I’ve always understood it best to use the high level input. See excerpt from manual on ST series:
The high level input* is designed to accept the stereo (two channel) signals from the speaker
terminals of your receiver, integrated amplifier, and basic amplifiers. This has the advantage
of ensuring that your subwoofer receives exactly the same signal as the main speakers,
which means that the character of the bass from the main system is carried forward into the
sub-bass. This is a very important point and together with the REL’s ABC circuitry, ensures
far superior system integration of the sub-bass with the main system.
*This input is labeled “Unbal Hi Input” on Stentor III and Studio III. This is the most
common or standard recommended means of connecting the sub for two channel use
Unfortunately it doesn’t make sense - it’s REL’s marketing department’s attempt to turn a design weakness into a marketing strength as a USP.
There are two main problems with using the High Level input:
1 the output from the sub isn’t being driven by the amp alone, it’s also being driven by and the sound coloured by the back emf from the speakers (i.e. the effect of the voltage generated by the movement of the bass cones).
2 You then have two amps in series and each adds a group delay, this means that the output from the sub is slower than it would be if it were driven from the preamp. If you try to properly integrate the sub in the time domain as well as the frequency domain (as you should if you want a seamless crossover rather than having a full but ‘detached’ bass response) then the necessary time correction is greater than using the low level input.
From an audio engineering perspective, using a stereo low level input will be always the better option if the sub is appropriately designed to allow this (but REL’s aren’t designed this way).
How long was the cable?
Wht capacitance did it have?
What precautions did you take to protect the SN2’s preamp from being loaded by the cable?
How was it protected form LF interference?
It appears to be a similar distance from the listening position as are the main speakers.
How did you determine the group delay for the sub and (since you haven’t compensated for it acoustically using the sub’s position) how did you include that delay in the feed to the main speakers?
If you haven’t done this then the sub won’t be time aligned and all high/low level arguments are moot as the sub’s crossover isn’t properly integrated anyway!
The best results I’ve had with my N-Sub were with the high level connection from the speakers. I also tried a Naim supplied long DIN to RCA lead on the low level input, and it didn’t sound quite as good, which is what Naim suggest, but the difference was quite small.
If you dont mind - can i push the envelope ( for me ) further
a. The Main speakers themselves are 1/5 out into the room and 1/5 from the side walls -
i saw a youtube video on such a setting and tried this for kicks… and the fonts are sounding very nice indeed…
The room is about 20 ft long by 12 ft wide ( with 1 side…right speaker side… fully open )
b. The sub is placed about 2 inches ahead of the speaker front baffle - i haven’t determined this by any scientific way - save to say its the max i can do given my constraints.
c. I am yet to determine if i can make do with the front speakers pushed back by 4 to 6 inches so that will make the subwoofer front baffle about 6 inches from the main speakers front baffle.
Question …
a. Whats the method to calculate group delay and therefore extrapolate the inches by which the sub should be ahead of the main speakers ?
It’s difficult to calculate the group delay (it depends on the crossover filter, the amp and the mechanical properties of the speaker cone and motor), so unless you already have the figures, it’s generally better to measure it.
Typically for subs, it’ll be in the region of 2-4ms for small subs and 3 to 7ms for larger ones like yours.
The easiest way to get the time alignment right is to look for discontinuities in the minimum phase plot of a frequency sweep. It can also be done from an impulse plot, but that’s harder to interpret; it is however more accurate and can be very useful for fine tuning and/or as a final check.
It’s incredibly difficult to set-up a sub by ear alone.
The best way to do it is to use an instrumentation mic and something like REW to establish physical position and crossover settings. You may also wish to use a DSP to remove the room resonance peaks from the sub’s response (in your case you’ll have quite a big peak in the response at 26.4Hz - corresponding to 20ft, and may have others).
