I got bored last night, so I decided to do a little "real world" data collection with my FW18.0 using CEA-2010 bursts I downloaded from Brent Butterworth's website (20Hz-63Hz) along with Omnimic. If nothing else, I figured it could help me justify an upgrade to a bigger, badder subwoofer
Doesn't look like it will work out that way though...
To preface... according to Brent's page, the bursts are recorded at -1dBFS, and in a standard stereo MP3 file. By my reckoning (and data), that suggests played back on a properly calibrated system, I should read a peak level of 110dB at the seats at full reference output (104dB + 104dB). For this test, I boosted the sub level by a further 10dB, which would peg reference level at a peak of 120dB, which is about 3dB off a theoretical "worst case scenario", and by my reckoning, probably about as bad as a real world soundtrack would really get. Further, I boosted the sub level an additional ~5dB because well... I like it a little hot.
The first run of CEA bursts was run at a master volume level of -25dB from reference, equating to an expected output level of 100dB (110dB+15dB boosting-25dB MV). The results were:
This is the baseline response I run, and consists of cutting a peak in the response around 45Hz, and filling in a dip around 63Hz, and little else.
The second run was taken at a master volume level of -10dB from reference, equating to a nominal level of 115dB. Results were:
Only thing of note was 0.8dB of compression at 63Hz, which seems to be little more than a blip going by the next run at -5dB from reference (120dB nominal level):
Pretty much a perfect 5dB rise across the board relative to the previous run...
So as it relates to my actual usage.... A MV level of -10dB is really about the limit for me these days (usually -20dB or under when the family is present). Needless to say, it just got a little harder for me to justify an upgrade to myself beyond the "but I want it..." factor. As is, it would appear that my system can comfortably deliver those levels with a useful amount of headroom to spare. I did run a few bursts with another 3dB on the MV, and there was still further output to be had. However, the 120dB nominal sweep appears to be close enough to the limits that it's probably best to leave anything further on the table.