Which sub-bass driver?

Kukulcan

Member
2014-06-27 10:00 am
Let's start from these points:

need to cover 20-180 hz

music (classical and jazz) and movies with LFE channel

modest SPL required, let's say max. 90-100db, but often much less (but good capability at 20 hz)

sealed box, volume ok up to 100 lt.

unlimited dsp capability

1000w amplifier already available

Sound quality is paramount


What kind of driver would you choose??? I have some ideas, but I intentionally give no brand names or driver type

Let's start!
 

eriksquires

Member
2013-05-10 4:11 pm
Hi Kukulcan,

The stopping point for speaker buyers is that high-quality bass requires the right acoustics and placement of the subwoofer. These requirements usually go un-heeded or are too fussy for most. They are simply a lot more trouble than smaller mains speakers are. Having said that, with the patience, it is possible and straightforward.

If sound quality really is paramount, start with measuring your room. Add appropriate bass traps until the energy storage in the room has been controlled.

The biggest mistake bass aficionado's do is go for size and power first. Tune the room and you'll be amazed how little subwoofer (if any ) you need.

Once you have tuned the room, then adding a subwoofer is a piece of cake, and you'll have a much better idea how much you need. Alternatively, you can add all the inches or meters of speaker diameter and watts and it will sound like muddy dog poop.

Let me show you what I mean with a graph of my main speakers in my living room at 3' away. I just took it last night as I was examining the best way to integrate a subwoofer.

Please keep in mind, this is in-room, far field response, this is not quasi anechoic! Meaning, these are great measurements from a single speaker with a 7" woofer.

attachment.php


I can only get results this good due to the pair of bass traps from GIK Acoustics I am using in the corner behind the speaker. Did I mention this was taken with NO subwoofer? The subwoofer was off for this measurement. In fact for music I currently do not use it at all, it only comes on for movies.


attachment.php



After measuring the sub though and using XSim to simulate the effects of adding a subwoofer I decided it was simple and worth trying to integrate the subwoofer by using a mini-DSP. However, this is only simple and easy and a luxury because of the bass traps.

The before results were terrible, and I had so much energy storage in the room my test tools couldn't measure it. :)

So, to summarize. Measure your room first. Apply appropriate treatment then re-visit the entire subwoofer question.

Usually what happens here though is people thuroughly ignore me, spend weeks with cabinets and drivers and crossovers and build something they are never quite happy with. Good luck with either of the paths you choose to take!

Best,


Erik
 

Attachments

  • FR_Speaker_Only.jpg
    FR_Speaker_Only.jpg
    71.5 KB · Views: 511
  • IMG_20160131_085055.jpg
    IMG_20160131_085055.jpg
    77.8 KB · Views: 507

eriksquires

Member
2013-05-10 4:11 pm
Unfortunately your posting comes when I'm in the middle of doing just this, so you will get much more of a reply than you wanted.

My goal when I add the sub is not to add bass to much as to integrate the bass. See the big dips in the speaker response around 100Hz and 70ish Hz? I want to remove those much more than I want to add bass. Now, let's look at the subwoofer response:

attachment.php


This is actually a very good plot for a subwoofer in a living room, however it still has some big peaks. If I were to try to add this to my main speaker response it would accentuate the room modes at 25 and 80ish.

I used XSim to simulate a pair of scenarios. The first where I just add the subwoofer:

attachment.php


The top blue line is the predicted final outcome. You can see both speakers are going to accentuate the same modes.

This is definitely a case where cutting those modes in the speakers and the subwoofer would be ideal. I'm reluctant to do any signal processing on the main speakers though so I will probably use the mini-DSP for the Center and Subwoofer channels instead. I reiterate though, the work I have left to do in adding EQ is so little only because of the bass traps. Also, DSP's rarely handle nulls well at all, an area where bass traps excell.

Where I made of money, I would add a pair of bass traps opposite the existing one's, and with a little luck those peaks and dips would both smooth themselves out.

But now, what if I had heavy duty DSP on the subwoofer, and crossed the main speakers over at 80 Hz? Hmmmm...

attachment.php


Of course, this assumes that I have absolutely perfect sufwoofer response that has been perfectly phase matched.

Anyway, my point is, if you really want ideal bass, you should consider more than just adding inches or watts, but measuring and spending the time to understand your system and room will yield the best possible results.

Best,


Erik

P. S. I used intereferometry to estimate the distance between the speakers as measured from my listening position, so the phase and amplitude matching has been taken into consideration.
 

Attachments

  • FR_Sub_Only.jpg
    FR_Sub_Only.jpg
    70.8 KB · Views: 267
  • Combined_Predicted_No_DSP.jpg
    Combined_Predicted_No_DSP.jpg
    85.5 KB · Views: 267
  • Combined_Predicted_Small_Satelites.jpg
    Combined_Predicted_Small_Satelites.jpg
    82.1 KB · Views: 259
Last edited:
Hi, I think the peaks at 80hz and 25hz, and the corresponding null at 50hz, seen in both the speaker and sub output, is due to the room mode peaks and nulls.
True, the peaks can be trimmed with DSP, but the nature of a room mode means there is still a lot of energy storage and ringing at those frequency. A DSP flattened room response will sound very different compared to a treatment flattened room response due to the modal ringing.
25hz is very tough to treat, so your best bet is the 50hz dip and 80hz peak.

Hope this helps!

Sent from my GT-I9505 using Tapatalk
 

eriksquires

Member
2013-05-10 4:11 pm
Hi Rhapsodee!

You are quite right. That's what I was trying to point out.

Alternatively, not flattening out those peaks makes it impossible to raise the overall bass level to match the rest of the range. However, this is a hybrid approach, relying first on bass traps, secondly on DSP. I'm hoping I can prevent that DSP sound.

I'm going to experiment with this over the next two weeks though. I will only EQ the subwoofer but whether I relegate it to movies only or incorporate it with the main speakers, which will not be DSP'd remains to be heard.

Crossover functions will be in the active stage thanks to my Halo P7 7.1 preamp which does this.

Best,

Erik
 

eriksquires

Member
2013-05-10 4:11 pm
Very sorry Kukulcan!

You are right! You didn't mention it and most who are here asking for help with bass drivers have an aversion to room tuning. :)

Steve has posted some interesting things about bass drivers in the thread about Snell and high mass drivers.

It seems the recommendation there is a driver with high Qm as having better transients than the normally damped drivers.

So, let me ask you, what you don't talk about is the room size, distance, number of drivers, and where you plan to put the sub/driver as well as what the room acoustics are or will be like. All of these have an impact in the choices. Also, with your large cabinet, is a folded horn a possibility?

Sorry again,

Erik
 

eriksquires

Member
2013-05-10 4:11 pm
Without knowing your room, the ScanSpeak 32W/4878T Revelator 13" Subwoofer may fit the bill. The sealed f3 is 45 Hz, but in a room may be perfectly flat to below 20 Hz. Being sealed it will be less stressed by bumping up the low end than a ported sub would be.

However, if you had room / speaker data you might be able to see just how "big" a subwoofer you need. That doesn't seem to be your concern though.


Erik
 

Kukulcan

Member
2014-06-27 10:00 am
Very sorry Kukulcan!



It seems the recommendation there is a driver with high Qm as having better transients than the normally damped drivers.

Hi erik, no problem at all, I'm glad you came back.

Yes exactly, Qms is one of the parameter I'm interested in. I also have the suspect that hi Qms driver, tipically pro audio drivers (or simply big cones), are designed with the proper cone area for their purpose, and not with the "WAF" cone area.. (small nice and tiny cones for slim floorstanders with exaggersted throw).

Now my question is: when Qms is high? 5-7 is enough? Or value of 15-20 are much more desirable? In other words, is it enough avoiding a Qms of 2,5 or the higher, the better?

I also think that very high Qms, and very low Qes (0,2-0,3) usually are associated with a pretty good motor and good suspension (please note, I wrote "associated" and not "imply"). What do you think?


About my room, well I rely much on Time Domain Room Correction, whose capability are incredible for the low frequency (up to 1khz). Basically the software available are: Audiolense, Acourate and Dirac.

Of course a good passive treatment is always very important, but right now I don't want to invest in it, because I'm going to move to a new house.
 

Kukulcan

Member
2014-06-27 10:00 am
Without knowing your room, the ScanSpeak 32W/4878T Revelator 13" Subwoofer may fit the bill. The sealed f3 is 45 Hz, but in a room may be perfectly flat to below 20 Hz. .


Erik

The room is 70 m3, but the room in my next house could big bigger.

That Scan is in the radar! I'm also considering the SEAS L26ROY, but it's just 10" and it's already a bit more "lossy" in terms of suspension than the Scan.
I think the Scan could work more effortlessly. ah.. the budget is about EUR 500,00 I'm in Europe.

On a completely different side, I'm watching stuff like a Precision Devices PD185N02 PD.185N02 - Sub Bass Driver

This is of course a completely different beast, whose max SPL capability I don't need at all. The reason is that in a sealed box (I consider only sealed box) is critically damped in just 70 litres, so a huge cone in an acceptable big cabinet. Of course I should rely on severe EQ to make it flat, but that's not a problem with Audiolense. WinISD says 103 db at 20hz, wow!
Qms >20 Qes about 0.2, an incredible powerful motor... It would work very effertlessly, I suppose that distortion (THD, IMD) would be negligible... what do you think.
I mean, the problem for a passive setup is that this driver simply does not fit, but in an active setup with unlimited dsp possibilities...
 

eriksquires

Member
2013-05-10 4:11 pm
I've just now read about high Qms woofers from Gravesen but have not read any data to back that up. The 7" SS woofers i am using from ScanSpeak have a Qms of 5, which is rather high for a speaker this size, so maybe there is something to it. All I can say is that with the room treatment applied (not really that much, just 2 Sofit traps, easy to move) they measure most excellently and sound terrific.

Why don't you wait until you move? That way you'll be able to measure the speakers in the new room and figure out what kind of a sub you need. You may do really well without one or a smaller 10." Hard to tell now.

I guess my point to all the rest was, get the most out of your current speakers first. I haven't tried time domain correction, but for the soffit traps, they made my speaker drivers grow several inches in diameter. :D If you have the same effect with your room correction then perhaps you have done as much as possible.

My concerns, without data, was that time domain correcting software eats into amplifier power and speaker capabilities.

I still think, measure your current situation first. It doesn't have to be expensive. Get Room EQ Wizard or the cheap Dayton mic for phones and an app. Post your speaker response here, then let's see what's best. :) Integration is better than an absolute solution.

Best,

Erik


Best,


Erik
 

Kukulcan

Member
2014-06-27 10:00 am
I got already the measurements, but not in the PC i'm writing right now.. I have the minidsp mic with calibration file...

My actual sub (B&W 610) is perfectly integrated with the satellites (diy with Accuton mid-woofer and seas tweeter), but I don't like the sound of the sub.

It's muddy! People may find it "musical", until they listen to something else... I tried a Yamaha IS1118, it was mine, but it was not suitable for HT, it's for a club. I had for other reason, then I sold it.

But wow, If you listen to a contrabass played with that Yamaha 18", you feel the pinch of the cords... it sound like: tung/tung/tong

With the B&W it sounds: wung/wung/wong

I hope it describes it... :) the attack is muddy, and it?s not a matter of integration.
On the other hand, the 6.5 inch of the Accuton, can go pretty low, but not loud, it means that you can forget to listen to realistc drums, even at only 70-80 decibel
 

eriksquires

Member
2013-05-10 4:11 pm
It's that muddiness that makes me call sub-woofers the third child of the Devil himself. :D

What I would love to see, from your listening position, the frequency response of each driver separately then together. Disable any high pass filtering on the satellites for this.

Of course, time domain waterfall plots of the bass would be extremely helpful as well.

Best,


Erik
 
looks nice, one of the few pro driver with low Fs. Only the Qes seems a bit high, I would prefer a very low Qes (and high BL/mms ratio), so a very powerful motor ofr a perfect control of the cone movement

It beats the Scan 32w4878 in all of those criteria except Qes for a third of the money.
The LAB12 was specifically designed for use in a large tapped horn sub woofer but works equally well in a relatively tiny sealed box (<30L!) or a reasonably sized ported one.
Would need eq in sealed box though.