The peaks were recorded with an Omnimic about 10 ft from the speakers. Also, actual volume loss in a small room are much less than the theoretical. We were playing back the gunfight in Open Range and the gunshots are what caused the peaks so we weren't listening to the regular content that loud. We only had the volume up all the way for a few minutes. JTR Speakers are rated at 101 dB at 2 volts. The signals in the midbass where the peaks were recorded were probably coherent so you can get up to 9-12 dB of volume increase from both mains and subwoofers. I was using a 40 Hz crossover with a 12 dB/octave slope so the gunshots where right where both IB subs and mains were producing output. Total power available for subs and mains was 6000 watts with possibly double available for peaks.
If you look at the
Reference Capable Speakers Spreadsheet you can see that the JTR Triple 12HT is also rated at 101 dB with 2 watts. It is shown to be able to produce 124.8 dB peaks at 12 ft with 3200 watts. The methodology/accuracy of the spreadsheet seems to have been fairly well vetted at AVSForum.
To recalculate:
101 dB at 1 M = 91.32 dB at mic position (worst case scenario)
This is with 1 watt. To calculate the dB increase between watts you use
dB = 10 x Log (watts1/watts2). If I have 1600 peak watts for a speaker then 10 Log(1/1600)=32.04 dB change. The voltage required can be much less than calculated because peak output doesn't require the input of more voltage due to the amp's transformers.
91.32 + 32.04 = 123.36 dB per speaker
The subs are a little trickier. The subs are 87 dB at 1w (87.3 @ 2.83v). With 8 drivers this equates to 96 dB at 1 w. However, for each doubling wired in parallel you are halving the impedance which gives another 3 dB (per Bill Fitzmaurice). Since there are 4 pairs you should gain another 9 dB by halving the impedance into each pair which would make them have an efficiency of 105 dB.
At 10 ft the subs would be 95.32 dB with 1 watt. If they gain the same as the mains, then they could output 95.32 + 32.04 = 127.36 with going from 1 watt to 1600. At peak power they could increase dB by 39.44 dB which would put them at 134.76 dB. Interestingly, this matches almost exactly what WinISD predicts as maximum output without exceeding driver excursion limits.
If you start at 91.32 dB and get just 6 dB from coherent signals to the subs and mains you need 32 dB of gain which is certainly possible.
If you want to double check with continuous volts, my subwoofer and mains amps are capable of a combined 250 volts continuous draw. 250 volts at 4 ohms = 15,625 watts. I can't do this continuously right now since I haven't added the additional circuit that I plan to add. However, peak watts shouldn't be an issue.