Adventures in Sub Calibration

[noddin0ff]

Well, I've started playing with my new sub, a 12" Titanic MKIII Kit (assembled) from their ebay auctions. I thought I'd post some graphs. My comments are below.
edited out some text here.

http://forums.audioreview.com/attachment.php?attachmentid=1083&stc=1

[kexodusc]

Congrats on the excellent find...that's a serious sub you have there, I build one for my parents a little over a year ago, it totally outclassed my Paradigm PDR-2200 v.2. I saw what you paid for it and I'm a little jealous.
The xmax on that driver is quite high so don't be afraid to turn it up...

[noddin0ff]

Thanks Kex, and thanks for posting your experiences. I've been using them as a rough guide. Yeah, about $3 more and it would've been someone elses! I'm sure I'll be happy with it. Sorry for the long post, I'm just hoping it might help someone else thinking about adding a sub.

So,I've got a Yamaha RX-V800 receiver. The crossover is fixed at 90Hz. My mains are Paradigm Studio 20s with the 470 center and titans for rears. But, all the sub calibration stuff only involves the mains. I got test tones and excel spreadsheet from the following links.
http://www.snapbug.ws/bfd.htm
http://www.snapbug.ws/subtestmodel.xls
http://www.snapbug.ws/sinewaves
I modified the spreadsheet to plot more sets of data. If any one wants it, I can upload assuming there's no copyright issue. The spreadsheet automatically corrects the values for Radio Shack's analog meter. I used C-weighting, slow for the SPL meter settings (C vs A makes a big difference) as suggested on the BFD link. I burned the test tones to an audio CD with iTunes and played them via a CD player w/ analog outs.

I got a BFD from Parts Express, 24 bit, $99.99. And a 6' and a 25' Dayton sub cable for ~$11 and $13, respectively. The cables are very nice. I wish I'd got their interconnects. Also grabbed some $0.90 RCA to 1/4" mono adapters to hook up the BFD. The BFD is in the signal path Receiver LFE -> 6' Cable > BFD > 25' cable > SUB. But the BFD is in bypass mode, so far. It's powered but I haven't engaged the processors yet.

Before taking measurements I rough adjusted the sub and did some A/B toggling the sub on/off and mains small/large via the receiver. It'll take more listening to be certain what's going on, but my initial perceptions of the imaging surprised me. The soundstage was more narrow and focused. I was about to go and disable the center channel, until I thought about it. When the mains are set to "full" and "no sub", the sound seemed more broad and perhaps full or softer. The term 'smearing' came to mind. Perhaps, I now know what 'smear' means. The difference was very clear. It'll take some mental adjustment.

On to the SPL!
I noticed that between 80-100 Hz measurements were very position-dependent (+/- ~db over 12 inches). I just held the meter up at face level in the centered listening position. I don't have a lot of placement options. The sub is centered immediately behind the speaker stand for one of my mains.
First I took measurements with Mains set to Full, No sub: M(f)-SUB. Blue line.
Just for kicks, I repeated this with mains set to small: M(s)-SUB. Grey.
In keeping with Paradigm's specs, I see flat-ish response to about 50Hz and a smooth-ish roll off to ~36, then junk (junk did not sound good). I also see at dip around 89Hz, which is where the x-over will be. The meter is pretty solid down to the mid 50db's.
Small/Large Mains overlapped essentially exactly. I'm guessing that the Main=Small, no sub option is illogical so there is no cut off applied by the receiver if no sub is present.

Then I added the SUB via the receiver menu. I'd already adjusted the sub volume by ear. And had the Sub's Amp x-over set up to the highest 200Hz so it would not engage: M(s)+SUB. Pink line. I forgot to take measurements with Mains=small, LFE=Sub, but turn the sub's power off.

Not too bad. It was reasonably flat and about the same level as my mains down to the expected roll off at 25Hz, where the mains fear to tread. But, there was a +12db bump at 50Hz which was quite audibly noticeable, and some smaller peaks at ~75 and 125.

I thought I'd try to tame the 50Hz peak. I decided to use the EQ on the SUB first, until I figure out the BFD. The sub has a parametric EQ of three dials. One to select the peak frequency, one to select the width of the peak, and one to select the gain/defeat. I dialed in 50Hz, a fairly narrow width (I don't quite get what the numeric value of the second means), and -10db. The results are the Dark Blue line.

Pretty decent guess on taking out the 50hz peak. BUT, a big trough appeared at 100Hz. I don't know but it's probably not just coincidence that the trough is 1 octave up from my target frequency. Since 100Hz is above the x-over, I'm not sure what this means. I narrowed the width of the peak in SubEQ2 (Green Line), which improved some areas. Then lunch was over and I had to go back to work.

I'm wondering if the 100Hz dip is due to the Sub's parametric EQ, a room effect, or some phase issue/interference of the sub with the mains. The sub has a two-position phase switch, and by ear, a lot of base disappeared in the other phase.

If anyone has a suggestion, I'd like to hear it. It'll probably be a while before I can really test out the BFD.