[DSK]

Here's an excerpt:

"High-level connection, using the enclosed cable with the Neutrik Speakon connector,
is always the first choice. By connecting to the amplifier’s speaker outputs the sonic
signature of the entire amplification chain is folded into the signal for the sub, thereby
keeping timing and timbre cues consistent. In other words, the signal sent to the REL
is exactly the same signal sent to the speakers, allowing for seamless integration. This
connection can be made without affecting the performance of the amplifier because
the sub’s amplifier input impedance is 100,000 ohms. This scheme also avoids adding
any detrimental effects by not interposing any additional electronics into the
amplification chain."

This would seem to be advantageous only if your main amp lacks transparency and accuracy ... ie. it has a readily apparent "sonic signature". Presumably, the better the amp the more sonically invisibly it is and the less need to add its colourations to the sub for consistency. I've never had problems integrating decent gear using the sub's line level inputs.

Subwoofer amps are usually not as high quality as main amps and are going to be less accurate, therefore changing the sound more than the main amp, so why add the main amp into the chain as well? Remember too that the sub is used for lower frequencies only, it rolls off soon enough and quickly enough that it doesn't muck up the mids. Even lower quality amps usually get the fundamentals of notes right but are not so accurate with the subtleties like harmonics and timbre etc. The frequencies of the harmonics are higher than those of the fundamentals and are usually above the sub's LP filter crossover ...ie. they are mainly taken care of by the main amp and main speakers. Directional cues largely come from the harmonics rather than the fundamentals. Most people would also agree that speaker cables are more likely to degrade the signal than interconnects. So again, I don't really see why you would want to add the main amp and some speaker cables into the signal path to the subwoofer.

[PaulFolbrecht]

It's a matter of timing and phase, *not* of "adding the main amp's sonic sig" and such things.  I can't say I've ever noticed any of *that* type of thing one way or another.

I've had better luck with *coherency* - probably meaning nothing but phase - using HL inputs.  As I think I pointed out before, though, phase is adjustable anyway.

Frankly, I don't think by any means that HL is the holy grail or the only way to go.  Perhaps I came on too strong before.  In a properly setup system - and ESPECIALLY if you're using a sub as I do which i just to fill in the last 20Hz - you can certainly get excellent results either way.  I've just found there to be less fiddling required for excellent results with HL  inputs.

In the case of the passive linestage or no linestage, HL offers other advantages, which is where this all started.  You do not want your source (thru a passive LS or not) driving 3M ICs and two sinks.  At least I wouldn't.

[DSK]

It's a matter of timing and phase, ...

Normal sub setup (sub location, setting of sub's phase dial) using something like ETF5 software/measurement should address phase.

*not* of "adding the main amp's sonic sig" and such things.  I can't say I've ever noticed any of *that* type of thing one way or another....

But your REL quote clearly says "By connecting to the amplifier’s speaker outputs the sonic
signature of the entire amplification chain is folded into the signal for the sub,". This can only mean the main amp, as anything ahead of that is already in the signal path to the sub regardless of whether you use line level or speaker level inputs.

I've had better luck with *coherency* - probably meaning nothing but phase - using HL inputs.  As I think I pointed out before, though, phase is adjustable anyway.

Again, I'm not sure how this is so given that, as you say, the sub's phase (and location) is adjustable anyway.

The only situation I can think of in which REL's above statement makes some sense, is in a system using a stereotypically 'old school' tube amp that is very lush and euphonic compared to the solid state amp in the sub. In this case, using line level inputs, you may hear a lack of coherence between main speakers and sub across the overlapping region, especially if using a steep LP filter on the sub and/or crossing the sub fairly high.

All that using high level inputs can do is to add any changes to the signal, made by the main amp, to the sub as well. Except in my example above, this will typically be very little from a decent amp and probably less than those made by the sub's amp. As far as phase, the sub location and phase dial setting needs to be optimised using something like ETF5 anyway, so I don't see how high level inputs can simplify sub setup.  High level inputs are certainly handy to have if you are using a passive pre-amp or no pre-amp at all, but in systems with sources or pre-amps with sufficient drive, I can't see any benefit to using them (again, except for my example above).

Paul, please don't feel that I'm attacking you here, I'm just thinking aloud. I'm always happy to be proven wrong and learn something new. Cheers!

[PaulFolbrecht]

But your REL quote clearly says "By connecting to the amplifier’s speaker outputs the sonic
signature of the entire amplification chain is folded into the signal for the sub,". This can only mean the main amp, as anything ahead of that is already in the signal path to the sub regardless of whether you use line level or speaker level inputs.

Of course.  I interepreted this to mean phase & timing, nothing more.

But, before we continue, let me clarify (or adjust) my position.  After reconsideration, I admit that to some extent I'd put some blind faith previously in REL's recommendation, considering that they're considered by many to be the masters of the subwoofer.  I'd found that HL inputs do give excellent results, and I'm using them now in my main system (which is the only one with a sub) as a matter of fact.

I am biased by the fact that I'm really no longer interested in active preamps.  This is a large consideration!  Although I know that people do drive subs through long ICs with TVCs, I'd rather not.  And I'd rather not use an active pre - I'm done with them, most likely.

Anyway - the transformer linestage certainly steers me towards using HL inputs for my sub.  Again, in the past, experimenting with both, I'd used HL out of pure preference anyway, but - you and Mr. Reynolds seem to be more knowledgable in the are of line-level inputs and EQ.

To summarize - I'd only argue, at this point, that it's possible to get stellar results with HL inputs, and frankly I can't imagine better results the other way.  I really can't.  But I would not argue that HL is superior and I think both of you have something in pointing out the holes in the quote from REL.

And, again, I'm talking about a sub that comes in at 40Hz.  If it was 100Hz, there'd be more to gain or lose in general.

The only situation I can think of in which REL's above statement makes some sense, is in a system using a stereotypically 'old school' tube amp that is very lush and euphonic compared to the solid state amp in the sub. In this case, using line level inputs, you may hear a lack of coherence between main speakers and sub across the overlapping region, especially if using a steep LP filter on the sub and/or crossing the sub fairly high.

Hey, do such "old school" tube amps really exist?  Well, I know they do, but I've never known anybody who uses one.  It seems there are so many tube amp stereotypes based on this type of thing but nobody ever encounters it in practice...

All that using high level inputs can do is to add any changes to the signal, made by the main amp, to the sub as well. Except in my example above, this will typically be very little from a decent amp and probably less than those made by the sub's amp. As far as phase, the sub location and phase dial setting needs to be optimised using something like ETF5 anyway, so I don't see how high level inputs can simplify sub setup.  High level inputs are certainly handy to have if you are using a passive pre-amp or no pre-amp at all, but in systems with sources or pre-amps with sufficient drive, I can't see any benefit to using them (again, except for my example above).

Well, now it sounds like we agree completely.  Again - I shouldn't have quoted REL as caviliarly as I did.  I did so in response to Bob telling me that my comment about coherency & HL inputs didn't make sense.

Last comment about phase - if the sub is near the main speakers and firing in the same direction, phase will likely  need no adjustment with HL inputs.  If it's corner-loaded, then of course it will.  The ingenious Velodyne I have now does it all automatically anyway, so, yes, I suppose it doesn't matter.

Paul, please don't feel that I'm attacking you here, I'm just thinking aloud. I'm always happy to be proven wrong and learn something new. Cheers!

Ah, well, my skin is thicker than that, but it sounds like we just about completely agree anyway.

[andyr]

Andy,

You may find this article of interest: http://www.mkprofessional.com/bass_mgmt.htm

When an active external xover (bass management controller) is used between the preamp and amp, you have the greatest chance of obtaining a seamless integration of the sub with your main speakers. The other advantages of using a BMC is that you are now biamping with the sub which will lessen distortion of the main amps and main speakers. It's really a win-win.

Just my $0.02.

-- Bob

Hi Bob,

Thanks for that link - I run a 3-way active system at the moment (highly tweaked - some have said butchered!! - MG IIIAs!  ) and so, yes, I would agree that handling the cross to a sub by another LP/HP active crossover is the way to go.

Now that article said "Bass Management psychoacoustically works because the ear-brain mechanism cannot detect direction at low frequencies, but takes its directional cues from the harmonics of the low frequency sound. In the monitoring process, frequencies below 80 Hz are redirected to the subwoofer. Frequencies above 80Hz are sent to the desired speaker. During the hearing process, our hearing mechanism integrates the sound into the correct spatial auditory image."

While, again, I would agree with this I personally think that a single sub is not the ideal ... a pair of subs means you can balance up room accoustic "defects" to get a more even sound - and bass nodes get worse the lower you go in frequency, so this can become a real problem with implementing a sub!

Also, I would point out that just saying "frequencies below 80 Hz are redirected to the subwoofer" is a bit meaningless as we don't know whether they are talking about the -3dB point being 80Hz or the -6dB point!     I personally would not like a sub in my "serious listening" system that had a -3dB frequency at 80Hz as male bass voices would still be able to reach the sub.  A -6dB point of 80Hz for the sub LP roll-off is tolerable but when and if I ever implement subs with my Maggies, I'll put the -6dB point down at 60Hz, max.

Regards,

Andy

[Bob Reynolds]

While, again, I would agree with this I personally think that a single sub is not the ideal ... a pair of subs means you can balance up room accoustic "defects" to get a more even sound - and bass nodes get worse the lower you go in frequency, so this can become a real problem with implementing a sub!

Hi Andy,

The single/multiple subs issue seems to be one of the true questions in audio. There is a difference between what is theoretically possible and what is practical. I believe it has been shown (paper published by a researcher at Harmon, which I've not seen, but read references to) that multiple subs will give the most uniform in room response. The stipulation is that there are no restrictions on placement within the room. In practice, that stipulation usually does not hold. I believe that M&K's stance is based on real world observations in many studios and I think the same observations hold for the typical home listening room, which are: finding one optimal location for a sub is difficult; finding two or more optimal locations is almost impossible.

Thus, if even that one optimal location can not be utilized, then placing the sub tight in a corner exictes all the room modes and will average out the response. This does not guarantee anything at the listening position, but seems to work well enough in practice. M&K also suggests that if higher levels are required than a single sub can provide, then stack another sub on top at the same location. It should be noted that in both cases, the sub(s) are being sent a mono signal.

Also, I would point out that just saying "frequencies below 80 Hz are redirected to the subwoofer" is a bit meaningless as we don't know whether they are talking about the -3dB point being 80Hz or the -6dB point!     I personally would not like a sub in my "serious listening" system that had a -3dB frequency at 80Hz as male bass voices would still be able to reach the sub.  A -6dB point of 80Hz for the sub LP roll-off is tolerable but when and if I ever implement subs with my Maggies, I'll put the -6dB point down at 60Hz, max.

That's a good point and I posted that question on the asylum. The consensus from the responses I received is that the -3dB point should be assumed when not specified. And the slope of the filter makes a big difference as well. Picking a xover point is always debatable. The THX specification chose 80 Hz; probably a reasonable compromise.

Question? When you talk about two subs are you speaking of stereo subs or two mono subs?

-- Bob

[andyr]

Hi Bob,

Thanks for your reply.  Yes, I guess I'd have to agree with your point that finding one optimal location for a sub is difficult; finding two or more optimal locations is almost impossible!!   

And I'd agree that you assume a published frequency roll-off point is the -3dB point unless specifically stated ... only 12 & 24dB L-R filters seem to be defined in terms of their -6dB point.

And when I talk about two subs am I speaking of stereo subs or two mono subs?

Well I actually don't know which is the correct term!!     When I say "2 subs" I mean one near the R speaker (or in the R corner) which is fed the R channel and a second near the L speaker (or in the L corner) which is fed the L channel!!   

I assume this is really two mono subs??  Or is a "mono sub" by definition one that is fed both the R & L channels?   

Regards,

Andy

[Bob Reynolds]

And when I talk about two subs am I speaking of stereo subs or two mono subs?

Well I actually don't know which is the correct term!!     When I say "2 subs" I mean one near the R speaker (or in the R corner) which is fed the R channel and a second near the L speaker (or in the L corner) which is fed the L channel!!   

I assume this is really two mono subs??  Or is a "mono sub" by definition one that is fed both the R & L channels?   

Good catch, Andy. Boy, semantics can be tricky, eh?

So you will use two subs in a "stereo" configuration just like your main left and right speakers. I've always wondered about that. Assuming that we can not localize sound below 100 Hz or so, what are the advantages of running two subs in this manner? I'm obviously missing something.

-- Bob

[PaulFolbrecht]

So you will use two subs in a "stereo" configuration just like your main left and right speakers. I've always wondered about that. Assuming that we can not localize sound below 100 Hz or so, what are the advantages of running two subs in this manner? I'm obviously missing something.

Some say that 1 sub will almost always have dead spots in the room but with 2 you are assured complete coverage (no troughs).  Properly setup, that is.  Also, you can then corner-load them both, of course, and get some very serious energy if you're really into that.

I've certainly never felt the need for more than one. 

[andyr]

So you will use two subs in a "stereo" configuration just like your main left and right speakers. I've always wondered about that. Assuming that we can not localize sound below 100 Hz or so, what are the advantages of running two subs in this manner? I'm obviously missing something.

-- Bob

Paul just provided the same answer I would've!   

Regards,

Andy

[Bob Reynolds]

Paul and Andy,

I was curious about these two configurations for two subs:

1) left channel going to one sub and right channel going to the other sub -- "stereo" configuration

versus

2) L+R going to one sub and L+R going to the other sub -- "mono" configuration

I understood from Andy that 1) is how he would use two subs. I'm curious about the advantages/disadvantages of these two scenarios.

[woodsyi]

I could locate my sub running 50 Hz lowpass when it was smack in the middle between the two speakers. (where the equipment rack is now)   I got a second one and played around with the positions.  This is what I settled with.  Now I can't localize the subs. http://www.audiocircle.com/index.php?action=gallery;area=browse;album=345&pos=7

Since this picture, I flipped the subs to have passive radiators firing out to the walls with the bottom plate removed.  I put 4" GIK panels in front of the wall to control the reflection from the passive.  The woofers on the sub and the speakers are hopefully in line and in phase. 

[PaulFolbrecht]

I would expect the differences in practice to be very small - how much channel-specific information is there below 100Hz - but I'd do it the 1st way if I ever had two subs.

[PaulFolbrecht]

woodsi - my single sub is located behind the left speaker, about 3' from the corner, crosses over at 40 hz (6dB), and I definitely cannot localize it at all by it's output.  FWIW.

My room is "open" so I have very little in the way of standing wave problems.

[gitarretyp]

I could locate my sub running 50 Hz lowpass when it was smack in the middle between the two speakers. (where the equipment rack is now)

Woodsyi, did you try facing the drivers in the sub along the front wall rather than straight at you? I've found this greatly improves the localizability of the sub (my guess would be that it further reduces the audibility of high frequencies and harmonics).

[woodsyi]

Paul,

I thought I could get away with 1.  I also tried different spots on the side but coudn't make it disappear.  I still think low bass is non directional --just not in my room-- and placing  it out in the middle of the room was out of question due to WAF.  She is cool with me doing all kinds of things to the room but she drew the line at a sub in the middle of the room. 

[woodsyi]

I could locate my sub running 50 Hz lowpass when it was smack in the middle between the two speakers. (where the equipment rack is now)

Woodsyi, did you try facing the drivers in the sub along the front wall rather than straight at you? I've found this greatly improves the localizability of the sub (my guess would be that it further reduces the audibility of high frequencies and harmonics).

Never tried that.  I guess it would be kind of a dipole sub idea with the woofers on my speakers firing forward.  Something to try in the future. 

[Bob Reynolds]

Another thing to remember about the setup of a sub that affects localization is level balancing. Regardless of the xover setting if the levels aren't matched the sub will either be ineffective or call attention to itself.

M&K provides test tones and their LFE4 bass management controller has level controls for making this calibration. I found that in my room with the furniture arrangement there was a 1.5 dB imbalance between Left & Right and about 1.0 dB imbalance with the sub.

[Den]

. . .I still think low bass is non directional --just not in my room--

I think it´s the same in most rooms.
The 'non-directional' theory is based on the fact that, below a certain frequency, the sound wave is larger than the human head.  Since the wavelength is longer than the distance between your ears, you are unable to triangulate on the location of the source.  Hence the THX spec of 80Hz for the crossover point to the sub.  In practice, however, most folks can sense/feel from where the bass energy is emanating. 
Another nice theory that attempts to dictate what we can and cannot 'hear' bites the dust. . .

[Gordy]

Since the wavelength is longer than the distance between your ears, you are unable to triangulate on the location of the source. 

Using that theory, anything below 1500hz would be nondirectional