s parameters suggest an efficiency of just under 90dB/1W/1M, which is quite efficient for a car audio subwoofer.

On design via t/s parameters, NPDang's comments should be given serious consideration. Not only do some manufacturers appear to engage in a bit of "specsmanship", it should be noted that t/s parameters are SMALL SIGNAL parameters (when last have you driven a car audio sub at 1W?

), and will shift as (1) the cone moves noticeably, and (2) if the coil gets warmer, both of which can be expected to occur at higher power levels. Coming up with a good car audio subwoofer alignment therefore requires not only considering the ACTUAL t/s parameters, but how you expect the driver to perform at higher power levels, and of course box size limitations as it's got to fit somewhere in the car of course. With that in mind, if I'm considering a sealed alignment for a particular driver, I usually go for a Q between 0.8 and 1.0 (higher Q for drivers with lower Xmax or observed excursion limitations), and fine-tune the results with polyester fiberfill stuffing.

See less See more

Save

monkeybutt

Discussion Starter · #13 · Apr 21, 2009

fourthmeal said:
I don't want to take this thread off topic so I'll just post this for now and we can cover it separately if you want to in another thread or in PM.

In nature (and by that I mean the natural laws of the universe as we know them), there are recurring patterns and numbers. Some feel that they are coincidence, but I feel that the force driving the patterns is due to "intelligent energy", or torsion waves as they are known in the Eastern Bloc countries.

At any rate, a Russian scientist took notice of the way plants, trees, shells, etc. all grew, and noticed they follow a distinct pattern. In fact, one of the most obvious of these to us humans is in our own structure. Take note of the distances you see between points in your outstretched arm. Your fingertips, knuckles, palm, wrist, elbow, and shoulder all follow the same pattern mathematically. The pattern you see there is exactly the same (numerically) as the phi spiral the galaxy has, or the spiral of a nautilus shell. This can go on and on, for instance the musical Octave scale uses the same principles exactly, and even the cycles of the stock market follow the same mathematical form. The core to all these is the golden ratio (.707) and the direct mathematical derivatives of that number.

If you want more info on this, I can point you in a few directions (beware that you are stepping into "fringe knowledge" at this point and it can get weird as hell and there is no turning back once you know how it ALL really works), so I'll leave that to you if that's what you want. In the meantime, to satiate your thirst for this info, you can google golden ratio or Fibonacci and see what its all about for a on-the-surface basic view.

Thanks Fourthmeal. Interesting, I'll check it out - especially that part about the stock market.

Now that you've mentioned it this does sound vaguely familiar.

Save

monkeybutt

Discussion Starter · #14 · Apr 21, 2009

Brian Steele said:
I'm going to guess that something's wrong with the quoted t/s parameters. What's the rated efficiency or sensitivity of that driver? The t/s parameters suggest an efficiency of just under 90dB/1W/1M, which is quite efficient for a car audio subwoofer.

On design via t/s parameters, NPDang's comments should be given serious consideration. Not only do some manufacturers appear to engage in a bit of "specsmanship", it should be noted that t/s parameters are SMALL SIGNAL parameters (when last have you driven a car audio sub at 1W? ), and will shift as (1) the cone moves noticeably, and (2) if the coil gets warmer, both of which can be expected to occur at higher power levels. Coming up with a good car audio subwoofer alignment therefore requires not only considering the ACTUAL t/s parameters, but how you expect the driver to perform at higher power levels, and of course box size limitations as it's got to fit somewhere in the car of course. With that in mind, if I'm considering a sealed alignment for a particular driver, I usually go for a Q between 0.8 and 1.0 (higher Q for drivers with lower Xmax or observed excursion limitations), and fine-tune the results with polyester fiberfill stuffing.

Here are the RF specs. http://www.rockfordfosgate.com/library/datasheets/rfa-412_812-ts.pdf It is the 4 ohm version on the left.

Save

Brian Steele

· #15 · Apr 21, 2009

fourthmeal said:
The core to all these is the golden ratio (.707)

The "golden ratio is" 1.618 (see Golden ratio - Wikipedia, the free encyclopedia) and I've seen it used a lot more with respect to box dimensions rather than box volume.

Save

Lanson

· #16 · Apr 21, 2009

.707 and 1.618 are directly related to each other. I can't remember exactly how (but I will if you want me to.)

I do know they are one and the same when it comes to what they mean in a mathematical sense.

Save

Brian Steele

· #17 · Apr 21, 2009

monkeybutt said:
Here are the RF specs. http://www.rockfordfosgate.com/library/datasheets/rfa-412_812-ts.pdf It is the 4 ohm version on the left.

Hmm, the t/s params check out, though they do incorrectly refer to "efficiency" as "sensitivity".

I'd still suggest measuring them though, as the sensitivity does seem a bit high...

Save

Lanson

· #18 · Apr 21, 2009

...Its been a while since I studied this in massive depth, but if I remember correctly that .707 associates with the golden mean on a 3D level, meaning for all intents and purposes on a spherical level. I'll have to keep digging but it is 100% connected, this much I know. Which is why when you model at .707, you are showing the most area under the curve in output.

Save

chad

· #19 · Apr 21, 2009

npdang said:
It may not offer the absolute results you are looking for, but it does offer an interesting point of comparison does it not? Especially given that vehicle cabin gain is a constant.

The one caveat, and one that I often encounter... is that many drivers have a wide variance in small signal parameters from the stated manuf. spec. You also have to consider how well small signal modeling correlates with large signal performance.

I agree and I'd like to add that ANY data at all is better than NO data at all.

correct, it may not be what it looks like in the car but for repeatable and usable data I feel that modeling must take place to back up claims.

Save

Brian Steele

· #20 · Apr 21, 2009

chad said:
I agree and I'd like to add that ANY data at all is better than NO data at all.

OTOH, Garbage In = Garbage Out. At least with no data you wouldn't be misled into buying a sub that has a low published Q for a vented or 6th order BP alignment, only to find out that it actually has a high Q and isn't suitable for either

Save

1 - 20 of 24 Posts

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 Contributors this Month