If you’re willing to spend a little more on processing and power (and accept a couple of other slight compromises), you can push the box size down a bit for the 15, if that helps (technically you could do it with the 12s too, but that doesn’t take away from the underlying relationship between the two in terms of output).

This relies on the fact that the driver is reaching its maximum linear excursion on rather less than the maximum power it can actually handle. So, that means there may be some headroom available (which you could “access” via a parametric EQ in a digital speaker processor combined with a larger amp) to boost what would otherwise be lower bass levels. This has to be done in conjunction with a high pass filter to prevent massive over-excursion if any lower frequencies are present however.

Below are the Output, Excursion and Processor response curves for the 3015LF in a 110l (net) box, or approx. 3.9ft³ tuned to 41Hz. Because of the boost put in to flatten the response, this would need an amp capable of 76V output or 720W @ 8Ω nominal rating – over the continuous rating of the driver but less than the commonly accepted “program” rating that is often used for matching amplifiers to speakers. As the actual impedance is a little less than 8Ω right around the frequency where the boost is applied, this would translate to almost exactly 900W (= the program rating for this 8Ω version of the driver) being applied to the driver any time a full strength signal at that frequency came along in your music. Note the max SPL (below 100Hz) is just as high as the bigger version I posted above, and it still gets to 40Hz.

The actual processor setting used to achieve this would be:

High Pass: 12dB Butterworth 45Hz

Low Pass: 24dB Butterworth 100Hz (this one can pretty much be varied to suit your tops without much effect on the driver at lower frequencies, so is just an example for the time being)

Parametric EQ: 40Hz +6.3dB, Q=3.

Although 6dB (and change) of boost looks like a lot, the fact that the high pass frequency overlaps it keeps the net result to a perfectly manageable amount.

Allowing for decent sized vent and bracing etc, this could be built into a total cabinet size of 500W x 600H x 580D (or 19.68” x 23.62” x 22.83” for the metrically challenged  ) for 15mm (5/8”) timber; I’m not 100% sure I’d trust thinner timber to be good for 120+dB @ 40Hz, at least not without a lot more bracing.

HTH,

David.

I’d agree with the 15 over the 12, unless your logistical requirements really justify more smaller boxes – there are perhaps a few situations which would do that (if your band is a retirement project for a collection of slightly more senior types who want to save what’s left of their backs and don’t mind spending the extra bucks, for example, or if you play in some venues with particularly tight access arrangements, but I think all but the largest of the 15’s mentioned in this thread will be manageable enough in that regard).

Of the options to which you’re narrowing it down, I’d scrap the sealed box straight away. In your similar thread on PSW you did say you favoured LF extension over absolute output, so starting with something with such a high F3 would almost certainly disappoint you.

Of the two vented ones, they’re actually very similar in size – Eminence gives the net box volume as 3.5ft³ or 99l, I presume the 3.89ft³ comes from adding in the volume occupied by the backside of the driver and the ports. That means the biggest difference would be in what tuning frequency might be used to achieve the 52Hz F3 of the first one. There are two ways to achieve it however given that we know Vb: tune high to create a peak above the tuning frequency with a steeper, higher roll-off or tune lower to allow that part of the frequency range above the tuning frequency to sit a little lower – this results in a more gradual roll-off. Of these two options, the latter is the one I’d choose as group delay is lower and the more gradual roll-off should sound more natural to the ear. It will also be more conducive to manipulating the final response with EQ if you wanted to try that later, as you can’t safely apply any boosts much below the tuning frequency without drastically increasing excursion, which would in turn increase distortion and risk damaging the driver.

So, here are the output curves for all three, I’ve bothered to load the parameters for the 4Ω version as I see they are a little different from the 8Ω I was using before. Blue is Sealed, Red is the Eminence 3.5ft³/49Hz and Green is 3.3ft³/40Hz, all driven to 550W. All of the options keep Excursion below the nominal Xmax at this power level until you’re below 40Hz so that shouldn’t be a worry, whichever option you choose. The 40Hz vented box gives up just under 2dB of maximum output to the 49Hz vented, but I think that’s a small enough difference not to worry about it. It is still 4-6dB louder than the sealed box towards the bottom end of its useable range. The shallower roll-off from 80Hz down into the mid 30’s will be more amenable to EQ if needed, and if not EQ’d will definitely be tighter/faster. The second set of curves is the Group Delay (in milliseconds), sealed wins here FWIW, but of the two vented options the lower tuning keeps it lower through all of the useable bandwidth of the speaker, only getting worse below 36Hz.

There is one slight downside to the lower tuned vented option, it will require longer vents which will offset some of the difference in box size and depending on the exact size and shape of vents you use may even require you to fit a bend in them to avoid hitting the back wall of the box. This can be where using a shelf vent built in along the base of the box can help, as the sides of the box create a virtual extension beyond the apparent end of the port, allowing you to use a shorter vent to get the same tuning frequency. Unfortunately the exact amount you can shorten the vent by is the subject of much debate, so there would inevitably be a bit of trial and error involved – the good news is that building to the formulae from any website or modelling program will give you a lower tuning frequency than expected, allowing you to shorten the vent progressively until tuning comes up to the target frequency. If you use “standard” vent length calculations such as on the Eminence designs, you need to make sure that the vent is at least one diameter away from any of the walls or the back of the box to avoid these complications.

Regards,

David.