My modeler (MacSpeakerz by Jon Murphy) says an 8" long vent. Produces a kinda ugly curve.
I can't see where the vents for the 2.8 ft^3 cabinet are but the cabinet that a local used for his stamped basic 12s specified a 6" vent x the material thickness (¾"). We installed a MUCH longer vent for a significant improvement.
tortugaranger, can you give us the detail of your cabinet?
dave
My default starting point for this project was to use AN's plain vanilla cabinet design specs for their 2.8 cubic feet (~79 liter) box with a 6 inch port and no port tube, just the ~0.75 wall thickness. It was never my intent to actually use this as the final cabinet, only to use it to guide next steps. The final cabinet design will be a stacked "boat hull" shape using layers of milled baltic birch. Thus the entire cabinet except for the tops and bottoms will be sanded and clear finished baltic birch edge grain. The walls will be ~1.5 inch thick laminate with periodic cross bracing. Stiff, stiffer, stiffest. You won't be hearing this cabinet, just the driver.
I've run several cabinet design calcs which more or less conclude that the ideal cabinet for this driver would be ~ 12 cubic feet with an F3 well under 30 Hz. A monster size box that is too big to be practical for what I'm looking for. Hence the 2.8 cubic feet box which when you see it standing there in front of you still looks plenty big. 40" tall by 12" wide by 13.5" deep.
With a 6 inch port the design calcs call for a port tube length in the 8-12" range. I currently have no port tube other than the 0.73" baffle itself. This morning I experimented with an 8" quick and dirty port tube made of stiff thick walled paper rolled into a tube reinforced with duct tape. Ugly as sin but fairly solid. Did some with and without listening and my conclusion was the port tube was not noticeably effective. Quite possibly a proper tube would lead to a different conclusion. By my hunch is that while a longer tube may provide incremental benefit it would be rather modest at best.
After further playing around with the miniDSP together the Room Eq Wizard ("REW") software (
http://www.roomeqwizard.com/), I've arrived at the following adjustments and thoughts.
1) Being a full range driver, the 10" Audio Nirvana Classic benefits greatly from the application of baffle step compensation. Using the DSP, a high shelf PEQ starting at 400 Hz with -6 dB gain really did the trick.
2) Whether it's the driver, the room, or combination of both, there was a noticeable boominess in the mid-bass. After running REW and loading the adjustment filters into the DSP, the boominess was completely gone. The filter was a fairly big notch. Huge plus and a snap to implement with REW and the miniDSP.
3) Overall bass satisfaction benefits from a low shelf PEQ boost starting at 100 Hz with a +2 to +3 dB gain. Adjust to suit. Cranking it up further really pours on the bass.
4) Using REW to try to flatten the upper frequency range (>1k and above) of the 10" AN Classic was not at all satisfying. It seemed to just kill the character of this driver. As soon as I turned off this filtering the sound came alive again. My conclusion is that once you take care of the baffle step compensation mentioned in 1) above, no further tweaking is warranted with this driver. As full range drivers go it's pretty smooth. 4) EDITed After further work with the DSP, RTA etc. it's quite clear that de-emphasizing the 2.5-5k range by ~2 dB resulted in enormous improvement by smoothing out a somewhat overly bright presentation in this audible range. This wasn't obvious based on the RTA but was most clearly evident while listening to live tracks and switching this filter in/out of the mix.
Edit: I should add that the room these speakers are in is arguably on the large size. Vaulted ceiling central great room with integral kitchen area. Plus the speakers are at least 3-4 foot out from the wall. In a smaller room with speakers closer to the walls, the need for baffle step compensation might well be far less. However, with full range drivers, a bit of high shelf lowering will likely always be of benefit.
Regards,
Morten