[Fanboyz]

THX seems to go by the -6db point, as do I.

Is that pretty much what everyone else does?

 

[Fanboyz]

anyone?

 

[A9X-308]

Electrically xover freqs are defined by the half voltage or power points of -6dB or -3dB respectively. The frequency chosen will depend upon the driver + enclosure characteristics and what sort of performance you want out of the system. The THX idea of setting the xover at a 'standard' frequency is nice for manufacturers and some consumers so they can mix and match components and know it will all work together, but it may not be the best for overall performance.

 

[tvrgeek]

Set based on system in-room measurements. The crossover is not always symmetrical, and crossover points not always the same. Things to tune. For a common syntax, a crossover is described by the -3dB point and the slope. (among other things outside a basic discussion)

 

[Jim Hef]

Audyssey recognizes the -3dB point, but interestingly, it has chosen the same speaker model at different crossovers for different locations. The sound is wonderful, but I find after using this feature that speaker capabilitly isn't necessarily the correct choice for manual setups. "Smart" systems are a wonderful thing when employed in our newer equipment.

 

[ENiGmA1987]

I set my crossover at the 100Hz mark, but thats just cause I love low end. My crossover is set as "plus" which means only frequencies lower than 100 go to the subs, while all frequencies still go to my main speakers.

 

[Magmatic]

I tend to set my crossover frequency about an octave above the lower corner frequency (f3) of the main front units. I just feel there's less chance of summing errors with the headroom before the speaker rolloff. With my RM7si signatures, this would be in the 90 Hz range. If the room correction program sets it to 80 Hz, I leave it alone. Anything else besides those two numbers in my instance gets changed to 90 or 80 Hz. I'll make a subjective test and pick which ever bias I feel that day.

 

[NekoAudio]

I also take measurements and adjust the crossover until I get a flat frequency response. I've found that going by the published specifications tends to end up with a crossover that is too low, and a resulting dip in measured frequency response.