|
Measuring
frequency response
For
measuring the frequency response of my system, the tone
generator in WinISD was used to generate a specific tone and the
resulting level read from a Radio Shack SPL meter, which was mounted on
a tripod.
When
measuring the subwoofer alone, it was done outside to avoid coloration
from the room. There is a picture,
but I put it off-page to save on loading time (extra 10secs @56k)The following
analysis was all done with the "Blast Furnace", but the results
can be generalised to any sub.
|
Integrating
a subwoofer with your Home Theatre
-
Will
the subwoofer be fed with only the LFE channel, or will a component
of the main channels be mixed in?
-
If
this mixing is to be utilised, what frequency should the amplifier's
crossover be set at? - 100hz, 150hz or 200hz
-
What
setting to use for the subwoofer's own low-pass filter? - in
my case a minimum of 40hz through to a maximum of 170hz
|
Block
diagram of surround amp (Pioneer VSX-D711), which is a typical arrangement
|
I
tested each of the Bass-management options in turn....
- Large
Speakers - no mixing takes place. LFE only is fed to sub out
- Small
speakers crossed over at 100Hz - Main channels have component
below 100hz mixed with LFE channel fed to sub out
- Small
speakers crossed over at 200Hz - Main channels have component
below 200hz mixed with LFE channel fed to sub out
|
|
Points
to note
- Sub-only
was measured at 1metre, outside, to avoid room nodes. This
was before stuffing was added
- The
other two responses are of sub + mains, measured in-room at
seating position
- Subwoofer
low-pass set for 170hz on all graphs
- Graphs
not corrected for Radio Shack SPL meter
|
Consider
the green graph for a moment. The
spike at 120hz is a standing wave within the cabinet which will be discussed
further in a section dealing with resonances. It is useful here because
it shows one of the problems encountered when running multiple filters
together, namely phase response.
The
blue graph shows that this peak is
eliminated, not in the enclosure, but in the signal fed to the sub. To
achieve this, the signal must have a sharp (but unseen here) dip at 120hz.
When you consider this in addition to the large dip in response at 70hz,
it suggests that the signal going to the sub is somewhat nasty.
The
red graph is OK for the 120hz situation,
but it has those nasty dips at 80-100hz and 170hz. The situation is somewhat
muddied by room nodes, which I will investigate a little later, but basically,
the combined responses don't look all that good.
While
were talking about setting up your surround system, here's
a good article on the subject from MK Sound
In
Part II of this section, I do some level testing with different DVDs.
While researching that data, I did a quick check to compare the sub output
using LFE only versus LFE + bass component of main channels. The results
showed that utilising only the LFE channel would not cost me any sub-bass
Considering
the results, and knowing that my front speakers give reasonable response
down to 40hz, I decided to not mix any component of the main channels
with the LFE channel - in other words set my amp to "large"
speakers. This is generally what you will do for a Home Theatre setup
if you have decent main speakers.
The
third question - what setting to use for the subwoofers low-pass filter,
becomes a bit academic now. I want to utilise all the LFE signal, so I
set the filter to it's
highest setting ie 170hz.
|
